[{"kind":"page","title":"Genezio documentation","heading":"","section":"Documentation","crumbs":["Documentation","Start here","Genezio documentation"],"url":"/docs","text":"Learn how Genezio measures and improves the perception of your brand in ChatGPT, Gemini, Perplexity, Claude and other AI answer engines. Genezio measures how AI answer engines mention, recommend and cite your brand, and it helps you improve these results. This documentation tells you how to configure Genezio, how to read its metrics, and how to use its data from your own code or from an AI assistant.","keywords":""},{"kind":"section","title":"Genezio documentation","heading":"Start with Genezio","section":"Documentation","crumbs":["Start here","Genezio documentation","Start with Genezio"],"url":"/docs#start-with-genezio","text":"Create your brand, topics and scenarios, and get your first results. Learn what AI visibility is, and why it is important for your brand. Learn how Genezio asks answer engines questions, as a real customer does. Read AI Visibility %, AI Recommendations % and Share of Voice correctly.","keywords":""},{"kind":"section","title":"Genezio documentation","heading":"Understand AI search","section":"Documentation","crumbs":["Start here","Genezio documentation","Understand AI search"],"url":"/docs#understand-ai-search","text":"Answer engines do not show a list of links. They write one answer, and they select the brands and the sources that go into it. These pages explain how they do this: - How LLM search works - Query fanouts explained - How LLMs select sources - How AI citations work","keywords":""},{"kind":"section","title":"Genezio documentation","heading":"Improve your AI visibility","section":"Documentation","crumbs":["Start here","Genezio documentation","Improve your AI visibility"],"url":"/docs#improve-your-ai-visibility","text":"Find the gaps where competitors win and your brand does not show. Write the content that answer engines read and cite. See which of your products the AI shopping answers show. Ask questions about your data, and get the cause of each change. Write the goals of your brand, and see each day where each goal is. Follow a full workflow from the first measurement to the result.","keywords":""},{"kind":"section","title":"Genezio documentation","heading":"Use Genezio from your code or from an AI assistant","section":"Documentation","crumbs":["Start here","Genezio documentation","Use Genezio from your code or from an AI assistant"],"url":"/docs#use-genezio-from-your-code-or-from-an-ai-assistant","text":"Read the AI visibility of your brands, and change their setup, with a REST API. Connect Claude, ChatGPT, Cursor or another MCP client to your Genezio data.","keywords":""},{"kind":"page","title":"Setup guide","heading":"","section":"Documentation","crumbs":["Documentation","Getting started","Setup guide"],"url":"/docs/getting-started/setup-guide","text":"Follow each step of the Genezio setup: create your account, confirm your brand, topics, and scenarios, and measure AI visibility on ChatGPT, Gemini, and more. This Genezio setup guide gives you each step from the sign-up to your first AI visibility results. The setup makes the data that Genezio needs to measure the AI Recommendations and the AI Visibility of your brand on answer engines such as ChatGPT, Claude, Gemini, and Perplexity. The setup takes only some minutes. The result is a full set of topics and scenarios. Genezio uses them to run conversations with answer engines.","keywords":""},{"kind":"section","title":"Setup guide","heading":"Step 1: Create your account","section":"Documentation","crumbs":["Getting started","Setup guide","Step 1: Create your account"],"url":"/docs/getting-started/setup-guide#step-1-create-your-account","text":"To start, go to the Genezio sign-up page. For the available sign-in methods, read Sign-in and SSO options. During the sign-up, Genezio asks you for some basic data about your brand: - Brand name - Website URL - Main customer country - Main customer language With this data, Genezio knows the market and the audience of your brand.","keywords":""},{"kind":"section","title":"Setup guide","heading":"Step 2: Confirm your brand description","section":"Documentation","crumbs":["Getting started","Setup guide","Step 2: Confirm your brand description"],"url":"/docs/getting-started/setup-guide#step-2-confirm-your-brand-description","text":"After you send the initial data, Genezio automatically makes a brand description. This description gives a summary of these items: - What your brand does - The category of your brand - The type of customers of your brand Genezio shows you the description for review. You can do one of these actions: - Confirm the description if it is correct. - Edit the description if it is necessary. When you confirm this description, Genezio knows how to show your brand in conversations with answer engines. For the contents of a good description, read Write a precise brand description.","keywords":""},{"kind":"section","title":"Setup guide","heading":"Step 3: Confirm the customer profile","section":"Documentation","crumbs":["Getting started","Setup guide","Step 3: Confirm the customer profile"],"url":"/docs/getting-started/setup-guide#step-3-confirm-the-customer-profile","text":"Then, Genezio makes a customer profile description from your brand and your market. This profile shows the typical person who can use an answer engine to look for solutions such as yours. The description can include these details: - The type of user - The goals of the user - The problems that the user tries to solve You can examine and edit this description before you confirm it. In the subsequent steps, Genezio uses this profile to simulate realistic questions of users.","keywords":""},{"kind":"section","title":"Setup guide","heading":"Step 4: Confirm topics","section":"Documentation","crumbs":["Getting started","Setup guide","Step 4: Confirm topics"],"url":"/docs/getting-started/setup-guide#step-4-confirm-topics","text":"After you confirm the customer profile, Genezio makes a set of topics. A topic is an area where users can ask answer engines questions about your brand or your category. These are examples of topics: - CRM for startups - sales automation tools - marketing automation platforms You can examine, edit, or remove topics before you confirm them. The topics set the areas where Genezio measures AI Recommendations and AI Visibility.","keywords":""},{"kind":"section","title":"Setup guide","heading":"Step 5: Confirm scenarios","section":"Documentation","crumbs":["Getting started","Setup guide","Step 5: Confirm scenarios"],"url":"/docs/getting-started/setup-guide#step-5-confirm-scenarios","text":"For each topic, Genezio makes prompts and scenarios that simulate different intents of users. The Genezio agent types use them in different ways.","keywords":""},{"kind":"section","title":"Setup guide","heading":"Prompter Agent","section":"Documentation","crumbs":["Getting started","Setup guide","Prompter Agent"],"url":"/docs/getting-started/setup-guide#prompter-agent","text":"With the Prompter Agent, Genezio sends the text to the answer engine as a direct prompt. The Prompter Agent does not use scenarios. The other agent types use scenarios. Example:","keywords":"What CRM should a startup use?"},{"kind":"section","title":"Setup guide","heading":"Recommender Agent","section":"Documentation","crumbs":["Getting started","Setup guide","Recommender Agent"],"url":"/docs/getting-started/setup-guide#recommender-agent","text":"This agent type uses scenarios. It looks at direct requests for a recommendation. Example:","keywords":"Recommend CRM tools for early-stage startups."},{"kind":"section","title":"Setup guide","heading":"Introspector Agent","section":"Documentation","crumbs":["Getting started","Setup guide","Introspector Agent"],"url":"/docs/getting-started/setup-guide#introspector-agent","text":"This agent type uses scenarios. It looks at how answer engines understand a specific brand. Example: You can examine and edit these items before you confirm them. Genezio also supports comparer scenarios. For example, a comparer scenario compares two brands. Usually, you add comparer scenarios later, during a deeper analysis. For tips on good scenarios, read Write effective scenarios.","keywords":"What is HubSpot used for?"},{"kind":"section","title":"Setup guide","heading":"Step 6: Start your free trial","section":"Documentation","crumbs":["Getting started","Setup guide","Step 6: Start your free trial"],"url":"/docs/getting-started/setup-guide#step-6-start-your-free-trial","text":"After you confirm your scenarios, your free trial starts. Then, Genezio automatically starts to run conversations with answer engines in the background. These conversations simulate users who ask the scenarios that you confirmed.","keywords":""},{"kind":"section","title":"Setup guide","heading":"Step 7: Watch the conversations in real time","section":"Documentation","crumbs":["Getting started","Setup guide","Step 7: Watch the conversations in real time"],"url":"/docs/getting-started/setup-guide#step-7-watch-the-conversations-in-real-time","text":"While Genezio runs the conversations, you can see them in real time in the Genezio interface. For each conversation, Genezio does these actions: - It sends the prompt of the scenario to the answer engine. - It records the answer. - It extracts the mentions of brands. - It finds the citations and the sources. - It extracts the perceptions (claims) and examines their accuracy. This process makes the dataset for the calculation of AI Recommendations and AI Visibility.","keywords":""},{"kind":"section","title":"Setup guide","heading":"Step 8: Explore your Brand Overview","section":"Documentation","crumbs":["Getting started","Setup guide","Step 8: Explore your Brand Overview"],"url":"/docs/getting-started/setup-guide#step-8-explore-your-brand-overview","text":"When the conversations start to run, Genezio opens your Brand Overview page. This dashboard gives a high-level view of your AI Recommendations and AI Visibility. On this page, you can do these actions: - See how frequently your brand appears in the answers of answer engines. - See the competitors that appear in the same conversations. - Explore the sources that have an effect on the answers. - Analyze the visibility on different topics. Then, you can examine specific conversations, topics, or insights. This shows you how answer engines show your brand. For a plan of your first days, read Your first week with Genezio.","keywords":""},{"kind":"section","title":"Setup guide","heading":"Tools that make the setup safer","section":"Documentation","crumbs":["Getting started","Setup guide","Tools that make the setup safer"],"url":"/docs/getting-started/setup-guide#tools-that-make-the-setup-safer","text":"Genezio has some functions that make the setup safer and easier to follow: - Recycle Bin: Genezio does not immediately lose the topics and scenarios that you delete. If you remove an item by mistake, you can restore it from the Recycle Bin. - Brand language: During the setup, you can set the language of your brand. Genezio then makes reports such as perceptions, SWOT, and AI insights in that language. - Predictable order and auto-scroll: New topics and scenarios appear in a predictable order. The view automatically scrolls to them. Thus, you immediately see the items that you added.","keywords":""},{"kind":"section","title":"Setup guide","heading":"In the API","section":"Documentation","crumbs":["Getting started","Setup guide","In the API"],"url":"/docs/getting-started/setup-guide#in-the-api","text":"The public API reads and changes the topics in List topics and the scenarios in List scenarios.","keywords":""},{"kind":"page","title":"Create an account","heading":"","section":"Documentation","crumbs":["Documentation","Getting started","Create an account"],"url":"/docs/getting-started/create-an-account","text":"Create your Genezio account, learn which actions only the owner can do, sign in with email or SSO, and give your team access to all brands or some brands. An account holds your brands, your team, and your plan. Most companies need only one account.","keywords":""},{"kind":"section","title":"Create an account","heading":"Sign up","section":"Documentation","crumbs":["Getting started","Create an account","Sign up"],"url":"/docs/getting-started/create-an-account#sign-up","text":"Create an account from the Genezio sign-up page. The person who creates the account becomes its owner. This is important later, because only owners can do some actions, for example to manage API keys and to change the plan. If your company already has an account, do not create a second account. Ask an owner to invite you. Two accounts give two sets of brands and two sets of numbers, and you cannot compare them.","keywords":""},{"kind":"section","title":"Create an account","heading":"Sign in","section":"Documentation","crumbs":["Getting started","Create an account","Sign in"],"url":"/docs/getting-started/create-an-account#sign-in","text":"By default, you sign in with an email and a password. Larger organizations can sign in through their own identity provider. Refer to Enterprise SSO and SCIM.","keywords":""},{"kind":"section","title":"Create an account","heading":"Invite your team","section":"Documentation","crumbs":["Getting started","Create an account","Invite your team"],"url":"/docs/getting-started/create-an-account#invite-your-team","text":"Invite colleagues from Settings → Users. You can give a member access to the full account, or limit the member to specific brands. Use the brand limit in these cases: - An agency gives a client access only to the brand of that client. - A regional team must see its own market and not the other markets. - An external contractor works on one product line. A member with access to some brands sees only these brands in all parts of the product.","keywords":""},{"kind":"section","title":"Create an account","heading":"Next step","section":"Documentation","crumbs":["Getting started","Create an account","Next step"],"url":"/docs/getting-started/create-an-account#next-step","text":"Create your first brand. All data in Genezio belongs to a brand. Thus, you cannot set up other items before you have a brand.","keywords":""},{"kind":"page","title":"Sign-in and SSO options","heading":"","section":"Documentation","crumbs":["Documentation","Getting started","Sign-in and SSO options"],"url":"/docs/getting-started/sign-in-and-sso-options","text":"Sign in to Genezio with email, Google, Microsoft, or enterprise SSO (SAML). Learn how to sign up, accept a team invitation, and set up SSO and SCIM. Genezio has four sign-in and SSO options: an email and a password, a Google account, a Microsoft account, or the single sign-on (SSO) of your organization. This guide tells you how to sign up, how to join a team that invited you, and what enterprise SSO gives you. Use the sign-in method that agrees with the usual procedures of your team.","keywords":""},{"kind":"section","title":"Sign-in and SSO options","heading":"Available sign-in methods","section":"Documentation","crumbs":["Getting started","Sign-in and SSO options","Available sign-in methods"],"url":"/docs/getting-started/sign-in-and-sso-options#available-sign-in-methods","text":"- Email and password: Create an account with your email address and a password. Before a new account becomes active, you must verify the email address. - Google: Sign in with your Google account. - Microsoft / Entra: Sign in with your Microsoft (Entra) account. This method operates on mobile devices and on desktop computers. - Enterprise SSO (SAML): Use this method if your organization wants all users to sign in through its own identity provider (IdP). Read Enterprise SSO with SAML 2.0. The sign-in, sign-up, verification, and invitation pages have the same clear layout. They show the icons of the sign-in providers and of the answer engines. Thus, you can easily find the correct option.","keywords":""},{"kind":"section","title":"Sign-in and SSO options","heading":"How to sign up for a Genezio account","section":"Documentation","crumbs":["Getting started","Sign-in and SSO options","How to sign up for a Genezio account"],"url":"/docs/getting-started/sign-in-and-sso-options#how-to-sign-up-for-a-genezio-account","text":"Enter your name, your email address, and a password. Accept the legal terms. These are the disclaimers on the sign-up page. Select your preference for marketing emails. This is an opt-in, thus the decision is yours. Genezio sends you a verification message. Click the link in the message to confirm your address. This makes your account active. After you verify your email, you can sign in and start the setup of your brand. To continue, read the Genezio setup guide.","keywords":""},{"kind":"section","title":"Sign-in and SSO options","heading":"How to accept a team invitation","section":"Documentation","crumbs":["Getting started","Sign-in and SSO options","How to accept a team invitation"],"url":"/docs/getting-started/sign-in-and-sso-options#how-to-accept-a-team-invitation","text":"If a colleague invited you to an existing team, you get an invitation by email. Open the invitation email and click the accept-invite link. If you do not have an account, complete the short sign-up step. Genezio adds you to the team and opens the shared workspace. For the roles that a team member can have, read Users.","keywords":""},{"kind":"section","title":"Sign-in and SSO options","heading":"Enterprise SSO with SAML 2.0","section":"Documentation","crumbs":["Getting started","Sign-in and SSO options","Enterprise SSO with SAML 2.0"],"url":"/docs/getting-started/sign-in-and-sso-options#enterprise-sso-with-saml-20","text":"Enterprise SSO lets your users sign in to Genezio through the existing identity provider of your company. Genezio supports single sign-on through SAML 2.0 for organizations that want central control of access. Examples of identity providers are Okta, Microsoft Entra ID, OneLogin, Ping, or any SAML 2.0 IdP. The users keep the credentials and the policies that you already manage. Enterprise SSO is not self-serve. You configure it together with the Genezio team. This is the procedure in summary: 1. Genezio sends you the Identifier (Entity ID) and the Reply URL. Use them to configure a SAML application in your IdP. 2. You send Genezio the SAML metadata XML of your IdP. Genezio uses it to complete the connection. After this, your users sign in with their work email. Genezio sends them to the usual login page of your company.","keywords":""},{"kind":"section","title":"Sign-in and SSO options","heading":"SCIM provisioning","section":"Documentation","crumbs":["Getting started","Sign-in and SSO options","SCIM provisioning"],"url":"/docs/getting-started/sign-in-and-sso-options#scim-provisioning","text":"Organizations that use SSO can also enable SCIM provisioning. With SCIM, your identity provider automatically creates and deactivates Genezio accounts. You do not have to invite each user manually. Know these two facts: - SSO alone does not create accounts. A user must exist in Genezio before the user can sign in. SCIM or an invitation creates the user. A successful authentication alone is not sufficient. - You set the roles in Genezio, not in your IdP. SCIM controls who has an account. An account Owner sets what each person can do, under Users. If your organization must have single sign-on, speak to the Genezio team. If you are the IT or identity administrator who does the setup, read Enterprise SSO and SCIM. It has the full procedure: SAML configuration, SCIM tokens, roles, and troubleshooting.","keywords":""},{"kind":"page","title":"Create your first brand","heading":"","section":"Documentation","crumbs":["Documentation","Getting started","Create your first brand"],"url":"/docs/getting-started/create-your-first-brand","text":"Decide what counts as a brand in Genezio, prepare the name, website, language, and market, and write a brand description that makes good topics and relevance. A brand is the unit that holds all other items. Topics, competitors, conversations, content, and scores all belong to one brand.","keywords":""},{"kind":"section","title":"Create your first brand","heading":"What counts as a brand","section":"Documentation","crumbs":["Getting started","Create your first brand","What counts as a brand"],"url":"/docs/getting-started/create-your-first-brand#what-counts-as-a-brand","text":"Usually, the brand is your company. But the correct unit is the item that buyers ask about, and that is not always the company: - A company with one product: one brand. - A company with different products that buyers evaluate separately: one brand for each product. - A company that operates in some markets or languages: one brand for each market. Teams make the most errors with the last case. One brand for two languages gives one unclear average, and not two pictures that you can use. The cause is that the questions, the competitors, and the sources are all different in each market.","keywords":""},{"kind":"section","title":"Create your first brand","heading":"What you need","section":"Documentation","crumbs":["Getting started","Create your first brand","What you need"],"url":"/docs/getting-started/create-your-first-brand#what-you-need","text":"The name. Write it as buyers write it. If people use a short form of the name, record it, because it changes what counts as a mention. The website. The primary domain. Genezio uses it to know what you do and to identify citations of your own pages. The language and the market. These change the questions that Genezio asks and the competitors that Genezio finds.","keywords":""},{"kind":"section","title":"Create your first brand","heading":"Write a good brand description","section":"Documentation","crumbs":["Getting started","Create your first brand","Write a good brand description"],"url":"/docs/getting-started/create-your-first-brand#write-a-good-brand-description","text":"Genezio uses the description to make topics and to find relevance. Thus, write it carefully. Describe what you do and for whom, in the words of a customer. Do not use internal positioning words. \"Project management for creative agencies\" is much better than \"the operating system for modern teams\". The first matches how a person asks the question.","keywords":""},{"kind":"section","title":"Create your first brand","heading":"After you create the brand","section":"Documentation","crumbs":["Getting started","Create your first brand","After you create the brand"],"url":"/docs/getting-started/create-your-first-brand#after-you-create-the-brand","text":"Genezio proposes topics from the description and the website. Examine each topic, and do not accept all of them without a check. Refer to Define topics.","keywords":""},{"kind":"page","title":"Describe your brand","heading":"","section":"Documentation","crumbs":["Documentation","Getting started","Describe your brand"],"url":"/docs/getting-started/describe-your-brand","text":"Write a precise brand description in Genezio: what a good description contains, how it improves your scenarios, and how to set the language of your brand. A precise brand description helps Genezio show your brand correctly in conversations with answer engines. This guide tells you what a good brand description contains and how to set the language of your brand. The brand description makes the generated scenarios better. It also makes sure that answer engines examine your brand in the correct context. When Genezio first makes a description for you, the description gives a summary of these items: - What your brand does - The category of your brand - The type of customers of your brand You examine this description, edit it where necessary, and confirm it.","keywords":""},{"kind":"section","title":"Describe your brand","heading":"What a good brand description contains","section":"Documentation","crumbs":["Getting started","Describe your brand","What a good brand description contains"],"url":"/docs/getting-started/describe-your-brand#what-a-good-brand-description-contains","text":"- Core positioning: What your brand does, in clear words that do not advertise. - Category and use cases: The market where you compete and the problems that you solve. - Differentiators: The qualities that make your brand different in comparisons. - Honest constraints: The price tier, the region, and the integrations. For more about the data that Genezio keeps for each brand, read Brands.","keywords":""},{"kind":"section","title":"Describe your brand","heading":"Set the language of your brand","section":"Documentation","crumbs":["Getting started","Describe your brand","Set the language of your brand"],"url":"/docs/getting-started/describe-your-brand#set-the-language-of-your-brand","text":"You can set a brand language for your brand. Genezio uses this language to make reports and conversations. Thus, the description and the measurement of your brand agree with the market that you serve.","keywords":""},{"kind":"section","title":"Describe your brand","heading":"Next steps","section":"Documentation","crumbs":["Getting started","Describe your brand","Next steps"],"url":"/docs/getting-started/describe-your-brand#next-steps","text":"After you confirm your brand description, continue with Write effective scenarios.","keywords":""},{"kind":"section","title":"Describe your brand","heading":"In the API","section":"Documentation","crumbs":["Getting started","Describe your brand","In the API"],"url":"/docs/getting-started/describe-your-brand#in-the-api","text":"The public API reads and changes the data of a brand in Get brand and Update brand.","keywords":""},{"kind":"page","title":"Define topics","heading":"","section":"Documentation","crumbs":["Documentation","Getting started","Define topics"],"url":"/docs/getting-started/define-topics","text":"Learn what makes a good topic in Genezio, where to find topics that matter, how many to start with, and why topics are different from keywords in AI search. A topic is a subject that buyers ask answer engines about. Topics control all that Genezio measures. A weak set of topics gives numbers that look good but have no meaning.","keywords":""},{"kind":"section","title":"Define topics","heading":"What makes a good topic","section":"Documentation","crumbs":["Getting started","Define topics","What makes a good topic"],"url":"/docs/getting-started/define-topics#what-makes-a-good-topic","text":"A person really asks it. It is not a product feature and it is not an internal category name. It is sufficiently narrow to win or lose. You cannot measure \"Software\". \"CRM for small sales teams\" has an answer, and you are in it or you are not in it. It is commercially important. Visibility on a topic that brings no sales is a vanity number.","keywords":""},{"kind":"section","title":"Define topics","heading":"Where to start","section":"Documentation","crumbs":["Getting started","Define topics","Where to start"],"url":"/docs/getting-started/define-topics#where-to-start","text":"Genezio proposes topics from your brand description and your website. Use these topics as a draft. Keep the topics that you know from sales calls. Get good topics from these three sources: - The questions that prospects really ask you. Your sales team has these questions. - The comparisons that you lose. If buyers compare you with a competitor, that comparison is a topic. - Your Search Console queries, if you connected Search Console. These queries show the real demand that you already get. Refer to Google Search Console.","keywords":""},{"kind":"section","title":"Define topics","heading":"How many topics","section":"Documentation","crumbs":["Getting started","Define topics","How many topics"],"url":"/docs/getting-started/define-topics#how-many-topics","text":"Start with five to ten topics. This number shows the shape of your category. It is also small enough for a reliable number for each topic within your conversation budget. If you divide a small budget across forty topics, you get forty unreliable values. Add more topics later. Start with the topics that you will act on.","keywords":""},{"kind":"section","title":"Define topics","heading":"What topics are not","section":"Documentation","crumbs":["Getting started","Define topics","What topics are not"],"url":"/docs/getting-started/define-topics#what-topics-are-not","text":"Topics are not keywords. A keyword is a text that a person types into a search box. A topic is a subject that a person asks an assistant about. The assistant then divides it into some queries of its own. Refer to Query fanouts. Next step: Creating scenarios. Scenarios are the specific ways to ask about each topic.","keywords":""},{"kind":"page","title":"Write effective scenarios","heading":"","section":"Documentation","crumbs":["Documentation","Getting started","Write effective scenarios"],"url":"/docs/getting-started/create-scenarios","text":"Learn how to write effective scenarios in Genezio that copy the real questions of your buyers, and how to change the type of a scenario after the setup. To write effective scenarios in Genezio, start from the real questions of your buyers. A scenario is the situation of a persona who asks an answer engine for help. This guide tells you how to write good scenarios and how to change the type of a scenario later. A good scenario is similar to the questions that your customers really ask answer engines. It includes the constraints and the follow-up questions that have an effect on which brands the answer engine recommends. For each topic, Genezio makes scenarios for different intents. The intents go from open discovery questions to direct requests for a recommendation and questions about a specific brand. You can examine and edit these scenarios before you confirm them.","keywords":""},{"kind":"section","title":"Write effective scenarios","heading":"How to write scenarios that copy real buyers","section":"Documentation","crumbs":["Getting started","Write effective scenarios","How to write scenarios that copy real buyers"],"url":"/docs/getting-started/create-scenarios#how-to-write-scenarios-that-copy-real-buyers","text":"- Copy a real buyer: Write scenarios from real customer questions, call transcripts, and FAQs. - Include constraints: The team size, the budget, the location, and the technology stack all have an effect on the recommendations. - Include branded and non-branded entry points: Measure how answer engines find your brand, and how they examine your brand by name. - Add follow-up turns: A conversation with many turns shows the deeper reasoning of the model. Each scenario is the situation of a persona. To make strong personas first, read Generate personas from documents.","keywords":""},{"kind":"section","title":"Write effective scenarios","heading":"Change the type of a scenario after the setup","section":"Documentation","crumbs":["Getting started","Write effective scenarios","Change the type of a scenario after the setup"],"url":"/docs/getting-started/create-scenarios#change-the-type-of-a-scenario-after-the-setup","text":"Your first selection is not permanent. You can later convert an existing scenario to a different type. For example, you can convert a Prompter into a Recommender. You select the persona for the converted scenario. For the available scenario types and when to use each type, read Scenarios. For the agent that runs each type, read How the 5 agents work.","keywords":""},{"kind":"section","title":"Write effective scenarios","heading":"Next steps","section":"Documentation","crumbs":["Getting started","Write effective scenarios","Next steps"],"url":"/docs/getting-started/create-scenarios#next-steps","text":"After you confirm your scenarios, Genezio runs conversations with answer engines. To monitor them, read Run conversations.","keywords":""},{"kind":"section","title":"Write effective scenarios","heading":"In the API","section":"Documentation","crumbs":["Getting started","Write effective scenarios","In the API"],"url":"/docs/getting-started/create-scenarios#in-the-api","text":"The public API reads and changes scenarios in List scenarios.","keywords":""},{"kind":"page","title":"Run your first conversations","heading":"","section":"Documentation","crumbs":["Documentation","Getting started","Run your first conversations"],"url":"/docs/getting-started/run-your-first-conversations","text":"Prepare and run your first Genezio conversations: check topics and scenarios, select answer engines, calculate the cost of a run, and read the first answers. A conversation is one scenario to one answer engine. Genezio makes each number from a set of conversations.","keywords":""},{"kind":"section","title":"Run your first conversations","heading":"Before you run","section":"Documentation","crumbs":["Getting started","Run your first conversations","Before you run"],"url":"/docs/getting-started/run-your-first-conversations#before-you-run","text":"Examine these three items, because they control if the results have a meaning: - You will act on your topics. Refer to Define topics. - Each topic has scenarios that are similar to real questions. Refer to Creating scenarios. - Your list of competitors is approximately correct. You can correct it later. But very incorrect entries make the share values incorrect from the start.","keywords":""},{"kind":"section","title":"Run your first conversations","heading":"Select answer engines","section":"Documentation","crumbs":["Getting started","Run your first conversations","Select answer engines"],"url":"/docs/getting-started/run-your-first-conversations#select-answer-engines","text":"Select the answer engines that your buyers really use. Genezio supports these answer engines: - ChatGPT - Claude - Gemini - Perplexity - Copilot - Grok - DeepSeek - Google AI Overview More answer engines give more conversations from your budget. Thus, start with two or three answer engines that are important for your market, not all of them. Refer to Selecting answer engines.","keywords":""},{"kind":"section","title":"Run your first conversations","heading":"The cost of a run","section":"Documentation","crumbs":["Getting started","Run your first conversations","The cost of a run"],"url":"/docs/getting-started/run-your-first-conversations#the-cost-of-a-run","text":"The number of conversations is topics × scenarios × answer engines. For example, ten topics with ten scenarios each on four answer engines give four hundred conversations in each run. Calculate this number before you start, because it increases faster than most people expect.","keywords":""},{"kind":"section","title":"Run your first conversations","heading":"During the run","section":"Documentation","crumbs":["Getting started","Run your first conversations","During the run"],"url":"/docs/getting-started/run-your-first-conversations#during-the-run","text":"Each answer is available when it is complete. You do not have to wait for the full run. Read the first answers. A badly written scenario is immediately clear, and it costs much less to correct it before the full set is complete.","keywords":""},{"kind":"section","title":"Run your first conversations","heading":"After the run","section":"Documentation","crumbs":["Getting started","Run your first conversations","After the run"],"url":"/docs/getting-started/run-your-first-conversations#after-the-run","text":"Go to Understanding your first results. Do not make conclusions from the main score before you read some real answers.","keywords":""},{"kind":"page","title":"Understanding your first results","heading":"","section":"Documentation","crumbs":["Documentation","Getting started","Understanding your first results"],"url":"/docs/getting-started/understanding-your-first-results","text":"Read your first Genezio results correctly: read answers before numbers, find where you are absent, see what is cited, and avoid early conclusions from one run. Your first run gives you a score. The score is the least useful part of the run.","keywords":""},{"kind":"section","title":"Understanding your first results","heading":"Read answers before numbers","section":"Documentation","crumbs":["Getting started","Understanding your first results","Read answers before numbers"],"url":"/docs/getting-started/understanding-your-first-results#read-answers-before-numbers","text":"Open five or six conversations and read them. In ten minutes, you will know more than the dashboard can tell you: - How answer engines describe your category - What answer engines cite - Which competitors answer engines take seriously - How answer engines describe your brand when it appears Teams that do not do this step use the next month to interpret a number that they do not understand.","keywords":""},{"kind":"section","title":"Understanding your first results","heading":"Then look at three items","section":"Documentation","crumbs":["Getting started","Understanding your first results","Then look at three items"],"url":"/docs/getting-started/understanding-your-first-results#then-look-at-three-items","text":"Where you are absent. Topics with a score of zero are the clearest result. Answer engines answer that question without you. The conversations show which brands they use in the answer. Where you appear but answer engines do not recommend you. Answer engines name you, but they select a different brand. This is a positioning problem, and the answers usually tell the reason in clear words. What answer engines cite. These are the sources that have the most effect in your category. This list is the most actionable list in the product. Refer to Most cited sources.","keywords":""},{"kind":"section","title":"Understanding your first results","heading":"Conclusions to avoid now","section":"Documentation","crumbs":["Getting started","Understanding your first results","Conclusions to avoid now"],"url":"/docs/getting-started/understanding-your-first-results#conclusions-to-avoid-now","text":"Do not benchmark the main score. A score of 40% has no meaning until you see the scores of your competitors on the same questions. Do not use one run as a trend. The answers of answer engines change from one run to the next. You need some runs before a direction is real. Refer to Why do LLM results change?. Do not try to correct all problems. The first run usually shows more gaps than a team can work on. Select the topics with real commercial importance, and ignore the other topics for now.","keywords":""},{"kind":"section","title":"Understanding your first results","heading":"A good first action","section":"Documentation","crumbs":["Getting started","Understanding your first results","A good first action"],"url":"/docs/getting-started/understanding-your-first-results#a-good-first-action","text":"Select one topic where a competitor wins. Read what the answer engine cited. Then decide if you can be a better source for that question. This is one week of work with a measurable result, and it is better than a strategy document.","keywords":""},{"kind":"page","title":"Generate personas from documents","heading":"","section":"Documentation","crumbs":["Documentation","Getting started","Generate personas from documents"],"url":"/docs/getting-started/generating-personas-from-documents","text":"Generate personas in Genezio from a document that you upload, such as buyer research or ICP notes. Learn what you get and how to make the result precise. To generate personas in Genezio from documents, upload the persona research that you already have. The platform reads the document and makes the persona for you. This guide tells you why to start from a document, what you get, and how to get a good result. We recommend this method for most new teams. Some documents already show how your business thinks about a customer segment. A persona from such a document is much faster to make than a persona from an empty form. Usually, the persona is also better.","keywords":""},{"kind":"section","title":"Generate personas from documents","heading":"Why personas from documents give better results","section":"Documentation","crumbs":["Getting started","Generate personas from documents","Why personas from documents give better results"],"url":"/docs/getting-started/generating-personas-from-documents#why-personas-from-documents-give-better-results","text":"In Genezio, personas have the largest effect on the quality of the results. A weak persona gives weak conversations, and weak conversations give weak insights. A strong persona comes from real knowledge of the audience. It lets the platform do its best work. Most teams already have this knowledge of the audience in written form. But the knowledge is outside Genezio, for example in these documents: - A buyer-persona document from the brand team or the research team - An internal presentation about customer research - A job description for the role that your product serves - Customer-success notes, summaries of sales discovery calls, one-page ICP documents, and descriptions of the ideal customer With the document import, you bring this work directly into Genezio. You do not have to make it again in a form.","keywords":""},{"kind":"section","title":"Generate personas from documents","heading":"What a generated persona contains","section":"Documentation","crumbs":["Getting started","Generate personas from documents","What a generated persona contains"],"url":"/docs/getting-started/generating-personas-from-documents#what-a-generated-persona-contains","text":"Upload a document, and Genezio makes a full persona for the brand. The persona has these items: - A name - A role - A country - A city - A language - The supporting context that controls how Genezio frames the conversations Genezio saves the persona immediately. When the generation is complete, the persona is a real persona of the brand that you can use. If it is necessary, you can edit the persona after the generation, the same as each other persona. Each upload makes one persona. If your source document has many different personas, upload the document one time for each persona. Alternatively, divide the document before the upload.","keywords":""},{"kind":"section","title":"Generate personas from documents","heading":"When to upload a document and when to use the form","section":"Documentation","crumbs":["Getting started","Generate personas from documents","When to upload a document and when to use the form"],"url":"/docs/getting-started/generating-personas-from-documents#when-to-upload-a-document-and-when-to-use-the-form","text":"You already have a persona document, a research presentation, or a job description: Upload the document. You have ICP notes in a Notion page, a sales playbook, or a different document: Upload the document. You start with no persona research: Complete the form, or write a short document first and then upload it. You want a quick variation of an existing persona: Clone the existing persona and edit it If you are not sure that your document is \"good enough\", upload it. The generation is fast and you can edit the result. Usually, a real document gives a better start than no document.","keywords":""},{"kind":"section","title":"Generate personas from documents","heading":"Tips for a precise persona","section":"Documentation","crumbs":["Getting started","Generate personas from documents","Tips for a precise persona"],"url":"/docs/getting-started/generating-personas-from-documents#tips-for-a-precise-persona","text":"- Use the most specific and current document that you have. Old documents give old personas. - Include the role and the context, not only the demographics. A persona that knows the job-to-be-done of the buyer gives better conversations. A persona that knows only the age and the location gives worse conversations. - If the source has many audiences, divide it first. When you upload only the applicable section, you get a more precise persona. - Examine the persona after the generation. Small edits usually give better conversations. Examples are a clearer name, a more specific city, or a better description. After you have your personas, write effective scenarios for them.","keywords":""},{"kind":"section","title":"Generate personas from documents","heading":"Related pages","section":"Documentation","crumbs":["Getting started","Generate personas from documents","Related pages"],"url":"/docs/getting-started/generating-personas-from-documents#related-pages","text":"- Personas - Setup guide","keywords":""},{"kind":"section","title":"Generate personas from documents","heading":"In the API","section":"Documentation","crumbs":["Getting started","Generate personas from documents","In the API"],"url":"/docs/getting-started/generating-personas-from-documents#in-the-api","text":"The public API reads and changes personas in List personas and Create persona.","keywords":""},{"kind":"page","title":"Customize your dashboard","heading":"","section":"Documentation","crumbs":["Documentation","Getting started","Customize your dashboard"],"url":"/docs/getting-started/customizing-your-dashboard","text":"Customize your Genezio dashboard: hide the charts, tables, and panels that a brand does not use, set a default for the team, and let each user change the view. To customize your Genezio dashboard, hide the charts, tables, and panels that a brand does not use. This guide tells you how Dashboard Components operate, where to configure them, and how each user can change the default view. The Genezio dashboard intentionally shows much data. It shows visibility, recommendations, citations, perceptions, competitors, Share of Voice (SOV), and more. But not all brands need all of these items all the time. With Dashboard Components, the dashboard of each brand shows only the items that are important for that brand.","keywords":""},{"kind":"section","title":"Customize your dashboard","heading":"Why to customize the dashboard of a brand","section":"Documentation","crumbs":["Getting started","Customize your dashboard","Why to customize the dashboard of a brand"],"url":"/docs/getting-started/customizing-your-dashboard#why-to-customize-the-dashboard-of-a-brand","text":"Different brands use different parts of the platform. Examples: - A consumer brand can use mostly Citations and the Recommender results. - A B2B brand can use mostly Introspector and Share of Voice. - A regulated brand can use mostly Perceptions and the Grounded badge. - An agency that manages many clients wants different defaults for each client. When each brand shows all sections, the dashboard has too much data. When you hide the sections that do not apply, the dashboard is easier to read and the daily work is faster.","keywords":""},{"kind":"section","title":"Customize your dashboard","heading":"How hidden dashboard components operate","section":"Documentation","crumbs":["Getting started","Customize your dashboard","How hidden dashboard components operate"],"url":"/docs/getting-started/customizing-your-dashboard#how-hidden-dashboard-components-operate","text":"You can show or hide each dashboard component (a chart, a table, or a panel) independently. Genezio fully removes a hidden component from the page: - The component does not render. - The component does not send its API calls. - Thus, the page loads faster. This is a real increase in performance, not only a visual change. With fewer components, the dashboard is lighter.","keywords":""},{"kind":"section","title":"Customize your dashboard","heading":"Configure the dashboard components in Brand Settings","section":"Documentation","crumbs":["Getting started","Customize your dashboard","Configure the dashboard components in Brand Settings"],"url":"/docs/getting-started/customizing-your-dashboard#configure-the-dashboard-components-in-brand-settings","text":"Configure the visibility of the dashboard components in Brand Settings → Dashboard Components. For each component, select if the brand shows or hides it. The change applies to all users who work on that brand. It is the default shape of the dashboard for the brand.","keywords":""},{"kind":"section","title":"Customize your dashboard","heading":"Overrides for each user","section":"Documentation","crumbs":["Getting started","Customize your dashboard","Overrides for each user"],"url":"/docs/getting-started/customizing-your-dashboard#overrides-for-each-user","text":"The configuration of the brand is a default, not a fixed rule. Each user can override what they see. A user can show a component that the default hides, or hide a component that the default shows. This gives you the advantages of the two levels: - Brand owners can set a good default. Thus, each member of the team sees the same focused dashboard. - Each user can change their own view when they must see a specific item. The default for all other users does not change. For the roles of the users of a brand, read Users.","keywords":""},{"kind":"section","title":"Customize your dashboard","heading":"A first procedure to clean the dashboard","section":"Documentation","crumbs":["Getting started","Customize your dashboard","A first procedure to clean the dashboard"],"url":"/docs/getting-started/customizing-your-dashboard#a-first-procedure-to-clean-the-dashboard","text":"Use this procedure to start: 1. Open the dashboard. For one week, record the items that you really use. 2. Go to Brand Settings → Dashboard Components. 3. Hide the panels that you never opened. 4. Keep the panels that you are not sure about. Examine them again after some more weeks. 5. Change the configuration again when the priorities of the brand change. You can always show the components again. The customization is reversible. Thus, you can safely hide all the items that you do not use today.","keywords":""},{"kind":"section","title":"Customize your dashboard","heading":"Related pages","section":"Documentation","crumbs":["Getting started","Customize your dashboard","Related pages"],"url":"/docs/getting-started/customizing-your-dashboard#related-pages","text":"- Setup guide - Your KPIs explained","keywords":""},{"kind":"page","title":"Topic tags","heading":"","section":"Documentation","crumbs":["Documentation","Getting started","Topic tags"],"url":"/docs/getting-started/topic-tags","text":"Use topic tags to group your topics by content pillar, product line, or customer segment, filter the dashboard by tag, and keep the tag library clean. Topic tags are labels that you define and attach to topics in Genezio. They keep your topics organized by content pillar, product line, or customer segment. This guide tells you how to make tags, apply them to topics, filter by them, and keep the tag library clean. After some weeks of a serious deployment, most brands have many topics: 40, 60, and sometimes more than 100. The topics are about content pillars, product lines, customer segments, regions, and tracking purposes. Without organization, the topics page becomes a flat list that nobody reads. Tags are the layer that solves this problem.","keywords":""},{"kind":"section","title":"Topic tags","heading":"What topic tags are","section":"Documentation","crumbs":["Getting started","Topic tags","What topic tags are"],"url":"/docs/getting-started/topic-tags#what-topic-tags-are","text":"A topic tag is a label with a name and a color that you select. A topic can have many tags. Thus, the same topic can be in many views at the same time. For example, a topic about prices for SMB customers can have the tag Pricing and the tag SMB. You manage tags for each brand. Each brand has its own tag library, thus the taxonomies of different brands stay separate. This is useful for agencies and teams with many brands, where each brand has its own logic of organization. Genezio does not set a taxonomy. You apply the structure that agrees with how your team thinks about the topics that you track. For more about topics, read Topics.","keywords":""},{"kind":"section","title":"Topic tags","heading":"Usual tag taxonomies","section":"Documentation","crumbs":["Getting started","Topic tags","Usual tag taxonomies"],"url":"/docs/getting-started/topic-tags#usual-tag-taxonomies","text":"These are some usual taxonomies of brands.","keywords":""},{"kind":"section","title":"Topic tags","heading":"Content pillars","section":"Documentation","crumbs":["Getting started","Topic tags","Content pillars"],"url":"/docs/getting-started/topic-tags#content-pillars","text":"Tag the topics by their pillar: Solutions, Use Cases, Platform, Industries. This is useful when the content plan and the brand tracking use the same words. It also shows you the performance of each pillar in answer engines.","keywords":""},{"kind":"section","title":"Topic tags","heading":"Product lines or SKUs","section":"Documentation","crumbs":["Getting started","Topic tags","Product lines or SKUs"],"url":"/docs/getting-started/topic-tags#product-lines-or-skus","text":"If your company has many products, tag each topic by its product: CRM, Marketing Hub, Service Hub. Then you can filter the dashboard by product. This is important when each product line has its own owner.","keywords":""},{"kind":"section","title":"Topic tags","heading":"Customer segments","section":"Documentation","crumbs":["Getting started","Topic tags","Customer segments"],"url":"/docs/getting-started/topic-tags#customer-segments","text":"Tag the topics by audience: SMB, Mid-market, Enterprise. This is useful when your topics include the full funnel and different teams own different segments. You can also combine taxonomies. One topic can have Pricing (pillar), CRM (product), and SMB (segment). When you filter, the topic shows in each of these views.","keywords":""},{"kind":"section","title":"Topic tags","heading":"How to create and apply tags","section":"Documentation","crumbs":["Getting started","Topic tags","How to create and apply tags"],"url":"/docs/getting-started/topic-tags#how-to-create-and-apply-tags","text":"You must create a tag in the tag library before you can apply it to topics. This keeps the library clean and the colors coordinated, with no unplanned duplicates.","keywords":""},{"kind":"section","title":"Topic tags","heading":"Create a tag","section":"Documentation","crumbs":["Getting started","Topic tags","Create a tag"],"url":"/docs/getting-started/topic-tags#create-a-tag","text":"Open the tag library of the brand. Click to add a new tag. Enter a name and select a color in the color picker. Save the tag. The tag is now available for all of the brand.","keywords":""},{"kind":"section","title":"Topic tags","heading":"Apply tags to a topic","section":"Documentation","crumbs":["Getting started","Topic tags","Apply tags to a topic"],"url":"/docs/getting-started/topic-tags#apply-tags-to-a-topic","text":"Open a topic and select one or more tags from the tag library. A topic can have as many tags as it needs.","keywords":""},{"kind":"section","title":"Topic tags","heading":"Filter the topics by tag","section":"Documentation","crumbs":["Getting started","Topic tags","Filter the topics by tag"],"url":"/docs/getting-started/topic-tags#filter-the-topics-by-tag","text":"You can filter the topics page by tag. Select one or more tags. The list then shows only the topics that agree with the filter. Tag filters combine with the other filters, for example the answer engine, the persona, and the time range. Thus, you can ask a question such as this one: \"Show me all topics with the tag SMB for ChatGPT in the last 30 days.\" For the other filters, read Master Filters.","keywords":""},{"kind":"section","title":"Topic tags","heading":"Manage the tag library","section":"Documentation","crumbs":["Getting started","Topic tags","Manage the tag library"],"url":"/docs/getting-started/topic-tags#manage-the-tag-library","text":"The tag library contains the taxonomy. From time to time, it needs maintenance: - Edit a tag to change its name or its color. - Delete a tag that you do not use. Genezio removes the tag from each topic that had it. - The scope is always the brand. Tags do not move between brands. Manage the library as you manage each taxonomy: remove old tags, keep the names short, and keep the set of colors easy to read.","keywords":""},{"kind":"section","title":"Topic tags","heading":"Best practices for topic tags","section":"Documentation","crumbs":["Getting started","Topic tags","Best practices for topic tags"],"url":"/docs/getting-started/topic-tags#best-practices-for-topic-tags","text":"These practices keep the tags useful and clear: - Start small. At the start, three to five tags are sufficient. Add more tags when you really need them, not before. - Use one axis at a time. First, add a taxonomy of pillars and make it stable. Then, add a taxonomy of segments or products. This is easier than a design of all three taxonomies on the first day. - Use colors with discipline. Use different colors for tags of different taxonomies. For example, give all pillar tags colors of one family and all segment tags colors of a different family. Then you can read the topics page quickly. - Delete the tags that you do not use. If nobody filtered on a tag for months, the tag is probably noise. Delete it. - Keep the names short. Tags are visual labels, not sentences. SMB, Pricing, and EU are easier to read than Small and Medium Business Customers.","keywords":""},{"kind":"section","title":"Topic tags","heading":"Related pages","section":"Documentation","crumbs":["Getting started","Topic tags","Related pages"],"url":"/docs/getting-started/topic-tags#related-pages","text":"- Setup guide","keywords":""},{"kind":"section","title":"Topic tags","heading":"In the API","section":"Documentation","crumbs":["Getting started","Topic tags","In the API"],"url":"/docs/getting-started/topic-tags#in-the-api","text":"The public API reads and changes tags in List tags and Set topic tags.","keywords":""},{"kind":"page","title":"Your first week","heading":"","section":"Documentation","crumbs":["Documentation","Getting started","Your first week"],"url":"/docs/getting-started/your-first-week","text":"A day-by-day plan for your first week with Genezio: set a baseline, read your AI visibility scores, act on insights, write a brief, and set a monthly report. Your first week with Genezio follows a practical day-by-day plan for Digital Marketing Managers. At the end of the first week, you have these items: - Your baseline scores - Your first analysis of competitors - A list of content actions in the order of priority Before you start, prepare these items: - The name of your brand - The names of your top 2–3 competitors - A list of 5–10 topics where you want to rank in the answers of answer engines Topics must be category-level phrases, not product names. Examples are \"AI visibility platform\", \"project management for remote teams\", or \"CRM for startups\".","keywords":""},{"kind":"section","title":"Your first week","heading":"Day 1 (Monday): Configure your brand and your baseline","section":"Documentation","crumbs":["Getting started","Your first week","Day 1 (Monday): Configure your brand and your baseline"],"url":"/docs/getting-started/your-first-week#day-1-monday-configure-your-brand-and-your-baseline","text":"Time: approximately 2 hours. Goal: the first conversations run at the end of the day. 1. Add your brand in Genezio. Include the URL of your website and a short description of what you do. Read Brands. 2. Add your top 2–3 competitors. Genezio tracks them in parallel. Thus, each score that you see already includes a comparison. Read Competitors. 3. Create your first 3–5 topics. Start narrow, with one topic for each core use case or buyer persona. You can add more topics later. Read Topics. 4. Create 2–3 scenarios for each topic. Write them as real questions of users, in the words that a buyer uses with ChatGPT. Do not write \"CRM features\". Write \"What CRM should a 10-person startup use?\". Read Create topics and scenarios. 5. Select your answer engines. Start with ChatGPT and Perplexity. Add Claude and Gemini after you have the baseline. Read Select answer engines. 6. Run your first conversations. Usually, the results are available in some minutes. Result of Day 1: - The first AI Recommendations % and AI Visibility % numbers for your brand and your competitors. - An initial list of the sources that answer engines already cite in your category.","keywords":""},{"kind":"section","title":"Your first week","heading":"Day 2 (Tuesday): Read your baseline scores","section":"Documentation","crumbs":["Getting started","Your first week","Day 2 (Tuesday): Read your baseline scores"],"url":"/docs/getting-started/your-first-week#day-2-tuesday-read-your-baseline-scores","text":"Time: approximately 1.5 hours. Goal: know your position and the reasons for it. 1. Examine the AI Recommendations % and the AI Visibility % for each topic. Record the topics where your score is good and the topics where you have gaps. 2. Open Citations. For each topic with a low score, find the URLs that answer engines cite. Record them. This list becomes your plan for PR and link-building. 3. Examine Perceptions. Find the claims that answer engines make about your brand. Flag each claim that is old, negative, or missing. 4. Compare your scores with the scores of your competitors. Record which competitor leads, by how much, and which sources cite them but not you. Find the topic where the gap between you and the top competitor is the smallest. That topic is your first opportunity for a quick win. Result of Day 2: - A scored list of topics in the order of priority. - A list of citation sources: 5–10 domains that answer engines trust in your category. - The perception issues, flagged.","keywords":""},{"kind":"section","title":"Your first week","heading":"Day 3 (Wednesday): Examine the actionable insights","section":"Documentation","crumbs":["Getting started","Your first week","Day 3 (Wednesday): Examine the actionable insights"],"url":"/docs/getting-started/your-first-week#day-3-wednesday-examine-the-actionable-insights","text":"Time: approximately 1 hour. Goal: Genezio tells you exactly what to repair. 1. Open Actionable Insights. Read all four categories: Growth Opportunities, Critical Visibility Gaps, Citation & Authority Leverage, and Positioning & Structural Optimization. Read Actionable insights. 2. Set the priority by impact. The high-priority insights are the ones where your competitors win now. Dismiss each insight that your team cannot act on. 3. Generate more insights on demand for specific topics or scenarios, if you want a focused view. You can add custom instructions, such as \"focus on quick wins\" or \"prioritize opportunities vs. [competitor name]\". For the manual generation of insights, read Actionable insights. 4. Connect each insight to a content action: create content, get a citation, repair the schema, or get press coverage. Result of Day 3: - A list of 5–10 specific actions in the order of priority. - For each action: the content type, the responsible person, and the estimated timeline.","keywords":""},{"kind":"section","title":"Your first week","heading":"Day 4 (Thursday): Make your first content brief","section":"Documentation","crumbs":["Getting started","Your first week","Day 4 (Thursday): Make your first content brief"],"url":"/docs/getting-started/your-first-week#day-4-thursday-make-your-first-content-brief","text":"Time: approximately 2 hours. Goal: the first content piece that uses Genezio data has a brief and an owner. Take the insight with the highest priority from Day 3. Make a correct content brief from it. The Content Hub of Genezio can also make a brief for you. Read Briefs. Include these items in the brief: - Title and target keyword: Use the exact phrase that the answer engine uses in the scenarios where your score is low. - Citation sources to refer to: List 3–5 sources from the Citations section of Genezio for this topic. Your content must have external links to these domains. - Incorrect claims to correct: Include each perception of answer engines about your brand that is incorrect. Better content must correct it. The content can also reinforce the correct claims. - Schema markup requirements: Flag if the page needs FAQPage, HowTo, or a different schema. - Owner and date of the new measurement: Assign the brief. Run the applicable Genezio conversations again 2–4 weeks after the publication. Result of Day 4: - One full content brief, with an owner. - The date of the new measurement, in your calendar.","keywords":""},{"kind":"section","title":"Your first week","heading":"Day 5 (Friday): Configure your monthly report","section":"Documentation","crumbs":["Getting started","Your first week","Day 5 (Friday): Configure your monthly report"],"url":"/docs/getting-started/your-first-week#day-5-friday-configure-your-monthly-report","text":"Time: approximately 1 hour. Goal: AI visibility becomes a permanent part of your marketing metrics. 1. Make a screenshot or an export of your baseline scores from Day 1. This is your benchmark for Week 0. 2. Add the AI Recommendations % and the AI Visibility % to your monthly marketing report. Track them together with the organic traffic and the volume of branded searches. 3. Schedule a monthly new run of all conversations. Connect the content actions to the changes of the scores over time. 4. Add new topics and scenarios from what you learned this week. At the end of Week 1, you have these items: - The baseline AI Recommendations % and AI Visibility % for your key topics - A list of citation sources: the domains to target for PR and link-building - A perception audit: what answer engines say about you today, and what you want them to say - A list of actions from Actionable Insights, in the order of priority - The first content brief, with an owner - A monthly report rhythm","keywords":""},{"kind":"section","title":"Your first week","heading":"Weekly, monthly, and quarterly rhythm after the first week","section":"Documentation","crumbs":["Getting started","Your first week","Weekly, monthly, and quarterly rhythm after the first week"],"url":"/docs/getting-started/your-first-week#weekly-monthly-and-quarterly-rhythm-after-the-first-week","text":"Weekly: Examine Actionable Insights for new recommendations. Make briefs for 1–2 content pieces from the current gaps, and assign them. 30 min. Monthly: Run a full batch of conversations. Examine the changes of the scores. Update your content calendar from the scores that changed and the scores that did not change. Report AI Recommendations %, AI Visibility %, and Share of Voice to the leadership. 2 hours. Quarterly: Add topics. Add new scenarios for seasonal or strategic priorities. Run a Comparer Agent analysis on your top 2 competitors to find changes in their narrative. Half day","keywords":""},{"kind":"page","title":"Introduction","heading":"","section":"Documentation","crumbs":["Documentation","Introduction","Introduction"],"url":"/docs/introduction","text":"Start with the basics: why answers from AI assistants replace search results, how Genezio measures your brand in them, and which two numbers matter the most. More and more people ask an assistant and do not search. If your brand is not in the answer, you lost that moment, even if you are first on Google for the same question. The Introduction section tells what this changes and how Genezio measures it. If you are a new user, start here: 1. Read What is Genezio?. 2. For the business case, read Genezio for marketers. 3. For the mechanics, read How LLM search works. All other data depends on two numbers: - AI Visibility: Do you appear? - AI Recommendation: Does the answer engine recommend you? These two numbers change independently. The difference between them is usually the most useful data on your dashboard.","keywords":""},{"kind":"page","title":"What is Genezio?","heading":"","section":"Documentation","crumbs":["Documentation","Introduction","What is Genezio?"],"url":"/docs/introduction/what-is-genezio","text":"Genezio is an AI visibility platform that measures how often ChatGPT, Claude, Gemini, and Perplexity recommend your brand, and shows how to improve it. Genezio is an AI visibility platform that helps organizations measure and improve how frequently answer engines recommend their brand. Examples of answer engines are ChatGPT, Claude, Gemini, and Perplexity. An answer engine recommends a brand when a user asks it for a solution. More users use AI assistants to research products, compare services, and make purchase decisions. Thus, a recommendation in an AI-generated answer becomes as important as a high rank on Google. Genezio helps teams measure their AI recommendation rate, understand the causes of that rate, and do actions that make it better.","keywords":""},{"kind":"section","title":"What is Genezio?","heading":"The change from search engines to AI answers","section":"Documentation","crumbs":["Introduction","What is Genezio?","The change from search engines to AI answers"],"url":"/docs/introduction/what-is-genezio#the-change-from-search-engines-to-ai-answers","text":"In the past, users found brands through search engines such as Google. This was the usual interaction: 1. A user enters a search query. 2. The search engine shows a list of links. 3. The user clicks one or more results. In AI search, the experience is different. The AI system does not show a list of links. It makes a direct answer from information that it gets from many sources. Example: User query: What are the best running shoes for marathon training? The AI assistant does not show links. It can give an answer such as this one: AI answer: The best running shoes for marathon training include the Nike Alphafly, Adidas Adios Pro, and Saucony Endorphin Elite due to their energy return and cushioning. The sources for this answer can include: - Product review sites - Ecommerce stores - Brand websites - Forums and communities - Editorial articles In this environment, the important question is this: does the AI system recommend your brand when users ask for a solution? A position in a list of search results is not sufficient.","keywords":""},{"kind":"section","title":"What is Genezio?","heading":"The two core metrics: AI Recommendations and AI Visibility","section":"Documentation","crumbs":["Introduction","What is Genezio?","The two core metrics: AI Recommendations and AI Visibility"],"url":"/docs/introduction/what-is-genezio#the-two-core-metrics-ai-recommendations-and-ai-visibility","text":"Genezio measures two items: 1. AI Recommendations %: how frequently answer engines recommend your brand when users ask for a solution. This is the primary metric. In AI search, it is the item that is nearest to a purchase decision. 2. AI Visibility %: how frequently your brand occurs in AI-generated answers at all (as a mention, a citation, or a comparison). This is the wider measure of the AI presence of your brand. For example, a user asks this question: User query: What is the best CRM for startups? An AI assistant can give this answer: AI answer: For a startup, I'd recommend HubSpot for its free tier and ease of setup. Salesforce and Pipedrive are also worth considering if you need more customization. In this conversation: - The AI recommends HubSpot. HubSpot is the choice that the AI suggests. - HubSpot, Salesforce, and Pipedrive all have visibility. They occur in the answer. - A CRM that the answer does not mention has no recommendation and no visibility for this conversation. Genezio measures the two metrics across thousands of AI conversations. It also shows you what to do to get more recommendations.","keywords":""},{"kind":"section","title":"What is Genezio?","heading":"How Genezio works in five steps","section":"Documentation","crumbs":["Introduction","What is Genezio?","How Genezio works in five steps"],"url":"/docs/introduction/what-is-genezio#how-genezio-works-in-five-steps","text":"Genezio simulates real user interactions with AI systems and analyzes the answers. The process has these steps.","keywords":""},{"kind":"section","title":"What is Genezio?","heading":"1. Define topics","section":"Documentation","crumbs":["Introduction","What is Genezio?","1. Define topics"],"url":"/docs/introduction/what-is-genezio#1-define-topics","text":"A topic is an area where a brand wants to occur in AI-generated answers. Examples: - \"running shoes for marathon training\" - \"best CRM for startups\" - \"luxury ski resorts in Switzerland\"","keywords":""},{"kind":"section","title":"What is Genezio?","heading":"2. Create scenarios","section":"Documentation","crumbs":["Introduction","What is Genezio?","2. Create scenarios"],"url":"/docs/introduction/what-is-genezio#2-create-scenarios","text":"A scenario is a realistic situation of a user who asks an AI assistant for help. Example scenario: User query: I am training for my first marathon. What running shoes should I buy?","keywords":""},{"kind":"section","title":"What is Genezio?","heading":"3. Run conversations","section":"Documentation","crumbs":["Introduction","What is Genezio?","3. Run conversations"],"url":"/docs/introduction/what-is-genezio#3-run-conversations","text":"Genezio runs conversations with the supported answer engines, for example: - ChatGPT - Claude - Gemini - Perplexity Each conversation simulates a real user who asks questions. Sometimes the user also asks follow-up questions.","keywords":""},{"kind":"section","title":"What is Genezio?","heading":"4. Analyze the answers","section":"Documentation","crumbs":["Introduction","What is Genezio?","4. Analyze the answers"],"url":"/docs/introduction/what-is-genezio#4-analyze-the-answers","text":"The platform analyzes each answer and extracts structured information: - The brands that the answer mentions - The sources that the answer cites - The claims (perceptions) about each brand - The accuracy of those claims, when compared with your brand knowledge","keywords":""},{"kind":"section","title":"What is Genezio?","heading":"5. Make insights","section":"Documentation","crumbs":["Introduction","What is Genezio?","5. Make insights"],"url":"/docs/introduction/what-is-genezio#5-make-insights","text":"With this data, Genezio calculates metrics and makes insights, for example: - The AI Recommendations score and the AI Visibility score - The share of voice, when compared with competitors - The sources that answer engines cite most - The opportunities to make your content better","keywords":""},{"kind":"section","title":"What is Genezio?","heading":"What Genezio measures in AI answers","section":"Documentation","crumbs":["Introduction","What is Genezio?","What Genezio measures in AI answers"],"url":"/docs/introduction/what-is-genezio#what-genezio-measures-in-ai-answers","text":"Genezio analyzes different aspects of AI-generated answers.","keywords":""},{"kind":"section","title":"What is Genezio?","heading":"Brand mentions","section":"Documentation","crumbs":["Introduction","What is Genezio?","Brand mentions"],"url":"/docs/introduction/what-is-genezio#brand-mentions","text":"Brand mentions show how frequently a brand occurs in the answers of AI assistants.","keywords":""},{"kind":"section","title":"What is Genezio?","heading":"Citations","section":"Documentation","crumbs":["Introduction","What is Genezio?","Citations"],"url":"/docs/introduction/what-is-genezio#citations","text":"Citations show which sources the AI system uses as references for its answers. These sources can include: - Brand websites - Review sites - News articles - Blogs For more about citations, read How AI citations work.","keywords":""},{"kind":"section","title":"What is Genezio?","heading":"Competitor presence","section":"Documentation","crumbs":["Introduction","What is Genezio?","Competitor presence"],"url":"/docs/introduction/what-is-genezio#competitor-presence","text":"Competitor presence shows which other brands occur in the same answers. This helps you identify your competitors in the AI landscape.","keywords":""},{"kind":"section","title":"What is Genezio?","heading":"Perception accuracy","section":"Documentation","crumbs":["Introduction","What is Genezio?","Perception accuracy"],"url":"/docs/introduction/what-is-genezio#perception-accuracy","text":"Perception accuracy shows if the claims of the AI about a brand are correct or incorrect. Genezio compares the claims with your brand knowledge base. This helps you find where answer engines give incorrect information about your brand, and where they give correct information.","keywords":""},{"kind":"section","title":"What is Genezio?","heading":"Share of voice","section":"Documentation","crumbs":["Introduction","What is Genezio?","Share of voice"],"url":"/docs/introduction/what-is-genezio#share-of-voice","text":"Share of voice is the relative visibility of a brand, when compared with its competitors, across many AI conversations.","keywords":""},{"kind":"section","title":"What is Genezio?","heading":"Why AI recommendations are important for your pipeline","section":"Documentation","crumbs":["Introduction","What is Genezio?","Why AI recommendations are important for your pipeline"],"url":"/docs/introduction/what-is-genezio#why-ai-recommendations-are-important-for-your-pipeline","text":"AI assistants become a usual start point for product discovery and purchase decisions. More users ask questions such as these: - \"What CRM should a startup use?\" - \"Recommend running shoes for marathon training.\" - \"What is the best credit card for travel?\" The brands that answer engines recommend in these answers get a direct pipeline advantage. If an answer engine recommends your competitor and not you, you lose deals before the buyer visits your website. Genezio helps you understand this new landscape and do actions to get more recommendations.","keywords":""},{"kind":"section","title":"What is Genezio?","heading":"What Genezio does not do","section":"Documentation","crumbs":["Introduction","What is Genezio?","What Genezio does not do"],"url":"/docs/introduction/what-is-genezio#what-genezio-does-not-do","text":"Genezio does not control how AI systems make answers. It helps organizations do these tasks: - Measure how frequently AI answers recommend their brand - Understand why AI answers recommend some brands more than other brands - Identify opportunities to increase their recommendation rate Usually, these items increase AI recommendations: better content, stronger authority signals, and more relevance across the web.","keywords":""},{"kind":"section","title":"What is Genezio?","heading":"Next steps","section":"Documentation","crumbs":["Introduction","What is Genezio?","Next steps"],"url":"/docs/introduction/what-is-genezio#next-steps","text":"To start with Genezio, read these pages: - Genezio setup guide: the steps from the sign-up to your first results. - Scenarios: how to prepare your first AI recommendations analysis. - Why AI recommendations matter: why the context of a recommendation is important for your pipeline.","keywords":""},{"kind":"page","title":"Genezio for marketers","heading":"","section":"Documentation","crumbs":["Documentation","Introduction","Genezio for marketers"],"url":"/docs/introduction/genezio-for-marketers","text":"See why AI visibility is now a marketing KPI, how AI search is different from SEO, and how marketers use Genezio to measure and improve brand presence. For CMOs, content strategists, SEO leads, and digital marketing managers. The summary: Genezio tells you how frequently your brand appears when real people ask AI assistants such as ChatGPT, Claude, or Perplexity about your category. It also tells you what you can do to appear more frequently.","keywords":""},{"kind":"section","title":"Genezio for marketers","heading":"Why AI visibility is now a marketing KPI","section":"Documentation","crumbs":["Introduction","Genezio for marketers","Why AI visibility is now a marketing KPI"],"url":"/docs/introduction/genezio-for-marketers#why-ai-visibility-is-now-a-marketing-kpi","text":"For the last ten years, marketing teams monitored their ranks on Google. A page-one rank gave you traffic. Without it, you were not visible. This model changes. More and more, people do not use the list of links. They ask an AI assistant directly: \"What's the best CRM for a startup of 20 people?\" ChatGPT gives a direct answer with 3 recommendations. The user does not have to click. If your brand is not in that answer, you lost that moment, even if you are 1 on Google for the same query. The answer engines: Genezio monitors ChatGPT, Claude, Gemini, Perplexity, Copilot, Grok, DeepSeek, and Google AI Overview. You select the answer engines that are important for your market. The two numbers: AI Visibility: how frequently you appear. AI Recommendation: how frequently the answer engine recommends you. A brand can be very visible and rarely recommended.","keywords":""},{"kind":"section","title":"Genezio for marketers","heading":"Traditional search and AI search","section":"Documentation","crumbs":["Introduction","Genezio for marketers","Traditional search and AI search"],"url":"/docs/introduction/genezio-for-marketers#traditional-search-and-ai-search","text":"How it works: The user searches, gets links, and clicks your link The user asks an AI assistant and gets a direct answer. Your brand is in the answer or it is not. What gives visibility: The page rank and the click-through rate If your brand is in the AI-generated answer. The risk: Page 2 gets no traffic There is no page 2. You are in the answer or you are not. Key insight: A brand can be 1 on Google for \"best project management tool\" and not be in the answer of ChatGPT to the same question. These are two different visibility contests, and most brands compete in only one of them.","keywords":""},{"kind":"section","title":"Genezio for marketers","heading":"What Genezio does","section":"Documentation","crumbs":["Introduction","Genezio for marketers","What Genezio does"],"url":"/docs/introduction/genezio-for-marketers#what-genezio-does","text":"Genezio simulates how real users use AI assistants. Then it analyzes what these AI assistants say about your brand. It is similar to mystery shopping for answer engines, with thousands of automatic conversations. 1. You define topics. Topics are the categories that you want people to know you for. 2. Genezio runs conversations. Simulated users ask AI assistants real questions. 3. Genezio analyzes the AI answers. It finds which brands the answers mentioned, cited, and recommended. 4. You get insights. These show the gaps, the opportunities, and what to correct.","keywords":""},{"kind":"section","title":"Genezio for marketers","heading":"What you can do with Genezio as a marketer","section":"Documentation","crumbs":["Introduction","Genezio for marketers","What you can do with Genezio as a marketer"],"url":"/docs/introduction/genezio-for-marketers#what-you-can-do-with-genezio-as-a-marketer","text":"Measure the AI presence of your brand. See your AI Visibility % and your AI Recommendations % across topics, competitors, and answer engines. Know where you win and where you are not visible. Understand why answer engines prefer competitors. Genezio shows which sources answer engines cite when they recommend competitors and not you. These sources are your content gap and your content roadmap. Know which content to make. The Actionable Insights section gives specific recommendations: the topics to target, the sources that must cite you, and the narratives to build. You do not have to guess what to write. Monitor improvements over time. When you publish new content, improve your site, or get new mentions, your scores change. You can see the direct effect of your marketing work on AI visibility. Compare your brand with competitors. See your Share of Voice. It shows how frequently your brand appears, compared to competitors, across the same set of AI conversations. It is one number that your CMO can monitor each month. For CMOs: Put AI Visibility % and Share of Voice on your monthly marketing dashboard, together with organic traffic and branded search volume. They are the next step in the measurement of brand presence.","keywords":""},{"kind":"section","title":"Genezio for marketers","heading":"Related pages","section":"Documentation","crumbs":["Introduction","Genezio for marketers","Related pages"],"url":"/docs/introduction/genezio-for-marketers#related-pages","text":"These pages change the technical concepts of Genezio into practical work: - Your KPIs explained: what AI Visibility % and AI Recommendations % mean for your business. - How the 5 agents work: the types of AI conversations that Genezio runs, and why they are important. - From data to content strategy: how to change Genezio insights into a content plan. - Your first week with Genezio: a practical guide for each day of your first week.","keywords":""},{"kind":"page","title":"Why AI recommendations matter","heading":"","section":"Documentation","crumbs":["Documentation","Introduction","Why AI recommendations matter"],"url":"/docs/introduction/why-ai-recommendations-matter","text":"AI recommendations matter because buyers now decide in conversations with answer engines. Learn why the context of a recommendation counts for your pipeline. AI recommendations matter because more buyer journeys now start in answer engines such as ChatGPT, Claude, Gemini, and Perplexity. This page explains why a recommendation in the correct context is important for your pipeline, and how Genezio measures it. This page is for CMOs, content strategists, SEO leads, and digital marketing managers. The summary in one sentence: If an AI assistant does not recommend your brand when the correct customer asks the correct question, you lose that deal before the buyer visits your website.","keywords":""},{"kind":"section","title":"Why AI recommendations matter","heading":"How buyers use AI assistants to make a purchase decision","section":"Documentation","crumbs":["Introduction","Why AI recommendations matter","How buyers use AI assistants to make a purchase decision"],"url":"/docs/introduction/why-ai-recommendations-matter#how-buyers-use-ai-assistants-to-make-a-purchase-decision","text":"This section shows what occurs when a person uses ChatGPT, Claude, Gemini, or Perplexity to make a purchase decision. The interaction is not one question. It is a conversation. A marketing director who looks for a CRM does not only type \"best CRM\" and accept the first answer. The interaction is more similar to this: Turn 1: \"I need a CRM for a B2B startup with a 10-person sales team.\" Turn 2: \"We need something under $50/month per seat with good HubSpot integration.\" Turn 3: \"Between Pipedrive and Close, which one is better for outbound-heavy teams?\" With each turn, the AI makes its list of recommendations smaller. Brands enter the conversation and leave it. In the last answer, the AI recommends one or two brands and rejects the other brands. In AI search, this is where you win or lose deals.","keywords":""},{"kind":"section","title":"Why AI recommendations matter","heading":"Why the context of a recommendation counts","section":"Documentation","crumbs":["Introduction","Why AI recommendations matter","Why the context of a recommendation counts"],"url":"/docs/introduction/why-ai-recommendations-matter#why-the-context-of-a-recommendation-counts","text":"Marketers must understand this: a recommendation is important only when it occurs in the correct context. A recommendation as \"a good CRM\" in the answer to a generic prompt is good. A recommendation as \"the best CRM for outbound-heavy B2B teams under $50/seat\" is better, when that is exactly your target customer. This type of recommendation increases your pipeline. For this reason, Genezio does not test your brand with random questions. Genezio builds each conversation from these items: - A persona: the person who asks (the role, the industry, the company size, the location, the language) - A scenario: the need of the persona (the specific constraints, the budget, the use case) - A multi-step conversation: the change of the recommendation of the AI when the persona adds details Thus, your AI Recommendations % shows the performance of your brand with the customers that you want to get. It shows this performance in the situations that are important for your business.","keywords":""},{"kind":"section","title":"Why AI recommendations matter","heading":"Absent or recommended: the two outcomes that count","section":"Documentation","crumbs":["Introduction","Why AI recommendations matter","Absent or recommended: the two outcomes that count"],"url":"/docs/introduction/why-ai-recommendations-matter#absent-or-recommended-the-two-outcomes-that-count","text":"If the AI does not recommend your brand, you lose the opportunity. This is the clearest signal in AI search. A persona that matches your ideal customer asks an AI assistant for a recommendation in your category. If the AI does not mention you, that buyer now examines your competitors. There is no second page. There is no \"scroll down.\" You were not in the conversation. If the AI recommends your brand in the correct scenario, you get influence. The AI can recommend your brand to the correct persona, with the correct constraints, in a realistic multi-step conversation. This is important. It shows that the AI thinks that your brand is a credible match for the specific needs of that buyer. More buyer journeys start in AI assistants. Thus, that recommendation has real weight.","keywords":""},{"kind":"section","title":"Why AI recommendations matter","heading":"Example: a generic mention and a real recommendation","section":"Documentation","crumbs":["Introduction","Why AI recommendations matter","Example: a generic mention and a real recommendation"],"url":"/docs/introduction/why-ai-recommendations-matter#example-a-generic-mention-and-a-real-recommendation","text":"Compare these two conversations. Conversation A: \"What are some CRM tools?\" The AI gives a list of 8 brands in alphabetical order. Your brand is one of them. Conversation B: \"I run a 15-person SaaS sales team. We do mostly outbound. I need a CRM under $50/seat that integrates with our existing email stack. What should I use?\" The AI recommends your brand as the first choice. It gives a specific explanation of why your brand is a good match. The two conversations both count as a \"mention.\" But only Conversation B is a recommendation. Only Conversation B shows a real purchase moment. Genezio focuses on conversations of type B. Each conversation has a persona and a scenario that copy how your real customers ask questions. For this reason, the AI Recommendations % is useful. It measures if your brand wins in the important moments, not if it occurs in generic lists.","keywords":""},{"kind":"section","title":"Why AI recommendations matter","heading":"How Genezio measures AI recommendations","section":"Documentation","crumbs":["Introduction","Why AI recommendations matter","How Genezio measures AI recommendations"],"url":"/docs/introduction/why-ai-recommendations-matter#how-genezio-measures-ai-recommendations","text":"Genezio runs thousands of AI conversations. Each conversation starts from a persona and a scenario that you define. The measurement has these steps: 1. You define your audience: who your customers are, where they are, and which language they speak. 2. You define your topics: the categories where you want recommendations. 3. Genezio creates realistic scenarios: from them, Genezio makes multi-step conversations that show real buyer questions with real constraints. 4. The answer engines answer: ChatGPT, Claude, Gemini, and Perplexity each give an independent answer. 5. Genezio analyzes each answer: which brands the AI recommended, which brands it mentioned, which sources it cited, and what it said about each brand. The result is two numbers: - AI Recommendations %: in the conversations where your brand occurred, how frequently did the AI recommend it as a solution? This is your conversion metric. - AI Visibility %: how frequently does your brand occur in all the eligible conversations? This is your presence metric. Together, the two numbers answer this question: do you occur in the answers, and when you occur, do you win?","keywords":""},{"kind":"section","title":"Why AI recommendations matter","heading":"Traditional search (SEO) and AI search (GEO) compared","section":"Documentation","crumbs":["Introduction","Why AI recommendations matter","Traditional search (SEO) and AI search (GEO) compared"],"url":"/docs/introduction/why-ai-recommendations-matter#traditional-search-seo-and-ai-search-geo-compared","text":"How it works: The user searches, gets links, and clicks your link The user has a conversation with the AI, and the AI recommends a brand. What gives the results: The page rank and the click-through rate If the AI recommends your brand for the specific needs of that persona. The risk: Page 2 gets no traffic There is no page 2. The AI recommends you or it does not Key insight: A brand can have rank 1 on Google for \"best project management tool\". The same brand can be fully absent from the recommendation of ChatGPT when a real buyer describes the situation. These are two different games, and most brands play only one of them.","keywords":""},{"kind":"section","title":"Why AI recommendations matter","heading":"What you can do with Genezio","section":"Documentation","crumbs":["Introduction","Why AI recommendations matter","What you can do with Genezio"],"url":"/docs/introduction/why-ai-recommendations-matter#what-you-can-do-with-genezio","text":"","keywords":""},{"kind":"section","title":"Why AI recommendations matter","heading":"See where the AI recommends you and where it does not","section":"Documentation","crumbs":["Introduction","Why AI recommendations matter","See where the AI recommends you and where it does not"],"url":"/docs/introduction/why-ai-recommendations-matter#see-where-the-ai-recommends-you-and-where-it-does-not","text":"Genezio divides your AI Recommendations % by topic, by persona, and by answer engine. You can see exactly which scenarios you win and which scenarios your competitors win.","keywords":""},{"kind":"section","title":"Why AI recommendations matter","heading":"Understand why the AI recommends competitors and not you","section":"Documentation","crumbs":["Introduction","Why AI recommendations matter","Understand why the AI recommends competitors and not you"],"url":"/docs/introduction/why-ai-recommendations-matter#understand-why-the-ai-recommends-competitors-and-not-you","text":"Genezio shows which sources answer engines cite when they recommend competitors. This is your content gap and your content roadmap.","keywords":""},{"kind":"section","title":"Why AI recommendations matter","heading":"Know exactly which content to create","section":"Documentation","crumbs":["Introduction","Why AI recommendations matter","Know exactly which content to create"],"url":"/docs/introduction/why-ai-recommendations-matter#know-exactly-which-content-to-create","text":"Actionable Insights shows specific recommendations: which topics to target, which sources must cite you, and which narratives to build. Each recommendation connects to the personas and scenarios where you lose.","keywords":""},{"kind":"section","title":"Why AI recommendations matter","heading":"Track the improvements over time","section":"Documentation","crumbs":["Introduction","Why AI recommendations matter","Track the improvements over time"],"url":"/docs/introduction/why-ai-recommendations-matter#track-the-improvements-over-time","text":"Your scores change when you publish new content, make your site better, or get new mentions. You can see the direct effect of your marketing work on AI recommendations for the important audiences.","keywords":""},{"kind":"section","title":"Why AI recommendations matter","heading":"Compare your brand with competitors","section":"Documentation","crumbs":["Introduction","Why AI recommendations matter","Compare your brand with competitors"],"url":"/docs/introduction/why-ai-recommendations-matter#compare-your-brand-with-competitors","text":"See your Share of Voice: how frequently the AI recommends your brand, when compared with competitors, across the same personas and scenarios. Your CMO can track this one number each month. For CMOs: Put AI Recommendations % and Share of Voice on your monthly marketing dashboard. These two numbers tell you if your brand wins the conversations that give pipeline.","keywords":""},{"kind":"section","title":"Why AI recommendations matter","heading":"What to read next","section":"Documentation","crumbs":["Introduction","Why AI recommendations matter","What to read next"],"url":"/docs/introduction/why-ai-recommendations-matter#what-to-read-next","text":"- How the 5 agents work: the five types of AI conversations that Genezio runs, and why each type is important - The two KPIs of AI visibility: what AI Recommendations and AI Visibility measure","keywords":""},{"kind":"page","title":"How LLM search works","heading":"","section":"Documentation","crumbs":["Documentation","Introduction","How LLM search works"],"url":"/docs/introduction/how-llm-search-works","text":"LLM search changes one user question into one AI answer in five steps, from query expansion to synthesis. Learn what each step means for brand visibility. LLM search is the process in which an answer engine such as ChatGPT, Claude, Gemini, or Perplexity gets information from many sources and combines it into one AI-generated answer. This page explains the five steps of LLM search and why they decide which brands occur in AI answers. Answer engines do not give a list of links. Thus, they answer questions in a different way from traditional search engines. When you understand this process, you understand why some brands, sources, and claims occur in AI-generated answers.","keywords":""},{"kind":"section","title":"How LLM search works","heading":"How traditional search engines work","section":"Documentation","crumbs":["Introduction","How LLM search works","How traditional search engines work"],"url":"/docs/introduction/how-llm-search-works#how-traditional-search-engines-work","text":"In traditional search engines, the interaction usually has these steps: 1. A user enters a search query. 2. The search engine gets pages from its index. 3. The search engine ranks the pages by relevance and authority. 4. The user selects one or more links. Thus, the rank position and the click-through rate control the visibility.","keywords":""},{"kind":"section","title":"How LLM search works","heading":"How answer engines work: the LLM search model","section":"Documentation","crumbs":["Introduction","How LLM search works","How answer engines work: the LLM search model"],"url":"/docs/introduction/how-llm-search-works#how-answer-engines-work-the-llm-search-model","text":"Answer engines have a different process. They do not give links. They make a direct answer from information that they get from many sources. This is a simplified flow: 1. The user asks a question. 2. The AI system expands the query into many related searches. 3. The system gets the relevant documents and sources. 4. The model combines the information from those sources. 5. The model makes an answer in natural language. 6. Some systems include citations or links to the sources. The final answer is AI-generated. Thus, the visibility depends on whether the answer includes a brand or a source. A high rank of a web page is not sufficient.","keywords":""},{"kind":"section","title":"How LLM search works","heading":"Step 1: The user question","section":"Documentation","crumbs":["Introduction","How LLM search works","Step 1: The user question"],"url":"/docs/introduction/how-llm-search-works#step-1-the-user-question","text":"The process starts when a user asks a question in natural language. Examples: User query: What are the best running shoes for marathon training? User query: Which CRM is best for startups? User query: What is the best ski resort in Switzerland for beginners? These questions are frequently longer and more conversational than the keywords of a traditional search.","keywords":""},{"kind":"section","title":"How LLM search works","heading":"Step 2: Query expansion","section":"Documentation","crumbs":["Introduction","How LLM search works","Step 2: Query expansion"],"url":"/docs/introduction/how-llm-search-works#step-2-query-expansion","text":"Answer engines almost never use only the original question. They expand it into many related queries to get more information. These expanded searches are query fanouts. For more, read Query fanouts explained. For example, take this question: User query: What are the best running shoes for marathon training? Internally, the AI system can expand it into queries such as these: - best marathon running shoes - long distance running shoes reviews - nike vs adidas marathon shoes - top marathon racing shoes With this process, the AI system examines many perspectives and gets a wider set of sources.","keywords":""},{"kind":"section","title":"How LLM search works","heading":"Step 3: Retrieval of sources","section":"Documentation","crumbs":["Introduction","How LLM search works","Step 3: Retrieval of sources"],"url":"/docs/introduction/how-llm-search-works#step-3-retrieval-of-sources","text":"Then the system gets information from different sources on the web. These sources can include: - Brand websites - Editorial articles - Product reviews - Comparison pages - Forums and communities - News sites The documents that the system gets give the factual material for the final answer. For the factors that control this selection, read How LLMs select sources.","keywords":""},{"kind":"section","title":"How LLM search works","heading":"Step 4: Synthesis","section":"Documentation","crumbs":["Introduction","How LLM search works","Step 4: Synthesis"],"url":"/docs/introduction/how-llm-search-works#step-4-synthesis","text":"After the model gets the relevant information, it analyzes and combines that information. The model does not always quote one page. It can merge information from many sources to make a clear explanation. For example, an answer can include: - A list of recommended products - Summaries of the key advantages - Comparisons between alternatives This synthesis step is one of the main differences between AI-generated answers and traditional search results.","keywords":""},{"kind":"section","title":"How LLM search works","heading":"Step 5: Answer generation","section":"Documentation","crumbs":["Introduction","How LLM search works","Step 5: Answer generation"],"url":"/docs/introduction/how-llm-search-works#step-5-answer-generation","text":"At the end, the model makes an answer in natural language. The answer can be similar to this one: Popular marathon running shoes include the Nike Alphafly, Adidas Adios Pro, and Saucony Endorphin Elite due to their energy return and lightweight design. Some systems also include these items in the answer: - Links to sources - Citations - Suggestions for follow-up questions - Additional context","keywords":""},{"kind":"section","title":"How LLM search works","heading":"Why LLM search changes brand visibility","section":"Documentation","crumbs":["Introduction","How LLM search works","Why LLM search changes brand visibility"],"url":"/docs/introduction/how-llm-search-works#why-llm-search-changes-brand-visibility","text":"Answer engines make an AI-generated answer. Thus, users can read only the answer and not click many links. For this reason, visibility depends on these items: - Whether the answer mentions a brand - Whether the answer cites the website of the brand - Whether the brand occurs in comparisons or recommendations Thus, a brand can have strong search ranks but a small presence in AI answers. This occurs if answer engines almost never refer to the brand or connect it to the topic.","keywords":""},{"kind":"section","title":"How LLM search works","heading":"What LLM search means for organizations","section":"Documentation","crumbs":["Introduction","How LLM search works","What LLM search means for organizations"],"url":"/docs/introduction/how-llm-search-works#what-llm-search-means-for-organizations","text":"Organizations that want to occur in AI-generated answers must understand how answer engines collect and combine information. These are the important factors: - How clearly answer engines connect a brand to a topic - Whether authoritative sources refer to the brand - Whether the content of the brand answers usual user questions - How answer engines show the competitors in the same category Answer engines make their answers from many sources. Thus, visibility depends on a wide web presence, not on one optimized page.","keywords":""},{"kind":"section","title":"How LLM search works","heading":"How Genezio analyzes LLM search","section":"Documentation","crumbs":["Introduction","How LLM search works","How Genezio analyzes LLM search"],"url":"/docs/introduction/how-llm-search-works#how-genezio-analyzes-llm-search","text":"Genezio helps teams understand how LLM search works in practice. Genezio does these tasks: - It runs realistic conversations with AI systems. - It extracts the queries that the model examines. - It identifies the sources that the answers use. - It finds the brand mentions and the citations. - It compares the visibility of the competitors. With this data, organizations can see how AI systems understand their category and how their brand occurs in the answers.","keywords":""},{"kind":"section","title":"How LLM search works","heading":"Next steps","section":"Documentation","crumbs":["Introduction","How LLM search works","Next steps"],"url":"/docs/introduction/how-llm-search-works#next-steps","text":"To learn more about how AI systems get and use information, read these pages: - Query fanouts explained - Citations - Perceptions These pages tell you how Genezio extracts structured insights from AI-generated answers.","keywords":""},{"kind":"page","title":"Query fanouts explained","heading":"","section":"Documentation","crumbs":["Documentation","Introduction","Query fanouts explained"],"url":"/docs/introduction/query-fanouts-explained","text":"Query fanouts are the related searches that an answer engine makes from one question. Learn how query fanouts change AI answers and how Genezio tracks them. Query fanouts are the related searches that an answer engine makes from one user question. An answer engine such as ChatGPT, Claude, Gemini, or Perplexity almost never uses only the exact question to make an answer. This page explains how query fanouts change AI answers and how Genezio uses them. With query fanouts, the AI system examines a topic from different angles, gets more relevant sources, and makes a more complete answer.","keywords":""},{"kind":"section","title":"Query fanouts explained","heading":"What is a query fanout?","section":"Documentation","crumbs":["Introduction","Query fanouts explained","What is a query fanout?"],"url":"/docs/introduction/query-fanouts-explained#what-is-a-query-fanout","text":"A query fanout is an additional search query that the AI system makes internally. The AI system uses it to help answer the question of the user. The system does not search for one phrase only. It makes different variants and related questions. These help the system get information from different sources. For example, a user can ask this question: User query: What are the best running shoes for marathon training? Internally, the AI system can make queries such as these: - best marathon running shoes - long distance running shoe reviews - nike vs adidas marathon shoes - best racing shoes for marathon runners - top cushioned marathon shoes With these queries, the system gets content that shows different perspectives on the topic.","keywords":""},{"kind":"section","title":"Query fanouts explained","heading":"Why answer engines use query fanouts","section":"Documentation","crumbs":["Introduction","Query fanouts explained","Why answer engines use query fanouts"],"url":"/docs/introduction/query-fanouts-explained#why-answer-engines-use-query-fanouts","text":"User questions are frequently wide, ambiguous, or incomplete. One search query can get too little information for a high-quality answer. Query fanouts help the AI system do these tasks: - Examine many interpretations of a question - Get information from different sources - Compare alternatives - Validate claims across many documents When the model makes the search space larger, it gets a more complete understanding of the topic before it makes its answer.","keywords":""},{"kind":"section","title":"Query fanouts explained","heading":"How query fanouts change the answers","section":"Documentation","crumbs":["Introduction","Query fanouts explained","How query fanouts change the answers"],"url":"/docs/introduction/query-fanouts-explained#how-query-fanouts-change-the-answers","text":"The query fanouts control which documents and sources the AI system gets when it collects information. Those sources then have an effect on these items: - Which brands occur in the answer - Which sources the answer cites - Which claims the answer includes - How the answer compares the competitors For example, a query fanout can be this one: Query fanout: best CRM for startups 2024 The AI system can get comparison pages, review articles, and vendor websites about CRMs for startups. Thus, the brands that those sources mention have a higher probability to occur in the final answer. For this reason, query fanouts have a strong effect on AI Recommendations and AI Visibility.","keywords":""},{"kind":"section","title":"Query fanouts explained","heading":"An example of query fanouts in practice","section":"Documentation","crumbs":["Introduction","Query fanouts explained","An example of query fanouts in practice"],"url":"/docs/introduction/query-fanouts-explained#an-example-of-query-fanouts-in-practice","text":"Look at this question: User query: Which CRM is best for startups? The query fanouts can include these: - Query fanout: best CRM software for startups - Query fanout: HubSpot vs Pipedrive for small businesses - Query fanout: affordable CRM tools for startups - Query fanout: startup sales management software - Query fanout: CRM comparison for early stage companies Each of these searches can get different articles, reviews, or vendor pages. Then the AI system combines the information from those sources into one answer.","keywords":""},{"kind":"section","title":"Query fanouts explained","heading":"Why query fanouts are important for brands","section":"Documentation","crumbs":["Introduction","Query fanouts explained","Why query fanouts are important for brands"],"url":"/docs/introduction/query-fanouts-explained#why-query-fanouts-are-important-for-brands","text":"Query fanouts control which sources the AI system gets. Thus, they also have an effect on which brands occur in AI-generated answers. If the sources for those query fanouts frequently mention a brand, the brand has a higher probability to occur in the final answer. If a brand is absent from those sources, it can be absent from the answer. This can occur also when the brand is a strong product in the category. Thus, when organizations understand query fanouts, they can identify these items: - The questions that AI systems examine for a topic - The sources that have an effect on the answer - The areas of the information landscape where their brand is missing","keywords":""},{"kind":"section","title":"Query fanouts explained","heading":"How Genezio uses query fanouts","section":"Documentation","crumbs":["Introduction","Query fanouts explained","How Genezio uses query fanouts"],"url":"/docs/introduction/query-fanouts-explained#how-genezio-uses-query-fanouts","text":"Genezio extracts and analyzes the query fanouts that answer engines make during AI conversations. For the reports in the platform, read Query fanouts. With this data, teams can understand these items: - What the AI system really searches for - Which queries control the sources that the AI system gets - Which topics have an effect on brand visibility - Where there are new content opportunities For example, Genezio can find that AI systems frequently examine queries such as these: - \"best CRM for early stage startups\" - \"CRM tools for SaaS founders\" If a brand has little content or coverage about those topics, that gap can be an opportunity to increase AI Recommendations and AI Visibility.","keywords":""},{"kind":"section","title":"Query fanouts explained","heading":"Query fanouts compared with SEO keywords","section":"Documentation","crumbs":["Introduction","Query fanouts explained","Query fanouts compared with SEO keywords"],"url":"/docs/introduction/query-fanouts-explained#query-fanouts-compared-with-seo-keywords","text":"Query fanouts are similar to keywords. But they show how AI systems examine a topic, not how users type short search queries. Traditional SEO frequently optimizes pages for individual keywords. In AI search, the wider set of query fanouts can be more important. These queries control which sources the AI system gets when it makes the answer.","keywords":""},{"kind":"section","title":"Query fanouts explained","heading":"Next steps","section":"Documentation","crumbs":["Introduction","Query fanouts explained","Next steps"],"url":"/docs/introduction/query-fanouts-explained#next-steps","text":"To learn how query fanouts change AI answers, read these pages: - How LLM search works - Citations - Perceptions These pages tell you how Genezio analyzes AI answers and changes them into structured insights.","keywords":""},{"kind":"section","title":"Query fanouts explained","heading":"In the API","section":"Documentation","crumbs":["Introduction","Query fanouts explained","In the API"],"url":"/docs/introduction/query-fanouts-explained#in-the-api","text":"The public API lists the query fanouts of a brand: List the searches.","keywords":""},{"kind":"page","title":"How LLMs select sources","heading":"","section":"Documentation","crumbs":["Documentation","Introduction","How LLMs select sources"],"url":"/docs/introduction/how-llms-select-sources","text":"LLMs select sources by relevance, authority, clarity, topic coverage, and consistency. Learn how source selection decides which brands occur in AI answers. Large language models (LLMs) select sources in a retrieval stage, before they make an answer. This page explains where the sources come from, the five factors that control source selection, and why source selection is important for brands. An answer engine such as ChatGPT, Claude, Gemini, or Perplexity usually uses information from many sources on the web. These sources give the factual material for the answer. Thus, they explain why some websites, brands, or claims occur in AI-generated answers.","keywords":""},{"kind":"section","title":"How LLMs select sources","heading":"Retrieval before generation: how answer engines get sources","section":"Documentation","crumbs":["Introduction","How LLMs select sources","Retrieval before generation: how answer engines get sources"],"url":"/docs/introduction/how-llms-select-sources#retrieval-before-generation-how-answer-engines-get-sources","text":"Most modern AI search systems use this pattern: the answer engine gets sources and uses them to build its answer. In this process: 1. A user asks a question. 2. The system gets the relevant documents or passages. 3. The answer engine analyzes the information that it got. 4. The model makes an AI-generated answer. The model uses the information that it got as context. Thus, the documents that the system selects in the retrieval stage have a strong effect on the final answer.","keywords":""},{"kind":"section","title":"How LLMs select sources","heading":"Where sources come from","section":"Documentation","crumbs":["Introduction","How LLMs select sources","Where sources come from"],"url":"/docs/introduction/how-llms-select-sources#where-sources-come-from","text":"Usually, AI systems get information from a combination of sources: - Websites that search engines index - Editorial articles - Product reviews - Comparison pages - Documentation sites - Forums and communities - Training data and web sources Some systems use external search engines. Other systems use their own internal indexes or knowledge sources.","keywords":""},{"kind":"section","title":"How LLMs select sources","heading":"Five factors that control source selection","section":"Documentation","crumbs":["Introduction","How LLMs select sources","Five factors that control source selection"],"url":"/docs/introduction/how-llms-select-sources#five-factors-that-control-source-selection","text":"Different factors have an effect on which sources the system gets when it answers a question.","keywords":""},{"kind":"section","title":"How LLMs select sources","heading":"Relevance to the query","section":"Documentation","crumbs":["Introduction","How LLMs select sources","Relevance to the query"],"url":"/docs/introduction/how-llms-select-sources#relevance-to-the-query","text":"The most important factor is the relevance of the content to the question, or to the related queries that the system makes. If a page clearly answers a specific question, the system has a higher probability to get it.","keywords":""},{"kind":"section","title":"How LLMs select sources","heading":"Authority and trust","section":"Documentation","crumbs":["Introduction","How LLMs select sources","Authority and trust"],"url":"/docs/introduction/how-llms-select-sources#authority-and-trust","text":"The system can give priority to sources that many people know as authoritative. These sources can include: - Well-known publications - Established websites - Industry experts - Resources that many sources cite Authority signals can come from links, citations, reputation, and reliability over time.","keywords":""},{"kind":"section","title":"How LLMs select sources","heading":"Content clarity","section":"Documentation","crumbs":["Introduction","How LLMs select sources","Content clarity"],"url":"/docs/introduction/how-llms-select-sources#content-clarity","text":"AI systems frequently find it easier to understand and use content that clearly explains a topic or gives structured information. For example, AI systems can more easily use pages that include these items: - Clear headings - Lists - Comparisons - Short explanations","keywords":""},{"kind":"section","title":"How LLMs select sources","heading":"Coverage of the topic","section":"Documentation","crumbs":["Introduction","How LLMs select sources","Coverage of the topic"],"url":"/docs/introduction/how-llms-select-sources#coverage-of-the-topic","text":"The system has a higher probability to get sources that fully cover a topic. A page can answer many related questions or give deep explanations. The system can select this page more frequently than a page that mentions the topic only briefly.","keywords":""},{"kind":"section","title":"How LLMs select sources","heading":"Consistency across sources","section":"Documentation","crumbs":["Introduction","How LLMs select sources","Consistency across sources"],"url":"/docs/introduction/how-llms-select-sources#consistency-across-sources","text":"Many sources can repeat similar claims or mention the same brands. The AI system can then use those signals as stronger evidence when it makes its answer.","keywords":""},{"kind":"section","title":"How LLMs select sources","heading":"How sources change AI answers","section":"Documentation","crumbs":["Introduction","How LLMs select sources","How sources change AI answers"],"url":"/docs/introduction/how-llms-select-sources#how-sources-change-ai-answers","text":"After the system gets the relevant documents, the model uses them as context to make its answer. In this stage, the model can do these tasks: - Summarize information from many sources - Compare products or services - Merge explanations from different documents - Extract key facts or recommendations The answer is AI-generated. Thus, the final answer can be different from all the individual pages. It shows a combination of the information that the system collected in the retrieval stage.","keywords":""},{"kind":"section","title":"How LLMs select sources","heading":"Citations in AI answers","section":"Documentation","crumbs":["Introduction","How LLMs select sources","Citations in AI answers"],"url":"/docs/introduction/how-llms-select-sources#citations-in-ai-answers","text":"Some AI systems include citations or links. These show which sources had an effect on the answer. These citations can point to: - Articles - Documentation pages - Product pages - Review sites Citations help users verify where the information came from. They also show clearly which sources the system used to make the answer. For more, read How AI citations work.","keywords":""},{"kind":"section","title":"How LLMs select sources","heading":"Why source selection is important for brands","section":"Documentation","crumbs":["Introduction","How LLMs select sources","Why source selection is important for brands"],"url":"/docs/introduction/how-llms-select-sources#why-source-selection-is-important-for-brands","text":"AI systems build answers from the sources that they get. Thus, brands that occur frequently in those sources have a higher probability to occur in AI-generated answers. For this reason, visibility in AI answers depends on these items: - Mentions by relevant sources - Content that answers usual questions - Presence in comparisons and reviews - A connection to the topic across the web If a brand almost never occurs in the sources for a topic, it can have a small visibility in AI-generated answers. To see the sources that answer engines cite for your brand, read Citations.","keywords":""},{"kind":"section","title":"How LLMs select sources","heading":"In the API","section":"Documentation","crumbs":["Introduction","How LLMs select sources","In the API"],"url":"/docs/introduction/how-llms-select-sources#in-the-api","text":"The public API lists the domains that answer engines cite: Cited domains.","keywords":""},{"kind":"page","title":"How AI citations work","heading":"","section":"Documentation","crumbs":["Documentation","Introduction","How AI citations work"],"url":"/docs/introduction/how-ai-citations-work","text":"AI citations are links to the sources behind an AI answer. Learn how ChatGPT, Gemini, and Perplexity select citations and how Genezio analyzes them for brands. AI citations are the links or references to the pages that gave information to an AI-generated answer. This page explains how answer engines such as ChatGPT, Claude, Gemini, and Perplexity select citations, what citations show, and how Genezio analyzes them. AI citations help explain where the information in an answer came from. They also let users examine the original sources.","keywords":""},{"kind":"section","title":"How AI citations work","heading":"What is an AI citation?","section":"Documentation","crumbs":["Introduction","How AI citations work","What is an AI citation?"],"url":"/docs/introduction/how-ai-citations-work#what-is-an-ai-citation","text":"An AI citation is a reference to a source that gave information to an AI-generated answer. Different systems show citations in different forms: - Links to websites - Numbered references - Inline source indicators - Source lists that you can expand For example, an AI answer can include a claim such as this one: The best CRM platforms for startups include HubSpot, Pipedrive, and Zoho CRM. The answer can include citations that point to: - Comparison articles - Software review sites - Vendor documentation - Editorial content These citations help users verify the information and learn more about the topic.","keywords":""},{"kind":"section","title":"How AI citations work","heading":"Why AI systems use citations","section":"Documentation","crumbs":["Introduction","How AI citations work","Why AI systems use citations"],"url":"/docs/introduction/how-ai-citations-work#why-ai-systems-use-citations","text":"Citations have different purposes in AI-generated answers.","keywords":""},{"kind":"section","title":"How AI citations work","heading":"Transparency","section":"Documentation","crumbs":["Introduction","How AI citations work","Transparency"],"url":"/docs/introduction/how-ai-citations-work#transparency","text":"Citations show users where the information came from. Thus, users can verify the claims of the answer.","keywords":""},{"kind":"section","title":"How AI citations work","heading":"Credibility","section":"Documentation","crumbs":["Introduction","How AI citations work","Credibility"],"url":"/docs/introduction/how-ai-citations-work#credibility","text":"Usually, people trust an answer more when reputable sources support it.","keywords":""},{"kind":"section","title":"How AI citations work","heading":"Exploration","section":"Documentation","crumbs":["Introduction","How AI citations work","Exploration"],"url":"/docs/introduction/how-ai-citations-work#exploration","text":"Users can click a citation to open the original article or resource and do more research.","keywords":""},{"kind":"section","title":"How AI citations work","heading":"How answer engines select citations","section":"Documentation","crumbs":["Introduction","How AI citations work","How answer engines select citations"],"url":"/docs/introduction/how-ai-citations-work#how-answer-engines-select-citations","text":"Usually, citations come from the documents that the system gets in the retrieval stage of the AI search process. The process usually has these steps: 1. The user asks a question. 2. The system makes related queries, the query fanouts. 3. The system gets the relevant documents. 4. The model analyzes those documents. 5. The model makes an answer from the information. 6. The system shows some of the sources that it got as citations. Citations come from the documents that the system got. Thus, they show the sources that the model thought were the most relevant or useful for the answer.","keywords":""},{"kind":"section","title":"How AI citations work","heading":"What citations show","section":"Documentation","crumbs":["Introduction","How AI citations work","What citations show"],"url":"/docs/introduction/how-ai-citations-work#what-citations-show","text":"Citations give important signals about how AI systems understand a topic. They show these items: - Which websites have an effect on AI answers - Which sources AI systems think are authoritative - Where the information about a topic is concentrated - Which brands AI systems connect to a topic For example, many AI systems can frequently cite the same review sites or editorial articles for a category. Then those sources probably have a strong effect on how AI systems understand that topic.","keywords":""},{"kind":"section","title":"How AI citations work","heading":"AI citations compared with brand mentions","section":"Documentation","crumbs":["Introduction","How AI citations work","AI citations compared with brand mentions"],"url":"/docs/introduction/how-ai-citations-work#ai-citations-compared-with-brand-mentions","text":"A brand can occur in an AI answer when the answer does not cite the website of the brand. Two cases are usual.","keywords":""},{"kind":"section","title":"How AI citations work","heading":"Brand mention without a citation","section":"Documentation","crumbs":["Introduction","How AI citations work","Brand mention without a citation"],"url":"/docs/introduction/how-ai-citations-work#brand-mention-without-a-citation","text":"The model mentions a brand from information that it got from third-party sources. Example: Popular project management tools include Notion, Asana, and Monday.com. In this case, the answer can cite review articles and not the official product websites.","keywords":""},{"kind":"section","title":"How AI citations work","heading":"Citation of the brand website","section":"Documentation","crumbs":["Introduction","How AI citations work","Citation of the brand website"],"url":"/docs/introduction/how-ai-citations-work#citation-of-the-brand-website","text":"The answer includes a direct reference to the website or the documentation of the brand. Example: According to the Notion documentation, teams can use databases to manage tasks and workflows. The two types of visibility are both important. But they show different types of influence.","keywords":""},{"kind":"section","title":"How AI citations work","heading":"Why citations are important for AI recommendations","section":"Documentation","crumbs":["Introduction","How AI citations work","Why citations are important for AI recommendations"],"url":"/docs/introduction/how-ai-citations-work#why-citations-are-important-for-ai-recommendations","text":"Citations help explain why some brands occur in AI answers. A brand or its website can get frequent citations. This is a signal that the AI system thinks those sources are useful or relevant for questions about the topic. Brands that almost never occur in cited sources can have a lower visibility in AI answers. Thus, when organizations know which sources get citations, they can identify these items: - The influential websites in their category - The content that has an effect on AI answers - The opportunities to increase visibility","keywords":""},{"kind":"section","title":"How AI citations work","heading":"How Genezio analyzes citations","section":"Documentation","crumbs":["Introduction","How AI citations work","How Genezio analyzes citations"],"url":"/docs/introduction/how-ai-citations-work#how-genezio-analyzes-citations","text":"Genezio extracts the citations from AI-generated answers. It analyzes them across many conversations. For the citation reports in the platform, read Citations. With this data, teams can see these items: - Which sources get the most citations - Which sources mention their brand - Which sources mention competitors - Which sources are dominant for a specific topic Genezio collects this information in one place. Thus, it helps organizations understand the information landscape that AI systems use.","keywords":""},{"kind":"section","title":"How AI citations work","heading":"Example: the citations of a sales pipeline question","section":"Documentation","crumbs":["Introduction","How AI citations work","Example: the citations of a sales pipeline question"],"url":"/docs/introduction/how-ai-citations-work#example-the-citations-of-a-sales-pipeline-question","text":"Look at this question: User query: What are the best tools for managing a startup sales pipeline? An AI system can make an answer that mentions these brands: - HubSpot - Pipedrive - Salesforce The citations can include: - Software comparison websites - SaaS review platforms - Blog articles that compare CRM tools The answer can have no direct citation of the official vendor websites. But the third-party sources that mention those brands still have an effect on the answer.","keywords":""},{"kind":"section","title":"How AI citations work","heading":"Next steps","section":"Documentation","crumbs":["Introduction","How AI citations work","Next steps"],"url":"/docs/introduction/how-ai-citations-work#next-steps","text":"To learn more about how Genezio analyzes AI answers, read these pages: - Perceptions - Share of Voice These pages tell you how Genezio changes citations and brand mentions into visibility insights that you can measure.","keywords":""},{"kind":"section","title":"How AI citations work","heading":"In the API","section":"Documentation","crumbs":["Introduction","How AI citations work","In the API"],"url":"/docs/introduction/how-ai-citations-work#in-the-api","text":"The public API lists the domains and the pages that answer engines cite: Cited domains and Cited pages.","keywords":""},{"kind":"page","title":"The KPIs of AI visibility","heading":"","section":"Documentation","crumbs":["Documentation","Introduction","The KPIs of AI visibility"],"url":"/docs/introduction/what-kpis-are-we-measuring","text":"Genezio measures brand performance in AI answers with two KPIs: AI Recommendations and AI Visibility. Learn what each KPI measures and where it comes from. Genezio uses two core KPIs to measure the performance of your brand in AI-generated answers: AI Recommendations and AI Visibility. This page tells you what each KPI measures, which conversations give the KPIs, and how to read the two KPIs together. The two KPIs answer different questions. Read them together.","keywords":""},{"kind":"section","title":"The KPIs of AI visibility","heading":"AI Recommendations: the conversion KPI","section":"Documentation","crumbs":["Introduction","The KPIs of AI visibility","AI Recommendations: the conversion KPI"],"url":"/docs/introduction/what-kpis-are-we-measuring#ai-recommendations-the-conversion-kpi","text":"AI Recommendations is the percentage of the conversations with your brand in which the AI recommends your brand. The denominator is the number of conversations where your brand was visible. Thus, this KPI is a conversion rate from presence to recommendation. AI Recommendations helps you understand if answer engines trust your brand sufficiently to suggest it as a solution when users ask for one.","keywords":""},{"kind":"section","title":"The KPIs of AI visibility","heading":"AI Visibility: the presence KPI","section":"Documentation","crumbs":["Introduction","The KPIs of AI visibility","AI Visibility: the presence KPI"],"url":"/docs/introduction/what-kpis-are-we-measuring#ai-visibility-the-presence-kpi","text":"AI Visibility is the percentage of eligible conversations where your brand occurs in AI-generated answers. This includes the cases where the answer does one of these actions with your brand: - It mentions your brand. - It puts your brand in a list of recommendations. - It includes your brand in comparisons. AI Visibility gives a wider view of how frequently your brand is present in AI answers across the relevant conversations. For the formula, read How Genezio measures visibility.","keywords":""},{"kind":"section","title":"The KPIs of AI visibility","heading":"Which agents give the KPIs","section":"Documentation","crumbs":["Introduction","The KPIs of AI visibility","Which agents give the KPIs"],"url":"/docs/introduction/what-kpis-are-we-measuring#which-agents-give-the-kpis","text":"The two metrics are brand-level measurements. Genezio calculates them from the conversations of the Prompter agent and the Recommender agent. In these two agent types, the AI selects without help which brands to show. The Comparer agent and the Introspector agent have different purposes: - The Comparer is a competitive analysis tool. It uses the metrics of your brand to show how you compare with specific competitors. - The Introspector analyzes how the AI describes your brand. These two agents do not have an effect on the calculation of AI Recommendations or AI Visibility. For the role of each agent, read How the 5 agents work.","keywords":""},{"kind":"section","title":"The KPIs of AI visibility","heading":"How to read the two KPIs together","section":"Documentation","crumbs":["Introduction","The KPIs of AI visibility","How to read the two KPIs together"],"url":"/docs/introduction/what-kpis-are-we-measuring#how-to-read-the-two-kpis-together","text":"Use the two metrics together: - AI Recommendations tells you how frequently answer engines recommend your brand when it occurs in a conversation. - AI Visibility tells you how frequently your brand occurs at all in the eligible conversations. A brand can have a high visibility and a lower recommendation rate, or the opposite. When you read the two KPIs together, you get a more complete view of the AI performance.","keywords":""},{"kind":"section","title":"The KPIs of AI visibility","heading":"Next steps","section":"Documentation","crumbs":["Introduction","The KPIs of AI visibility","Next steps"],"url":"/docs/introduction/what-kpis-are-we-measuring#next-steps","text":"To understand where these KPIs come from, read these pages: - Brand recommendation - Brand visibility - Topic and scenario relevance - Your KPIs explained","keywords":""},{"kind":"section","title":"The KPIs of AI visibility","heading":"In the API","section":"Documentation","crumbs":["Introduction","The KPIs of AI visibility","In the API"],"url":"/docs/introduction/what-kpis-are-we-measuring#in-the-api","text":"The public API reads the two KPIs over time: Recommendation over time and Visibility over time.","keywords":""},{"kind":"page","title":"How visibility is measured","heading":"","section":"Documentation","crumbs":["Documentation","Introduction","How visibility is measured"],"url":"/docs/introduction/how-genezio-measures-visibility","text":"AI Visibility is the percentage of eligible AI conversations that mention your brand. Learn the formula, what counts as an appearance, and which agents count. Genezio measures AI Visibility as the percentage of eligible AI conversations in which your brand occurs. This page explains the visibility formula, what counts as a brand appearance, and which Genezio agent types give the KPIs. The most important signal that Genezio tracks is AI Recommendations: does the answer engine recommend the brand as a solution? AI Visibility measures the wider presence. It shows how frequently your brand occurs in AI answers at all, as a mention, a comparison, or a recommendation. The visibility metric answers one simple question: in how many AI conversations does your brand occur?.","keywords":""},{"kind":"section","title":"How visibility is measured","heading":"The AI Visibility formula","section":"Documentation","crumbs":["Introduction","How visibility is measured","The AI Visibility formula"],"url":"/docs/introduction/how-genezio-measures-visibility#the-ai-visibility-formula","text":"Genezio calculates AI Visibility with this formula: For example, a brand occurs in 42 of the 100 conversations that Genezio analyzes. The AI Visibility of the brand for that set of scenarios is: This percentage shows clearly how frequently the brand occurs when AI systems answer questions about a specific topic.","keywords":"Visibility = (Conversations where the brand appears / Total eligible conversations) x 100 42%"},{"kind":"section","title":"How visibility is measured","heading":"What counts as a brand appearance","section":"Documentation","crumbs":["Introduction","How visibility is measured","What counts as a brand appearance"],"url":"/docs/introduction/how-genezio-measures-visibility#what-counts-as-a-brand-appearance","text":"A brand is visible in a conversation when the AI assistant does one of these actions: - It mentions the brand explicitly. - It puts the brand in a list of recommendations. - It compares the brand with competitors. - It refers to the brand as part of an explanation. The brand does not have to be the main subject of the answer. Each clear mention of the brand counts as visibility for that conversation.","keywords":""},{"kind":"section","title":"How visibility is measured","heading":"Genezio agent types and AI Visibility","section":"Documentation","crumbs":["Introduction","How visibility is measured","Genezio agent types and AI Visibility"],"url":"/docs/introduction/how-genezio-measures-visibility#genezio-agent-types-and-ai-visibility","text":"Genezio runs different types of conversations to simulate realistic interactions with AI systems. Each Genezio agent type shows a different user intent. For all five agents, read How the 5 agents work.","keywords":""},{"kind":"section","title":"How visibility is measured","heading":"Prompter agent","section":"Documentation","crumbs":["Introduction","How visibility is measured","Prompter agent"],"url":"/docs/introduction/how-genezio-measures-visibility#prompter-agent","text":"A Prompter agent conversation simulates a user who asks an open question about a category. Example: User query: What CRM should a startup use? The AI assistant answers with recommendations or with a list of relevant products. Prompter agent conversations are one of the primary signals for the measurement of visibility. They show how brands occur when users ask general discovery questions.","keywords":""},{"kind":"section","title":"How visibility is measured","heading":"Recommender agent","section":"Documentation","crumbs":["Introduction","How visibility is measured","Recommender agent"],"url":"/docs/introduction/how-genezio-measures-visibility#recommender-agent","text":"A Recommender agent conversation simulates a user who explicitly asks the AI to recommend solutions. Example: User query: Recommend project management tools for small teams. Genezio also uses these conversations to calculate visibility. They show which brands AI systems recommend for usual requests.","keywords":""},{"kind":"section","title":"How visibility is measured","heading":"Introspector agent","section":"Documentation","crumbs":["Introduction","How visibility is measured","Introspector agent"],"url":"/docs/introduction/how-genezio-measures-visibility#introspector-agent","text":"An Introspector agent conversation asks the AI system about a specific brand. Example: User query: What is HubSpot used for? The question includes the brand name. Thus, the AI assistant almost always mentions the brand in its answer. For this reason, Genezio does not use Introspector agent conversations to calculate AI Visibility. These conversations make the visibility score artificially high.","keywords":""},{"kind":"section","title":"How visibility is measured","heading":"Comparer agent","section":"Documentation","crumbs":["Introduction","How visibility is measured","Comparer agent"],"url":"/docs/introduction/how-genezio-measures-visibility#comparer-agent","text":"A Comparer agent conversation asks the AI system to compare specific brands. Example: User query: HubSpot vs Pipedrive: which CRM is better for startups? In this type of conversation, the brands are in the prompt. The model almost always refers to those brands in the answer. Thus, Genezio does not use Comparer agent conversations to calculate the KPIs. The Comparer is a competitive analysis tool. It uses the brand-level metrics of your brand to show how you compare with specific competitors. It does not have an effect on the calculation of AI Visibility or AI Recommendations.","keywords":""},{"kind":"section","title":"How visibility is measured","heading":"Which agent types give KPIs","section":"Documentation","crumbs":["Introduction","How visibility is measured","Which agent types give KPIs"],"url":"/docs/introduction/how-genezio-measures-visibility#which-agent-types-give-kpis","text":"Prompter: ✓ AI Visibility Open discovery. The AI selects which brands to mention. Recommender: ✓ AI Recommendations + AI Visibility Driven by the scenario. The AI selects which brands to recommend. Introspector: — Narrative analysis. It examines how the AI describes your brand. Comparer: — Competitive analysis. It compares your brand with specific competitors The two KPIs are brand-level metrics. Genezio calculates them from the Prompter and Recommender conversations, where the AI selects without help which brands to show. Thus, the metrics show the organic presence of the brand and real recommendation decisions.","keywords":""},{"kind":"section","title":"How visibility is measured","heading":"AI Visibility for each topic","section":"Documentation","crumbs":["Introduction","How visibility is measured","AI Visibility for each topic"],"url":"/docs/introduction/how-genezio-measures-visibility#ai-visibility-for-each-topic","text":"Usually, Genezio measures visibility in the context of a specific topic or set of scenarios. For example, a brand can have these values: - 60% visibility for \"CRM for startups\" - 35% visibility for \"sales tools for small businesses\" - 10% visibility for \"marketing automation platforms\" This helps organizations find the topics where AI systems strongly connect their brand to the topic, and the topics where the visibility is small. To group many topics by theme, use topic tags.","keywords":""},{"kind":"section","title":"How visibility is measured","heading":"Why AI Visibility is a useful metric","section":"Documentation","crumbs":["Introduction","How visibility is measured","Why AI Visibility is a useful metric"],"url":"/docs/introduction/how-genezio-measures-visibility#why-ai-visibility-is-a-useful-metric","text":"AI Visibility gives a high-level signal about how AI systems see a brand in a category. A higher visibility score shows these items: - AI systems frequently connect the brand to the topic. - The brand occurs in recommendations. - The sources that AI systems use frequently refer to the brand. A lower score can show that competitors are dominant in the conversation. It can also show that the information sources of the AI models refer to the brand too little.","keywords":""},{"kind":"section","title":"How visibility is measured","heading":"Next steps","section":"Documentation","crumbs":["Introduction","How visibility is measured","Next steps"],"url":"/docs/introduction/how-genezio-measures-visibility#next-steps","text":"To understand how your visibility compares with competitors, read this page: - Share of Voice This page tells you how Genezio expands the visibility analysis into deeper competitive insights.","keywords":""},{"kind":"section","title":"How visibility is measured","heading":"In the API","section":"Documentation","crumbs":["Introduction","How visibility is measured","In the API"],"url":"/docs/introduction/how-genezio-measures-visibility#in-the-api","text":"The public API reads the visibility of a brand over time and by topic: Visibility over time and Visibility by topic.","keywords":""},{"kind":"page","title":"How the 5 agents work","heading":"","section":"Documentation","crumbs":["Documentation","Introduction","How the 5 agents work"],"url":"/docs/introduction/how-the-5-agents-work","text":"Genezio runs five agents: Prompter, Recommender, Introspector, Comparer, and Fact Checker. Learn what each agent simulates and which agents give your KPIs. The 5 Genezio agents are five types of AI conversations that measure the presence of your brand in answer engines. Each agent answers a different question. Not all agents count for your score. This is intentional, and this page tells you why. A comparison: Think of five different mystery shoppers that examine your brand. Each one goes into a store (or asks an AI) with a different brief. Their results show very different things about how people see your brand.","keywords":""},{"kind":"section","title":"How the 5 agents work","heading":"Agent 1: Prompter agent for open discovery","section":"Documentation","crumbs":["Introduction","How the 5 agents work","Agent 1: Prompter agent for open discovery"],"url":"/docs/introduction/how-the-5-agents-work#agent-1-prompter-agent-for-open-discovery","text":"Counts for: AI Visibility % What it simulates: A user at the top of the funnel. The user examines a category and did not select a brand. The user asks open discovery questions. Example queries: - \"What are the best CRM tools for a startup?\" - \"What software should I use for project management?\" - \"What tools do marketing teams use for analytics?\" Why it is important for you: Most buyers start their research in this way. If you are absent here, the AI removes you before the shortlist exists. Prompter conversations have the largest effect on AI Visibility %. For the full feature page, read Prompter agent. Content implication: Category pillar pages · \"Best of\" content · In-depth guides · Original research","keywords":""},{"kind":"section","title":"How the 5 agents work","heading":"Agent 2: Recommender agent for purchase decisions","section":"Documentation","crumbs":["Introduction","How the 5 agents work","Agent 2: Recommender agent for purchase decisions"],"url":"/docs/introduction/how-the-5-agents-work#agent-2-recommender-agent-for-purchase-decisions","text":"Counts for: AI Recommendations % and AI Visibility % What it simulates: A user who is ready to make a decision. The user did some research. Now the user wants the AI to select a specific item for their situation. These conversations frequently have many steps. Example queries: - \"I'm a solo founder with a team of 3. Recommend a CRM that's easy to set up and under $50/month.\" - \"What's the best project management tool for a remote design team of 10?\" Why it is important for you: This is the moment of highest intent in the AI buyer journey. A user asks the AI to recommend a solution. If the AI does not give your name, you lose a qualified lead. AI Recommendations % is your conversion metric. Content implication: Use-case landing pages · Persona-specific content · ROI calculators","keywords":""},{"kind":"section","title":"How the 5 agents work","heading":"Agent 3: Introspector agent for brand narrative","section":"Documentation","crumbs":["Introduction","How the 5 agents work","Agent 3: Introspector agent for brand narrative"],"url":"/docs/introduction/how-the-5-agents-work#agent-3-introspector-agent-for-brand-narrative","text":"Does not count for the KPIs What it simulates: A user who knows your brand and wants to learn more. The user asks the AI directly about your brand, by name. Example queries: - \"What is HubSpot?\" - \"What does Salesforce do?\" - \"Tell me about Notion's pricing model.\" Why it does not count for your score: If the user mentions your brand in the question, the AI almost always mentions it in the answer. Thus, this result shows no real visibility. These conversations make your Visibility score artificially high and useless as a benchmark. What it is useful for: It helps you understand how the AI describes your brand: - Which words does the AI use? - Which strengths and weaknesses does it mention? - Which claims does it make? - Are those claims accurate? This data goes directly into your Perceptions analysis, the accuracy layer below the score. It also goes into your Statements analysis, the qualitative layer below the score. Content implication: Brand narrative analysis · Claim accuracy monitoring · Sentiment monitoring · Positioning audit","keywords":""},{"kind":"section","title":"How the 5 agents work","heading":"Agent 4: Comparer agent for competitive analysis","section":"Documentation","crumbs":["Introduction","How the 5 agents work","Agent 4: Comparer agent for competitive analysis"],"url":"/docs/introduction/how-the-5-agents-work#agent-4-comparer-agent-for-competitive-analysis","text":"Does not count for the KPIs. This is a competitive analysis tool. What it simulates: A user who compares you with a specific competitor. The user has fewer options now and wants a direct verdict between two brands. Example queries: - \"HubSpot vs Pipedrive: which is better for a startup?\" - \"Notion vs Asana for a remote team?\" - \"Should I use Salesforce or HubSpot?\" Why it does not count for your score: The question names the two brands. Thus, the two brands always occur in the answer. As with the Introspector, these conversations make your metrics incorrect. What it is for: The Comparer is an analysis tool. It takes the brand-level KPIs that Genezio calculated from the Prompter and Recommender conversations, and puts them in a competitive context. The Comparer also runs direct AI conversations between two brands. These conversations show how answer engines compare you with specific competitors: - Which strengths does the answer engine show? - Which weaknesses does it mention? - Which brand does it select as the better match? - Does it show you as better for large enterprises? As more expensive? As easier to use? This shows the narrative that you must make stronger, or correct. Prompter and Recommender tell you your score. Comparer tells you why you win or lose against a specific competitor. Content implication: Competitive intelligence · \"vs\" comparison page strategy · Win/loss narrative","keywords":""},{"kind":"section","title":"How the 5 agents work","heading":"Agent 5: Fact Checker agent for factual accuracy","section":"Documentation","crumbs":["Introduction","How the 5 agents work","Agent 5: Fact Checker agent for factual accuracy"],"url":"/docs/introduction/how-the-5-agents-work#agent-5-fact-checker-agent-for-factual-accuracy","text":"Does not count for the KPIs. This is an accuracy measurement tool. What it simulates: A factual probe. The agent does not ask the answer engine open questions. You give it specific claims about your brand that it can verify, for example pricing, features, founding year, customer counts, or certifications. The agent then measures if the answer engine confirms, contradicts, or avoids each claim. Example claims: - \"Our product offers a free tier.\" - \"We were founded in 2019.\" - \"We are SOC 2 Type II certified.\" Results for each claim: - True: the answer engine confirms the claim. - False: the answer engine contradicts the claim. - Indecisive: the answer engine does not take a clear position. Why it does not count for your score: The prompt states the claims. Thus, the conversation does not measure discovery. It measures accuracy, which is a different axis from visibility and recommendation. What it is for: The Fact Checker is most useful for brands where factual accuracy has a direct effect on buyer trust or regulatory standing. Examples are security, financial services, healthcare, and other regulated industries. The Fact Checker also works with the Knowledge Base and Grounded Perceptions. Together, they make a complete accuracy stack: - The Knowledge Base defines the truth. - The Fact Checker tests if answer engines confirm the truth. - Grounded Perceptions show if answer engines match the truth without a prompt about it. Content implication: Update the product, pricing, and certification pages, so that the correct facts are public and crawlers can read them · Correct third-party listings · Publish authoritative reference content for high-risk claims For the full feature page, read Fact Checker agent.","keywords":""},{"kind":"section","title":"How the 5 agents work","heading":"The 5 agents compared","section":"Documentation","crumbs":["Introduction","How the 5 agents work","The 5 agents compared"],"url":"/docs/introduction/how-the-5-agents-work#the-5-agents-compared","text":"Prompter: Examines a category No ✓ Measures AI Visibility Awareness measurement. Recommender: Ready to decide No ✓ Measures AI Recommendations + AI Visibility Intent measurement. Introspector: Learns about a brand Yes Narrative analysis (not KPIs) Narrative, sentiment, and accuracy audit. Comparer: Compares options Yes (both) Competitive analysis (not KPIs) Competitive positioning. Fact Checker: Verifies specific claims Yes Accuracy measurement (not KPIs) Monitoring of hallucinations and facts Two agents measure. Three agents analyze. Prompter and Recommender give your KPIs. Introspector, Comparer, and Fact Checker help you understand the narratives, the competitive dynamics, and the factual accuracy behind those numbers. All five agents are necessary, but they have different roles.","keywords":""},{"kind":"section","title":"How the 5 agents work","heading":"Content strategy for each agent","section":"Documentation","crumbs":["Introduction","How the 5 agents work","Content strategy for each agent"],"url":"/docs/introduction/how-the-5-agents-work#content-strategy-for-each-agent","text":"Each agent type maps directly to content that you must create: - Prompter: content at the category level. - Recommender: use-case landing pages. - Introspector: your own brand content and PR. - Comparer: \"vs\" comparison pages. The next page tells you more about this. Next: From data to content strategy","keywords":""},{"kind":"page","title":"Core Concepts","heading":"","section":"Documentation","crumbs":["Documentation","Core concepts","Core Concepts"],"url":"/docs/core-concepts","text":"Learn the data model behind each number in Genezio: brands, topics, scenarios, conversations, citations, competitors, and perceptions. The core concepts are the data model behind each number in Genezio. Read them early. They take approximately twenty minutes. Most problems of understanding later come from one of these concepts.","keywords":""},{"kind":"section","title":"Core Concepts","heading":"How the concepts connect","section":"Documentation","crumbs":["Core concepts","Core Concepts","How the concepts connect"],"url":"/docs/core-concepts#how-the-concepts-connect","text":"- A brand is the item that buyers ask about. - A topic is a subject that buyers ask about. - A scenario is one realistic way to ask about a topic. - When Genezio runs a scenario on an answer engine, the result is a conversation. Each metric comes from conversations.","keywords":""},{"kind":"section","title":"Core Concepts","heading":"What Genezio finds in an answer","section":"Documentation","crumbs":["Core concepts","Core Concepts","What Genezio finds in an answer"],"url":"/docs/core-concepts#what-genezio-finds-in-an-answer","text":"In each answer, Genezio finds these items: - Citations: the sources that the answer uses. - Competitors: the brands that the answer names as alternatives. - Perceptions: the claims that the answer makes about you.","keywords":""},{"kind":"section","title":"Core Concepts","heading":"Where to start","section":"Documentation","crumbs":["Core concepts","Core Concepts","Where to start"],"url":"/docs/core-concepts#where-to-start","text":"If you read only two pages, read Topics and Conversations.","keywords":""},{"kind":"page","title":"Brands","heading":"","section":"Documentation","crumbs":["Documentation","Core concepts","Brands"],"url":"/docs/core-concepts/brands","text":"A brand is the company, product, or service that Genezio monitors in AI answers. Learn the data of a brand and how to model a company with many products. In Genezio, a brand is the organization, the product, or the service that you monitor in the answers of answer engines. Each analysis in Genezio starts from a brand, and each metric shows how that brand appears in AI answers. Genezio examines topics, scenarios, conversations, and insights in relation to one brand.","keywords":""},{"kind":"section","title":"Brands","heading":"What a brand represents","section":"Documentation","crumbs":["Core concepts","Brands","What a brand represents"],"url":"/docs/core-concepts/brands#what-a-brand-represents","text":"A brand is usually one of these items: - A company - A product - A service - A platform Examples: - HubSpot - Nike - Stripe - Notion The brand is the entity that you monitor when users ask questions to answer engines such as ChatGPT, Claude, Gemini, or Perplexity.","keywords":""},{"kind":"section","title":"Brands","heading":"The data of a brand","section":"Documentation","crumbs":["Core concepts","Brands","The data of a brand"],"url":"/docs/core-concepts/brands#the-data-of-a-brand","text":"When you create a brand in Genezio, Genezio collects or makes these items of data. For the procedure, refer to Describe your brand.","keywords":""},{"kind":"section","title":"Brands","heading":"Brand name","section":"Documentation","crumbs":["Core concepts","Brands","Brand name"],"url":"/docs/core-concepts/brands#brand-name","text":"The brand name is the official name of the organization, the product, or the service. Example: Genezio uses this name to find the mentions of the brand in the answers of answer engines.","keywords":"HubSpot"},{"kind":"section","title":"Brands","heading":"Website URL","section":"Documentation","crumbs":["Core concepts","Brands","Website URL"],"url":"/docs/core-concepts/brands#website-url","text":"The website URL is the primary website of the brand. Example: The website tells Genezio the domain of the brand. Genezio uses it to find the citations that point to the content of the brand.","keywords":"https://hubspot.com"},{"kind":"section","title":"Brands","heading":"Brand description","section":"Documentation","crumbs":["Core concepts","Brands","Brand description"],"url":"/docs/core-concepts/brands#brand-description","text":"Genezio writes a short description. The description tells what the brand does and the category of the brand. Example: HubSpot is a CRM and marketing automation platform designed to help businesses manage customer relationships, marketing campaigns, and sales pipelines. This description helps Genezio know how the brand must show in conversations.","keywords":""},{"kind":"section","title":"Brands","heading":"Customer profile","section":"Documentation","crumbs":["Core concepts","Brands","Customer profile"],"url":"/docs/core-concepts/brands#customer-profile","text":"Each brand also has a customer profile. The customer profile describes the typical user or buyer. Example: - Startup founders - Small business owners - Marketing teams Genezio uses this profile to make realistic user questions when it creates scenarios.","keywords":""},{"kind":"section","title":"Brands","heading":"Why the brand is the center of Genezio","section":"Documentation","crumbs":["Core concepts","Brands","Why the brand is the center of Genezio"],"url":"/docs/core-concepts/brands#why-the-brand-is-the-center-of-genezio","text":"Genezio examines the answers of answer engines to find: - If the brand shows in the answer - How frequently the brand shows - Which sources mention the brand - How the answer describes the brand - Which competitors show together with the brand Thus, the brand is the reference point for all the visibility measurements. Genezio calculates each insight in relation to one brand. Examples of insights are visibility scores, share of voice, and competitor analysis.","keywords":""},{"kind":"section","title":"Brands","heading":"Many brands in one account","section":"Documentation","crumbs":["Core concepts","Brands","Many brands in one account"],"url":"/docs/core-concepts/brands#many-brands-in-one-account","text":"An account can have many brands. This is useful for: - Agencies that manage many clients - Companies that monitor many products - Organizations that monitor many markets Each brand has its own topics, scenarios, conversations, and insights.","keywords":""},{"kind":"section","title":"Brands","heading":"Companies with many products","section":"Documentation","crumbs":["Core concepts","Brands","Companies with many products"],"url":"/docs/core-concepts/brands#companies-with-many-products","text":"If your company has many different products, you can model them in Genezio in two ways. Select the correct way, because it changes how Genezio shows each metric.","keywords":""},{"kind":"section","title":"Brands","heading":"Option 1: one brand for each product","section":"Documentation","crumbs":["Core concepts","Brands","Option 1: one brand for each product"],"url":"/docs/core-concepts/brands#option-1-one-brand-for-each-product","text":"Use this option when each product has its own brand presence in answer engines: - Answer engines know the products by their product names. - Customers use the product name, not the name of the parent company. - Each product has its own competitors. In this case, make each product a separate brand. Each brand has its own topics, scenarios, AI Recommendations score, Visibility score, and competitors. Example: Microsoft. Do not make one \"Microsoft\" brand. Make these brands: - Microsoft Word: competes with Google Docs, Notion, and others. - Microsoft Excel: competes with Google Sheets, Airtable, and others. - Microsoft Teams: competes with Slack, Zoom, and others. Example: Bitdefender. This cybersecurity company has many product lines: - Bitdefender Total Security: consumer antivirus. It competes with Norton and McAfee. - Bitdefender GravityZone: enterprise endpoint protection. It competes with CrowdStrike and SentinelOne. Separate brands keep these groups of competitors separate.","keywords":""},{"kind":"section","title":"Brands","heading":"Option 2: one brand with Master Filters","section":"Documentation","crumbs":["Core concepts","Brands","Option 2: one brand with Master Filters"],"url":"/docs/core-concepts/brands#option-2-one-brand-with-master-filters","text":"Use this option when the parent brand is strong, but users do not know the products as separate brands: - Customers know the company name, not the product names. - Answer engines describe the products as products of the parent brand. - The products frequently have the same competitors. In this case, keep all the products in one brand. Use the Master Filters feature of Genezio to filter the data for each product line. The master filter picker shows in the main header. With it, you can change between the product lines, or keep the \"all\" view for the full brand. Refer to Master Filters for the full configuration.","keywords":""},{"kind":"section","title":"Brands","heading":"One brand without Master Filters","section":"Documentation","crumbs":["Core concepts","Brands","One brand without Master Filters"],"url":"/docs/core-concepts/brands#one-brand-without-master-filters","text":"Use one brand, and do not use Master Filters, when: - The company name and the product name are the same (for example Notion, Stripe, Slack). - You want to measure the visibility of the company, not of each product. - Your products have the same competitors and the same buyer persona.","keywords":""},{"kind":"section","title":"Brands","heading":"Quick decision","section":"Documentation","crumbs":["Core concepts","Brands","Quick decision"],"url":"/docs/core-concepts/brands#quick-decision","text":"Each product has its own brand recognition and its own competitors: Separate brands. The parent brand is strong, and the products do not have their own brand presence: One brand + Master Filters. The company and the product have the same name: Single brand, no Master Filters","keywords":""},{"kind":"section","title":"Brands","heading":"Brand users and access","section":"Documentation","crumbs":["Core concepts","Brands","Brand users and access"],"url":"/docs/core-concepts/brands#brand-users-and-access","text":"The owner of a Genezio account can invite team members to the full account or to specific brands in the account. Thus, organizations control who can see the analysis of each brand. For example: - Growth leads and SEO leads possibly need access to all brands. - A client stakeholder possibly needs access to one brand only. - An agency can give each team member access to the brand of one client. For the two access levels, refer to Users.","keywords":""},{"kind":"section","title":"Brands","heading":"Competitor brands","section":"Documentation","crumbs":["Core concepts","Brands","Competitor brands"],"url":"/docs/core-concepts/brands#competitor-brands","text":"During the analysis, Genezio also finds the competitor brands that show in the same answers. Genezio finds these competitors automatically from: - The mentions in the answers - The recommendations - The comparisons When you monitor competitors, you can see the position of your brand in the category. Refer to Competitors.","keywords":""},{"kind":"section","title":"Brands","heading":"Example: a brand and its competitors in one answer","section":"Documentation","crumbs":["Core concepts","Brands","Example: a brand and its competitors in one answer"],"url":"/docs/core-concepts/brands#example-a-brand-and-its-competitors-in-one-answer","text":"The analyzed brand is: A user asks an answer engine: The answer engine can give this answer: Popular CRM tools for startups include HubSpot, Pipedrive, and Salesforce. In this conversation: - HubSpot is the analyzed brand. - Pipedrive and Salesforce are the competitors that Genezio finds. Genezio uses this data in the visibility metrics and in the competitive analysis.","keywords":"HubSpot What CRM should a startup use?"},{"kind":"section","title":"Brands","heading":"Next steps","section":"Documentation","crumbs":["Core concepts","Brands","Next steps"],"url":"/docs/core-concepts/brands#next-steps","text":"To know how Genezio makes the conversations that measure the visibility of a brand, refer to: - Scenarios - Conversations","keywords":""},{"kind":"section","title":"Brands","heading":"Brands in the public API","section":"Documentation","crumbs":["Core concepts","Brands","Brands in the public API"],"url":"/docs/core-concepts/brands#brands-in-the-public-api","text":"To read and change brands, refer to List brands.","keywords":""},{"kind":"page","title":"Brand recommendation","heading":"","section":"Documentation","crumbs":["Documentation","Core concepts","Brand recommendation"],"url":"/docs/core-concepts/brand-recommendation","text":"Brand Recommendation measures how frequently answer engines recommend your brand when they mention it. Learn the formula, a worked example, and how to read it. Brand Recommendation is the Genezio metric that measures how frequently answer engines recommend your brand when users ask for a solution. It is the percentage of the conversations that show your brand in which the answer engine also recommends your brand. The metric answers the most important question for a marketing team: When users ask an answer engine to recommend a product or a service in your category, how frequently does the answer engine recommend your brand?.","keywords":""},{"kind":"section","title":"Brand recommendation","heading":"What Brand Recommendation represents","section":"Documentation","crumbs":["Core concepts","Brand recommendation","What Brand Recommendation represents"],"url":"/docs/core-concepts/brand-recommendation#what-brand-recommendation-represents","text":"Genezio runs conversations that simulate users who look for recommendations. In these conversations, the answer engine recommends your brand or does not recommend it. Examples of a recommendation: - The answer engine names your brand as a top choice. - The answer engine puts your brand in its list of recommended solutions. - The answer engine tells the user that your brand is a good fit for the specific needs of the user. If the answer recommends the brand, the conversation counts as a recommendation. If the answer does not recommend the brand, the conversation does not count. This is also true if the answer only mentions the brand.","keywords":""},{"kind":"section","title":"Brand recommendation","heading":"How Genezio calculates Brand Recommendation","section":"Documentation","crumbs":["Core concepts","Brand recommendation","How Genezio calculates Brand Recommendation"],"url":"/docs/core-concepts/brand-recommendation#how-genezio-calculates-brand-recommendation","text":"Genezio uses this formula: The denominator is not all the conversations. It is only the conversations in which your brand shows. Thus, Brand Recommendation measures your conversion rate from visibility to recommendation.","keywords":"Brand Recommendation = (Conversations where the brand is recommended / Conversations where the brand is visible) x 100"},{"kind":"section","title":"Brand recommendation","heading":"Worked example","section":"Documentation","crumbs":["Core concepts","Brand recommendation","Worked example"],"url":"/docs/core-concepts/brand-recommendation#worked-example","text":"Genezio runs 200 conversations for your topics: - Your brand shows in 80 of them. AI Visibility = 80 / 200 = 40%. - In 48 of these 80 visible conversations, the answer engine recommends your brand. AI Recommendation = 48 / 80 = 60%. This means: - Your brand shows in 40% of all the conversations (Visibility). - When your brand shows, the answer engine recommends it 60% of the time (Recommendation). - In the other 40% of the visible conversations, the answer engine mentions your brand but does not recommend it. The denominator is the visible conversations, not all the conversations. Thus, Recommendation % measures your conversion rate from a mention to the recommended choice.","keywords":""},{"kind":"section","title":"Brand recommendation","heading":"Which conversations Genezio uses","section":"Documentation","crumbs":["Core concepts","Brand recommendation","Which conversations Genezio uses"],"url":"/docs/core-concepts/brand-recommendation#which-conversations-genezio-uses","text":"Brand Recommendation is a metric at the brand level. Genezio calculates it from the conversations in which the answer engine selects the brands to recommend. The answer engine makes this decision from the needs of the user. Genezio uses the conversations of the Recommender Agent for this metric. These conversations start from a scenario and have many steps. In them, a persona asks the answer engine for a recommendation that agrees with the specific situation of the persona. Example: User query: Recommend CRM tools for early-stage startups with a small sales team. The answer engine replies with its own choices. If your brand is in the choices, the conversation is a recommendation. If your brand is not in the choices, you lost that opportunity.","keywords":""},{"kind":"section","title":"Brand recommendation","heading":"The other agent types","section":"Documentation","crumbs":["Core concepts","Brand recommendation","The other agent types"],"url":"/docs/core-concepts/brand-recommendation#the-other-agent-types","text":"- Prompter Agent: Its conversations count for AI Visibility (the answer mentions the brand). The Prompter Agent does not ask for a recommendation. Thus, Genezio does not use it for AI Recommendations. - Introspector Agent: Genezio uses it for the analysis of the narrative and of the accuracy of claims. The question contains the name of your brand. Thus, the answer does not show a real recommendation decision. - Comparer Agent: This agent is a competitive analysis tool, not a measurement engine. It uses the recommendation data and the visibility data of your brand. With this data, it shows your performance in relation to specific competitors. It has no effect on the calculation of AI Recommendations or of AI Visibility.","keywords":""},{"kind":"section","title":"Brand recommendation","heading":"Why Brand Recommendation is important","section":"Documentation","crumbs":["Core concepts","Brand recommendation","Why Brand Recommendation is important"],"url":"/docs/core-concepts/brand-recommendation#why-brand-recommendation-is-important","text":"Of all the metrics, Brand Recommendation is the nearest to a buying decision in AI search. A user asks an answer engine \"What should I use?\". If the answer engine recommends your brand, it is the same as when a trusted advisor sends a buyer to you. A higher recommendation score means: - Answer engines trust your brand as a good match for the needs of users. - Answer engines recommend your brand before your competitors. - Answer engines send to you the users who use them for purchase decisions. A lower score means that answer engines recommend your competitors. You lose deals before the buyer goes to your website.","keywords":""},{"kind":"section","title":"Brand recommendation","heading":"Brand Recommendation and Brand Visibility compared","section":"Documentation","crumbs":["Core concepts","Brand recommendation","Brand Recommendation and Brand Visibility compared"],"url":"/docs/core-concepts/brand-recommendation#brand-recommendation-and-brand-visibility-compared","text":"These two metrics work together, but they measure different things: What it measures: Of the conversations in which you are visible, how frequently the answer engine recommends you How frequently your brand shows in the answers. Denominator: The conversations in which the brand shows All the eligible conversations. Which conversations: Recommender Prompter + Recommender. Why it is important: Your conversion rate from presence to purchase intent It measures the general presence and awareness of the brand. Business analogy: Conversion rate: of the people who see you, how many select you Brand awareness: how many people know about you A brand can have a high visibility (many mentions) and a low recommendation rate (few recommendations). This difference is your priority. It means that answer engines know about you, but they do not trust you enough to recommend you.","keywords":""},{"kind":"section","title":"Brand recommendation","heading":"Brand Recommendation over time","section":"Documentation","crumbs":["Core concepts","Brand recommendation","Brand Recommendation over time"],"url":"/docs/core-concepts/brand-recommendation#brand-recommendation-over-time","text":"Genezio monitors Brand Recommendation continuously when it runs new conversations. Thus, teams can see trends such as: - More recommendations after you publish better comparison content - Competitors that get or lose a share of the recommendations - Changes in how answer engines compare your brand with the alternatives When you monitor the recommendations over time, you see the direct effect of your marketing work on the purchase decisions that answer engines influence.","keywords":""},{"kind":"section","title":"Brand recommendation","heading":"Next steps","section":"Documentation","crumbs":["Core concepts","Brand recommendation","Next steps"],"url":"/docs/core-concepts/brand-recommendation#next-steps","text":"- For the wider metric of how frequently your brand shows in the answers, refer to Brand Visibility. - To compare the recommendation rates of your competitors, refer to Share of Voice. - For all the metrics in one place, refer to Your KPIs explained.","keywords":""},{"kind":"section","title":"Brand recommendation","heading":"Brand Recommendation in the public API","section":"Documentation","crumbs":["Core concepts","Brand recommendation","Brand Recommendation in the public API"],"url":"/docs/core-concepts/brand-recommendation#brand-recommendation-in-the-public-api","text":"To read the recommendation data, refer to Recommendation over time and Recommendation by topic.","keywords":""},{"kind":"page","title":"Brand visibility","heading":"","section":"Documentation","crumbs":["Documentation","Core concepts","Brand visibility"],"url":"/docs/core-concepts/brand-visibility","text":"Brand Visibility is the percentage of AI conversations that mention your brand. Learn the formula, which agent types count, and how the accuracy range works. Brand Visibility is the Genezio metric that measures how frequently your brand shows in the answers of answer engines. It is the percentage of the eligible conversations that mention your brand. The metric answers a simple question: When users ask answer engines questions about your category, how frequently does your brand show?.","keywords":""},{"kind":"section","title":"Brand visibility","heading":"What Brand Visibility represents","section":"Documentation","crumbs":["Core concepts","Brand visibility","What Brand Visibility represents"],"url":"/docs/core-concepts/brand-visibility#what-brand-visibility-represents","text":"Genezio runs conversations with answer engines. The answers can include your brand, or not. Examples: - The answer recommends your brand. - Your brand shows in a comparison. - The answer mentions your brand together with competitors. If the brand shows at any location in the answer, the conversation counts as a visible occurrence. If the brand does not show, the conversation counts as not visible.","keywords":""},{"kind":"section","title":"Brand visibility","heading":"How Genezio calculates Brand Visibility","section":"Documentation","crumbs":["Core concepts","Brand visibility","How Genezio calculates Brand Visibility"],"url":"/docs/core-concepts/brand-visibility#how-genezio-calculates-brand-visibility","text":"Genezio uses this formula: Example: - Total of the analyzed conversations: 100 - Conversations that mention the brand: 78 Brand Visibility: This means that the brand shows in 78% of the answers for the analyzed topics.","keywords":"Brand Visibility = (Conversations where the brand appears / Total eligible conversations) x 100 78%"},{"kind":"section","title":"Brand visibility","heading":"Which conversations count for Brand Visibility","section":"Documentation","crumbs":["Core concepts","Brand visibility","Which conversations count for Brand Visibility"],"url":"/docs/core-concepts/brand-visibility#which-conversations-count-for-brand-visibility","text":"Not all the Genezio Agent types count for Brand Visibility. Genezio uses only the conversations in which the answer engine selects the brands without help. The prompt does not push the answer engine to a brand. Thus, Genezio uses only these Genezio Agent types.","keywords":""},{"kind":"section","title":"Brand visibility","heading":"Prompter Agent","section":"Documentation","crumbs":["Core concepts","Brand visibility","Prompter Agent"],"url":"/docs/core-concepts/brand-visibility#prompter-agent","text":"The conversations of the Prompter Agent ask a direct discovery question. Example: User query: What CRM should a startup use? The answer engine selects the brands that it mentions.","keywords":""},{"kind":"section","title":"Brand visibility","heading":"Recommender Agent","section":"Documentation","crumbs":["Core concepts","Brand visibility","Recommender Agent"],"url":"/docs/core-concepts/brand-visibility#recommender-agent","text":"The conversations of the Recommender Agent simulate a user who asks for recommendations. The answer engine selects the brands that it recommends from the scenario.","keywords":""},{"kind":"section","title":"Brand visibility","heading":"Excluded agent types","section":"Documentation","crumbs":["Core concepts","Brand visibility","Excluded agent types"],"url":"/docs/core-concepts/brand-visibility#excluded-agent-types","text":"Genezio does not use some agent types for Brand Visibility, because the prompt contains the brand name. These agent types are: Introspector Agent Example: User query: What is HubSpot used for? Comparer Agent Example: User query: HubSpot vs Salesforce If Genezio used these conversations, the visibility would be incorrectly high, because the prompt already contains the brand. The Comparer is a competitive analysis tool. It uses the KPIs of your brand, but it has no effect on the calculation of AI Recommendations or of AI Visibility. For more data, refer to Brand Recommendation.","keywords":""},{"kind":"section","title":"Brand visibility","heading":"Accuracy and sample size","section":"Documentation","crumbs":["Core concepts","Brand visibility","Accuracy and sample size"],"url":"/docs/core-concepts/brand-visibility#accuracy-and-sample-size","text":"Genezio calculates Brand Visibility with an accuracy model. Genezio does not use a fixed number of conversations. It analyzes the most recent conversations until the estimate gets to a reliable accuracy threshold. By default, Genezio uses the latest conversations until the estimate gets to an accuracy level of 90%. Thus, the visibility percentage is reliable.","keywords":""},{"kind":"section","title":"Brand visibility","heading":"Accuracy range","section":"Documentation","crumbs":["Core concepts","Brand visibility","Accuracy range"],"url":"/docs/core-concepts/brand-visibility#accuracy-range","text":"Genezio estimates Brand Visibility from a sample of conversations. Thus, the result has an accuracy range. For example: From the analyzed conversations, Genezio is 90% sure that the true visibility of the brand is in that range. When Genezio runs more conversations, the range usually becomes narrower, and the estimate becomes more precise.","keywords":"Brand Presence: 78% Accuracy range: 54% - 100% Accuracy level: 90%"},{"kind":"section","title":"Brand visibility","heading":"Why Brand Visibility is important","section":"Documentation","crumbs":["Core concepts","Brand visibility","Why Brand Visibility is important"],"url":"/docs/core-concepts/brand-visibility#why-brand-visibility-is-important","text":"Brand Visibility gives a high-level signal of how answer engines see a brand in a category. A higher visibility score usually means: - Answer engines frequently mention the brand. - The brand shows in comparison lists. - Answer engines strongly connect the brand with the topic. A lower score can show that competitors are the most frequent brands in the answers. Visibility counts mentions. To know how frequently a mention becomes a recommendation, refer to Brand Recommendation.","keywords":""},{"kind":"section","title":"Brand visibility","heading":"Brand Visibility over time","section":"Documentation","crumbs":["Core concepts","Brand visibility","Brand Visibility over time"],"url":"/docs/core-concepts/brand-visibility#brand-visibility-over-time","text":"Genezio monitors Brand Visibility continuously when it runs new conversations. Thus, teams can see trends such as: - More visibility after you publish new content - Competitors that get or lose presence - Changes in how answer engines interpret a category When you monitor the visibility over time, you see how your presence changes across answer engines.","keywords":""},{"kind":"section","title":"Brand visibility","heading":"Next steps","section":"Documentation","crumbs":["Core concepts","Brand visibility","Next steps"],"url":"/docs/core-concepts/brand-visibility#next-steps","text":"- To compare the visibility of your brand with the visibility of competitors in the same topic, refer to Share of Voice. - For the full method, refer to How Genezio measures visibility.","keywords":""},{"kind":"section","title":"Brand visibility","heading":"Brand Visibility in the public API","section":"Documentation","crumbs":["Core concepts","Brand visibility","Brand Visibility in the public API"],"url":"/docs/core-concepts/brand-visibility#brand-visibility-in-the-public-api","text":"To read the visibility data, refer to Visibility over time and Visibility by topic.","keywords":""},{"kind":"page","title":"Users","heading":"","section":"Documentation","crumbs":["Documentation","Core concepts","Users"],"url":"/docs/core-concepts/users","text":"Users are the people who can access a Genezio account and work in it. Learn the difference between account-level and brand-level users and when to use each. In Genezio, users are the persons who can access an account and work in it. You can give a user access to all the brands of the account or only to selected brands. You can invite users at two levels: - Account level: The user can access all brands in the account. - Brand level: The user can access only the selected brands. Thus, teams control the collaboration and the visibility across many brands.","keywords":""},{"kind":"section","title":"Users","heading":"Account-level users","section":"Documentation","crumbs":["Core concepts","Users","Account-level users"],"url":"/docs/core-concepts/users#account-level-users","text":"You invite account-level users to the full workspace of the account. This is useful for team members who must see much data, for example: - Account owners - Operations leads - Central growth teams or analytics teams","keywords":""},{"kind":"section","title":"Users","heading":"Brand-level users","section":"Documentation","crumbs":["Core concepts","Users","Brand-level users"],"url":"/docs/core-concepts/users#brand-level-users","text":"You invite brand-level users to one or more specific brands. This is useful when you must limit the access, for example: - Client stakeholders who must see only their brand - Agency specialists who work on selected client brands - Internal teams that are responsible for one product line","keywords":""},{"kind":"section","title":"Users","heading":"Why the user access level is important","section":"Documentation","crumbs":["Core concepts","Users","Why the user access level is important"],"url":"/docs/core-concepts/users#why-the-user-access-level-is-important","text":"The design of user access has an effect on the governance and on the daily work. A clear assignment of users helps teams to do these actions: - Prevent the accidental exposure of data across brands - Keep the ownership and the accountability clear - Collaborate faster in the correct scope","keywords":""},{"kind":"section","title":"Users","heading":"Next steps","section":"Documentation","crumbs":["Core concepts","Users","Next steps"],"url":"/docs/core-concepts/users#next-steps","text":"- Brands - Personas - Sign-in and SSO options - Enterprise SSO and SCIM","keywords":""},{"kind":"page","title":"Personas","heading":"","section":"Documentation","crumbs":["Documentation","Core concepts","Personas"],"url":"/docs/core-concepts/personas","text":"A persona is the type of user who asks answer engines about a brand. Learn how the language and the location of a persona shape realistic Genezio conversations. In Genezio, a persona is the type of user who can ask an answer engine questions about a brand. A persona defines who asks the questions, where the user is, and which language the user speaks, so that Genezio can simulate realistic conversations. Each persona is part of a specific brand, and most topics have a persona. Thus, Genezio makes the conversations from the point of view of the real customers of the brand. For Introspector topics, the persona is optional. Refer to Introspector agent.","keywords":""},{"kind":"section","title":"Personas","heading":"What a persona represents","section":"Documentation","crumbs":["Core concepts","Personas","What a persona represents"],"url":"/docs/core-concepts/personas#what-a-persona-represents","text":"A persona describes a typical user or buyer of the analyzed brand. For example, a persona can be: - A startup founder who looks for CRM software - A marketing manager who looks for automation tools - A runner who examines shoes for marathon training Genezio models the user behind the question. Thus, the conversations are more similar to the real interactions with answer engines.","keywords":""},{"kind":"section","title":"Personas","heading":"The attributes of a persona","section":"Documentation","crumbs":["Core concepts","Personas","The attributes of a persona"],"url":"/docs/core-concepts/personas#the-attributes-of-a-persona","text":"Each persona has attributes that change how Genezio makes the conversations.","keywords":""},{"kind":"section","title":"Personas","heading":"Brand","section":"Documentation","crumbs":["Core concepts","Personas","Brand"],"url":"/docs/core-concepts/personas#brand","text":"Each persona is part of a specific brand. Thus, the persona is a typical customer or user of the products or the services of that brand. Example: - Brand: HubSpot - Persona: Startup founder who manages sales and marketing","keywords":""},{"kind":"section","title":"Personas","heading":"Language","section":"Documentation","crumbs":["Core concepts","Personas","Language"],"url":"/docs/core-concepts/personas#language","text":"Each persona speaks a specific language. Genezio uses this language when it runs conversations with answer engines. Example: - English - Spanish - French Genezio writes the prompts, gets the answers, and does the analysis in that language.","keywords":""},{"kind":"section","title":"Personas","heading":"Location","section":"Documentation","crumbs":["Core concepts","Personas","Location"],"url":"/docs/core-concepts/personas#location","text":"Each persona also has a country and a city. Genezio runs the conversations as if the user is in that specific location. This makes the conditions realistic, because answer engines can change their answers for a geographic location. For example: - Country: United States - City: San Francisco Or: - Country: Germany - City: Berlin The location can change the recommendations, the available products, the regulatory context, or the regional preferences.","keywords":""},{"kind":"section","title":"Personas","heading":"Genezio acts as the persona in conversations","section":"Documentation","crumbs":["Core concepts","Personas","Genezio acts as the persona in conversations"],"url":"/docs/core-concepts/personas#genezio-acts-as-the-persona-in-conversations","text":"When Genezio talks to an answer engine, it acts as the defined persona. Genezio writes the prompts and the follow-up questions as if it is that user. Genezio speaks to the answer engine as the persona. For example, the persona is: - A startup founder - In San Francisco - A speaker of English Genezio writes the conversation with the answer engine from that point of view. The questions, the tone, and the context agree with how that specific user asks for recommendations or for information. Thus, the answers of the answer engines show realistic user behavior, not the behavior of generic test prompts.","keywords":""},{"kind":"section","title":"Personas","heading":"How Genezio uses personas","section":"Documentation","crumbs":["Core concepts","Personas","How Genezio uses personas"],"url":"/docs/core-concepts/personas#how-genezio-uses-personas","text":"Personas change how Genezio runs topics and scenarios in conversations. The process is: 1. A brand defines one or more personas. 2. Each topic has a persona. For Introspector topics, the persona is optional. 3. Genezio makes scenarios from that topic and that persona. 4. Genezio runs conversations with the answer engine and acts as the persona. 5. The conversation uses the language and the geographic context of the persona. Thus, the conversations show how real users from specific markets can interact with answer engines.","keywords":""},{"kind":"section","title":"Personas","heading":"Example: one persona and one topic","section":"Documentation","crumbs":["Core concepts","Personas","Example: one persona and one topic"],"url":"/docs/core-concepts/personas#example-one-persona-and-one-topic","text":"A brand sells CRM software for startups. A persona can be: - Persona: Early-stage startup founder - Language: English - Country: United States - City: San Francisco A topic for this persona can be: When Genezio runs conversations for this topic, the answer engine gets prompts that simulate a user in San Francisco. The user asks questions in English about CRM solutions.","keywords":"CRM tools for startups"},{"kind":"section","title":"Personas","heading":"Example: different personas for the same brand","section":"Documentation","crumbs":["Core concepts","Personas","Example: different personas for the same brand"],"url":"/docs/core-concepts/personas#example-different-personas-for-the-same-brand","text":"Personas are very important when the same brand serves different types of users. For example, a bank offers debit cards. Two personas can interact with answer engines in very different ways. Persona 1: Teenager (15 years old) - Goal: get a debit card that is connected to the account of the parents - Language: informal, simple - Possible questions: Persona 2: Adult (40 years old) - Goal: open a new personal bank account - Language: more direct and financial - Possible questions: The two personas can interact with the same brand and similar products. But they ask questions in different ways, and thus the answer engines can give very different answers. When you model personas, Genezio can find these differences. It simulates how answer engines reply to different types of users.","keywords":"Can I get a debit card if I'm 15? Is there a card for teens connected to my parents account? Which banks offer debit cards for teenagers? What bank offers the best checking account? Which banks have good debit cards and mobile apps? What are the fees for opening a new bank account?"},{"kind":"section","title":"Personas","heading":"Create a persona from a document","section":"Documentation","crumbs":["Core concepts","Personas","Create a persona from a document"],"url":"/docs/core-concepts/personas#create-a-persona-from-a-document","text":"You possibly have persona research in a document. Examples are a buyer-persona document, an internal customer-research deck, a job description, or other text that describes your customers. You can upload this document, and Genezio makes a persona from it automatically. This is the fastest procedure to get a usable persona. It is the recommended start for teams that: - Keep persona documents in a research repository or in marketing collateral - Have customer-research output from a CRM, a customer success tool, or a sales discovery process - Want to start without a form of many fields","keywords":""},{"kind":"section","title":"Personas","heading":"How persona import works","section":"Documentation","crumbs":["Core concepts","Personas","How persona import works"],"url":"/docs/core-concepts/personas#how-persona-import-works","text":"Upload the document on the persona creation screen. Genezio reads the document and makes a persona from its content. The persona has a name, a role, a country, a city, a language, and supporting context. Genezio saves the persona immediately as a real persona of the brand. You can edit it after, as any other persona. Each upload makes one persona. If your document describes many different personas, upload it one time for each persona, or divide the document. For the procedure, refer to Generating personas from documents.","keywords":""},{"kind":"section","title":"Personas","heading":"Why persona import is important","section":"Documentation","crumbs":["Core concepts","Personas","Why persona import is important"],"url":"/docs/core-concepts/personas#why-persona-import-is-important","text":"The largest problem when you start with the platform is to make good personas. A weak persona gives weak conversations, and weak conversations give weak insights. A real document already shows how your team thinks about a customer segment. When you start from it, you get a useful persona in one step. Your team does not have to write a persona in an empty form. For customers who already have persona research, this is an import in one step.","keywords":""},{"kind":"section","title":"Personas","heading":"Manage personas","section":"Documentation","crumbs":["Core concepts","Personas","Manage personas"],"url":"/docs/core-concepts/personas#manage-personas","text":"Genezio helps you keep personas organized and portable: - Export personas: You can export all your personas at one time, or export one specific persona. This is useful when you share persona definitions, or when you keep them together with other research. - Summary field: Each persona can have a summary. The summary shows quickly who the persona is and what is important to the persona. - Consistent country list: The persona creation uses a consistent country list. Thus, the locations are the same across all personas. Genezio then makes and uses personas in the language and the location of each persona. Thus, the conversations stay in the correct market.","keywords":""},{"kind":"section","title":"Personas","heading":"Filter reports by persona","section":"Documentation","crumbs":["Core concepts","Personas","Filter reports by persona"],"url":"/docs/core-concepts/personas#filter-reports-by-persona","text":"When the topics of a brand use two or more personas, a persona filter shows in the header of the report. When you select a persona, the report shows only the data of that persona. Thus, you can see how the answers are different for a startup founder and for an enterprise buyer. The filter stays the same across screens. It is next to the Master Filters list. If you have only one persona, there is nothing to compare, and the filter does not show.","keywords":""},{"kind":"section","title":"Personas","heading":"Why personas are important","section":"Documentation","crumbs":["Core concepts","Personas","Why personas are important"],"url":"/docs/core-concepts/personas#why-personas-are-important","text":"Personas let Genezio simulate realistic discovery situations. Different users ask different questions. The questions change with: - The role of the user - The goals of the user - The market of the user - The language of the user When Genezio models these factors, it shows better how answer engines reply to real users.","keywords":""},{"kind":"section","title":"Personas","heading":"Next steps","section":"Documentation","crumbs":["Core concepts","Personas","Next steps"],"url":"/docs/core-concepts/personas#next-steps","text":"To know how Genezio uses personas to make the conversations that measure visibility, refer to: - Scenarios - Conversations","keywords":""},{"kind":"section","title":"Personas","heading":"Personas in the public API","section":"Documentation","crumbs":["Core concepts","Personas","Personas in the public API"],"url":"/docs/core-concepts/personas#personas-in-the-public-api","text":"To read and change personas, refer to List personas.","keywords":""},{"kind":"page","title":"Topics","heading":"","section":"Documentation","crumbs":["Documentation","Core concepts","Topics"],"url":"/docs/core-concepts/topics","text":"A topic is a subject area in which users ask answer engines about a brand or a category. Learn how topics, scenarios, and the Genezio agent types work together. In Genezio, a topic is a subject area in which users can ask answer engines questions about a brand or a product category. Topics define the scope of the analysis: they set where Genezio measures the AI Recommendations and the Visibility of a brand. Topics set the questions that Genezio asks answer engines. Most topics have a persona. Genezio runs the conversations of the topic from the point of view of that persona, in the language and the location of the persona. Introspector topics are different: for them, a persona is optional. Refer to Introspector Agent.","keywords":""},{"kind":"section","title":"Topics","heading":"What a topic represents","section":"Documentation","crumbs":["Core concepts","Topics","What a topic represents"],"url":"/docs/core-concepts/topics#what-a-topic-represents","text":"A topic is a problem space or a category in which users look for solutions. Examples: - CRM for startups - Running shoes for marathon training - Debit cards for teenagers - Project management tools for small teams For each topic, Genezio makes conversations with answer engines to find: - Which brands the answers mention - Which sources the answers cite - How the answers describe the brands - Which competitors show","keywords":""},{"kind":"section","title":"Topics","heading":"Topics, scenarios, and conversations","section":"Documentation","crumbs":["Core concepts","Topics","Topics, scenarios, and conversations"],"url":"/docs/core-concepts/topics#topics-scenarios-and-conversations","text":"Each topic has one or more scenarios. A scenario is a realistic situation in which the persona has a goal that is related to the topic. Scenarios have more detail than topics. A topic defines the subject area. A scenario describes: - The situation of the persona - The goal of the persona - The context behind the question When Genezio runs the analysis, each conversation starts from a scenario. The structure is: 1. A topic defines the subject area (for example: \"CRM for startups\"). 2. A scenario describes the situation and the goal of the persona. 3. Genezio starts a conversation with an answer engine from that scenario. 4. The conversation has one or more prompts. A prompt is one message that Genezio sends to the answer engine. For example: Topic: Scenario: From that scenario, Genezio can make prompts such as: User query: What CRM would you recommend for a small SaaS startup with only a few sales leads per week? Or follow-up prompts such as: User query: Which of these tools is easiest to set up for a small team? The scenario defines the situation. The prompts are the individual messages that Genezio sends to the answer engine in the conversation.","keywords":"CRM for startups The persona is a founder of a small SaaS startup with two co-founders. They recently started getting inbound leads and need a simple way to track contacts and follow up with potential customers."},{"kind":"section","title":"Topics","heading":"Genezio Agent types","section":"Documentation","crumbs":["Core concepts","Topics","Genezio Agent types"],"url":"/docs/core-concepts/topics#genezio-agent-types","text":"Genezio has many Genezio Agent types. Each type is a different intent of the user. For a summary of all the types, refer to How the 5 agents work.","keywords":""},{"kind":"section","title":"Topics","heading":"Prompter Agent","section":"Documentation","crumbs":["Core concepts","Topics","Prompter Agent"],"url":"/docs/core-concepts/topics#prompter-agent","text":"A Prompter Agent sends one prompt to the answer engine. For this agent type, you write the prompt directly. Genezio sends it word for word, exactly as you wrote it. The interaction is single-turn. There are no follow-up prompts. Example: User query: What CRM should a startup use? The conversations of the Prompter Agent are the most simple and direct discovery questions that users ask answer engines.","keywords":""},{"kind":"section","title":"Topics","heading":"Recommender Agent","section":"Documentation","crumbs":["Core concepts","Topics","Recommender Agent"],"url":"/docs/core-concepts/topics#recommender-agent","text":"A Recommender Agent simulates a user who asks the answer engine to recommend solutions. The Recommender Agent is different from the Prompter Agent. It uses conversations of many steps. The scenario defines the situation and the goal. Genezio makes many prompts in the conversation to simulate how a real user interacts with the answer engine. Example scenario: From that scenario, Genezio can make prompts that: - Ask for tool recommendations - Ask about prices - Ask which tool is the easiest to use - Ask about alternatives These interactions simulate realistic recommendation conversations.","keywords":"George runs a small startup and is evaluating tools to manage incoming sales leads."},{"kind":"section","title":"Topics","heading":"Introspector Agent","section":"Documentation","crumbs":["Core concepts","Topics","Introspector Agent"],"url":"/docs/core-concepts/topics#introspector-agent","text":"An Introspector Agent examines a specific brand. The scenario puts the persona in a situation in which the persona wants to learn about a specific product or company. Example scenario: Genezio then makes prompts that: - Ask what the product does - Ask for which users the product is designed - Ask about the advantages and the limits of the product These topics help you analyze how answer engines describe and explain a brand. The scenario contains the brand name. Thus, Genezio does not use the conversations of the Introspector Agent to calculate AI Visibility.","keywords":"Mary has heard about HubSpot but is not sure what it does or whether it would fit their startup."},{"kind":"section","title":"Topics","heading":"Comparer Agent","section":"Documentation","crumbs":["Core concepts","Topics","Comparer Agent"],"url":"/docs/core-concepts/topics#comparer-agent","text":"A Comparer Agent examines how answer engines compare competitor brands. The scenario puts the persona in a situation in which the persona compares alternatives. Example scenario: From this scenario, Genezio makes prompts that ask the answer engine to: - Compare the two products - Show the strong points and the weak points - Recommend the better option for the situation of the persona These conversations help you analyze the competitive position in the answers. The scenario contains the brands. Thus, Genezio does not use the conversations of the Comparer Agent to calculate AI Visibility or AI Recommendations. The Comparer is a competitive analysis tool. It uses the brand-level KPIs that Genezio already calculated from the Prompter and Recommender conversations. With these KPIs, it shows how your brand compares with specific competitors.","keywords":"I am deciding between HubSpot and Pipedrive for managing startup sales."},{"kind":"section","title":"Topics","heading":"Fact Checker Agent","section":"Documentation","crumbs":["Core concepts","Topics","Fact Checker Agent"],"url":"/docs/core-concepts/topics#fact-checker-agent","text":"A Fact Checker Agent tests if answer engines confirm specific claims about your brand. The claims come from your Knowledge Base, or you type them directly. The prompt names the brand and the claims. Thus, Genezio does not use the conversations of the Fact Checker Agent to calculate AI Visibility or AI Recommendations.","keywords":""},{"kind":"section","title":"Topics","heading":"Single-turn and multi-turn agent types","section":"Documentation","crumbs":["Core concepts","Topics","Single-turn and multi-turn agent types"],"url":"/docs/core-concepts/topics#single-turn-and-multi-turn-agent-types","text":"The Genezio Agent types run conversations in different ways: - Prompter Agent: One prompt, sent exactly as written - Recommender Agent: A conversation of many steps, made from a scenario - Introspector Agent: A conversation of many steps that examines a brand - Comparer Agent: A conversation of many steps that compares brands With this combination, Genezio simulates simple discovery queries and realistic research in a conversation.","keywords":""},{"kind":"section","title":"Topics","heading":"How Genezio creates topics","section":"Documentation","crumbs":["Core concepts","Topics","How Genezio creates topics"],"url":"/docs/core-concepts/topics#how-genezio-creates-topics","text":"During onboarding, Genezio automatically makes a first set of topics from: - The brand description - The customer persona - The product category Then, users can: - Edit topics - Add more topics - Remove topics that are not relevant Over time, teams frequently add topics for more markets, products, or user intents.","keywords":""},{"kind":"section","title":"Topics","heading":"Generate more topics","section":"Documentation","crumbs":["Core concepts","Topics","Generate more topics"],"url":"/docs/core-concepts/topics#generate-more-topics","text":"You do not have to think of each topic yourself. Generate topics tells Genezio to suggest new topics for you. When you generate topics, you: - Select the persona for the topics. Thus, the suggestions show how that audience asks questions. - Optionally, add a hint. A hint points Genezio to a market, a product line, or an angle that you want. This is the fastest procedure to fill gaps. Select a persona that you do not serve enough, give a hint about the area that is important to you, and examine the suggestions. You can still edit or discard each suggestion that does not fit.","keywords":""},{"kind":"section","title":"Topics","heading":"Select answer engines for each topic","section":"Documentation","crumbs":["Core concepts","Topics","Select answer engines for each topic"],"url":"/docs/core-concepts/topics#select-answer-engines-for-each-topic","text":"In Settings, when you create or edit a topic, you can select the specific answer engines for that topic. If you do not, the topic uses the default answer engines of the brand. You select the answer engines for each topic. Thus, different topics can use different sets of answer engines. For example, most of your topics can run on ChatGPT and five other answer engines, and one topic runs on ChatGPT only. This is useful when a topic is important only on some answer engines. For example, a regional topic or an experimental topic does not need all your answer engines. Refer to Selecting Answer Engines.","keywords":""},{"kind":"section","title":"Topics","heading":"Why topics are important","section":"Documentation","crumbs":["Core concepts","Topics","Why topics are important"],"url":"/docs/core-concepts/topics#why-topics-are-important","text":"Topics define where Genezio measures the AI Recommendations and the Visibility of your brand. A brand can have a strong visibility in some topics and a low visibility in other topics. For example: - Strong visibility for \"CRM for startups\" - Weaker visibility for \"sales automation platforms\" When you analyze each topic separately, you can see where answer engines show your brand well and where you have opportunities. To put topics into groups for your reports, use Topic Tags or Master Filters.","keywords":""},{"kind":"section","title":"Topics","heading":"Next steps","section":"Documentation","crumbs":["Core concepts","Topics","Next steps"],"url":"/docs/core-concepts/topics#next-steps","text":"To know how Genezio changes topics into structured interactions with answer engines, refer to: - Scenarios - Conversations","keywords":""},{"kind":"section","title":"Topics","heading":"Topics in the public API","section":"Documentation","crumbs":["Core concepts","Topics","Topics in the public API"],"url":"/docs/core-concepts/topics#topics-in-the-public-api","text":"To read and change topics, refer to List topics.","keywords":""},{"kind":"page","title":"Scenarios","heading":"","section":"Documentation","crumbs":["Documentation","Core concepts","Scenarios"],"url":"/docs/core-concepts/scenarios","text":"A scenario is the situation and the goal of a persona in Genezio. Learn how to write a good scenario and how to change, pause, or restore your scenarios. In Genezio, a scenario is a realistic situation in which a persona has a goal that is related to a specific topic. Each conversation that Genezio runs with an answer engine starts from a scenario. Topics define the general subject area. Scenarios describe the specific situation that causes a user to ask questions.","keywords":""},{"kind":"section","title":"Scenarios","heading":"Topics and scenarios","section":"Documentation","crumbs":["Core concepts","Scenarios","Topics and scenarios"],"url":"/docs/core-concepts/scenarios#topics-and-scenarios","text":"It is important to know the difference between topics and scenarios. A topic defines the analyzed subject area. Example topic: A scenario is a short story that describes a realistic situation. It gives the context of the persona, the limits of the persona, and the goal of the persona. It is similar to a brief that you give to a person to explain the full situation. Example scenario: A scenario is not a prompt. It is not a question that you send to an answer engine. Genezio reads the scenario and adds the details of the persona. Then it makes natural prompts from them. These prompts are similar to the messages that a real person types in ChatGPT or Claude, one at a time.","keywords":"CRM for startups Mary's team doesn't have any technical background. They are looking for a marketing automation platform that is easy for a non-technical team to set up and start using without developer help. Their budget is $800/month and she wants her team of 6 to be able to use it in parallel. She is looking for something that is SOC2 compliant."},{"kind":"section","title":"Scenarios","heading":"How Genezio uses scenarios","section":"Documentation","crumbs":["Core concepts","Scenarios","How Genezio uses scenarios"],"url":"/docs/core-concepts/scenarios#how-genezio-uses-scenarios","text":"Each conversation that Genezio runs starts with a scenario. The process is: 1. A persona defines who the user is (language, location, context). 2. A topic defines the subject area. 3. A scenario defines the situation and the goal of the persona. 4. Genezio makes prompts from that scenario. 5. Genezio sends the prompts to the answer engine in the conversation. With this structure, the conversations start from realistic user contexts, not from abstract prompts.","keywords":""},{"kind":"section","title":"Scenarios","heading":"The structure of a scenario","section":"Documentation","crumbs":["Core concepts","Scenarios","The structure of a scenario"],"url":"/docs/core-concepts/scenarios#the-structure-of-a-scenario","text":"A scenario describes the situation and the goal of the persona. Write it as a short story: one paragraph with specific details, limits, and context. A scenario does not include details about the persona (such as age, country, occupation, or language). The persona defines these details separately. The scenario can use the name of the persona. Genezio adds all the other persona details when it runs the conversation. Thus, these items of the persona are already defined before Genezio runs the scenario: - The role or the profile - The language - The geographic location The scenario describes only what occurs and what the persona needs.","keywords":""},{"kind":"section","title":"Scenarios","heading":"What to include","section":"Documentation","crumbs":["Core concepts","Scenarios","What to include"],"url":"/docs/core-concepts/scenarios#what-to-include","text":"A good scenario includes: - The situation: the conditions that caused this need (for example, the spreadsheets are too small for the team now, or a product launch is near) - The goal: what the persona wants to get (for example, find a tool, compare options, or solve a problem) - Specific limits: budget, team size, technical requirements, compliance needs, and the time limit","keywords":""},{"kind":"section","title":"Scenarios","heading":"Goal description and intent","section":"Documentation","crumbs":["Core concepts","Scenarios","Goal description and intent"],"url":"/docs/core-concepts/scenarios#goal-description-and-intent","text":"Each scenario also has a goal description and an intent, in addition to the story: - The goal description tells, in plain words, what the persona wants to get at the end. - The intent tells the type of question that the persona asks the answer engine. Examples are: find a tool, compare options, or examine a fact. These two fields give Genezio more context when it makes prompts. Thus, Genezio makes more realistic prompts. The questions agree better with what the persona wants and with how a real person writes them.","keywords":""},{"kind":"section","title":"Scenarios","heading":"What not to include","section":"Documentation","crumbs":["Core concepts","Scenarios","What not to include"],"url":"/docs/core-concepts/scenarios#what-not-to-include","text":"Do not repeat the persona details in the scenario. The persona already defines who the user is. The scenario tells why the persona asks and what the persona needs. With this separation, Genezio can use the same scenario again in different persona contexts. The conversations still start from specific goals. For a step-by-step procedure, refer to Creating scenarios.","keywords":""},{"kind":"section","title":"Scenarios","heading":"Example: from a scenario to a conversation","section":"Documentation","crumbs":["Core concepts","Scenarios","Example: from a scenario to a conversation"],"url":"/docs/core-concepts/scenarios#example-from-a-scenario-to-a-conversation","text":"This example shows how a scenario becomes a conversation. Topic: Marketing automation Persona: Marketing manager, located in New York, language: English Scenario: The scenario does not mention the role, the location, or the language of Mary. These details come from the persona. What Genezio makes from this scenario: Genezio adds the persona details to the scenario. Then it makes natural prompts, similar to the messages that a real person sends to an answer engine, one at a time: User query: I'm looking for a marketing automation platform that's easy to set up without any developer help. My team of 6 isn't very technical. What would you recommend? User query: Our budget is around $800/month and we need all 6 team members to be able to work in the platform at the same time. Which tools support that? User query: Does any of these have SOC2 compliance? That's a requirement for us. The scenario is the full story. The prompts show how a person tells that story in a conversation: one question at a time, and each question continues from the previous answers.","keywords":"Mary's team doesn't have any technical background. They are looking for a marketing automation platform that is easy for a non-technical team to set up and start using without developer help. Their budget is $800/month and she wants her team of 6 to be able to use it in parallel. She is looking for something that is SOC2 compliant."},{"kind":"section","title":"Scenarios","heading":"Scenarios and Genezio Agents","section":"Documentation","crumbs":["Core concepts","Scenarios","Scenarios and Genezio Agents"],"url":"/docs/core-concepts/scenarios#scenarios-and-genezio-agents","text":"Each Genezio Agent type uses scenarios in a different way.","keywords":""},{"kind":"section","title":"Scenarios","heading":"Prompter Agent","section":"Documentation","crumbs":["Core concepts","Scenarios","Prompter Agent"],"url":"/docs/core-concepts/scenarios#prompter-agent","text":"For Prompter Agent topics, you write a direct prompt, not a story. Genezio sends this prompt to the answer engine exactly as written. The interaction is single-turn.","keywords":""},{"kind":"section","title":"Scenarios","heading":"Recommender Agent","section":"Documentation","crumbs":["Core concepts","Scenarios","Recommender Agent"],"url":"/docs/core-concepts/scenarios#recommender-agent","text":"For Recommender Agent topics, the scenario defines the start situation. Genezio makes many prompts in the conversation to simulate a realistic recommendation flow.","keywords":""},{"kind":"section","title":"Scenarios","heading":"Introspector Agent","section":"Documentation","crumbs":["Core concepts","Scenarios","Introspector Agent"],"url":"/docs/core-concepts/scenarios#introspector-agent","text":"For Introspector Agent topics, the scenario asks directly about a specific brand: - If the topic has a persona, the scenario puts the persona in a situation in which the persona wants to know that brand. - If the topic has no persona, the scenario asks the answer engine to describe the brand for no specific audience. Introspector topics can have no persona. Genezio makes prompts that examine: - What the brand does - For which users the brand is designed - The strong points and the weak points of the brand","keywords":""},{"kind":"section","title":"Scenarios","heading":"Comparer Agent","section":"Documentation","crumbs":["Core concepts","Scenarios","Comparer Agent"],"url":"/docs/core-concepts/scenarios#comparer-agent","text":"For Comparer Agent topics, the scenario describes a situation in which the persona compares many brands. Genezio makes prompts that compare these alternatives and analyze their differences.","keywords":""},{"kind":"section","title":"Scenarios","heading":"Why scenarios are important","section":"Documentation","crumbs":["Core concepts","Scenarios","Why scenarios are important"],"url":"/docs/core-concepts/scenarios#why-scenarios-are-important","text":"With scenarios, Genezio simulates the behavior of real users. In the real world, people almost never ask general questions such as \"best CRM\". They ask questions that come from their situation. For example: - A founder who looks for a first CRM - A teenager who looks for a debit card - A runner who prepares for a marathon When Genezio models these situations, it can analyze how answer engines reply to realistic user needs and contexts. To find scenarios that do not make answer engines talk about brands, refer to Topic / Scenario Relevance.","keywords":""},{"kind":"section","title":"Scenarios","heading":"Change a scenario to a different type","section":"Documentation","crumbs":["Core concepts","Scenarios","Change a scenario to a different type"],"url":"/docs/core-concepts/scenarios#change-a-scenario-to-a-different-type","text":"A good scenario is valuable work. It takes effort to write a good situation with the correct limits and goal. Genezio lets you change a scenario from one type to a different type. Thus, you can use the same work for different intents, and you do not write the scenario again. For example, you can change a Prompter scenario to a Recommender scenario. You can change a scenario between all five canonical types: - Prompter - Recommender - Introspector - Comparer - Fact Checker When you change the type of a scenario, you can also select the persona of the new scenario. This is useful when the same situation is correct for a different audience. It is also useful when you want the new scenario to run from a different point of view. Thus, one good scenario can give many types of analysis. You can examine the same customer situation through recommendation, comparison, introspection, or fact checking, and you do not start again each time.","keywords":""},{"kind":"section","title":"Scenarios","heading":"Pause and resume a scenario","section":"Documentation","crumbs":["Core concepts","Scenarios","Pause and resume a scenario"],"url":"/docs/core-concepts/scenarios#pause-and-resume-a-scenario","text":"You can pause a scenario to stop it without a deletion. You can resume it later when you want the data again. A paused scenario keeps its text. Genezio does not run it in the daily runs. This is useful when you want to: - Keep a seasonal scenario or a campaign scenario until it is relevant again - Stop the runs of a scenario that gives you no useful data now - Make your analysis smaller for a short time, and keep the text and the history of the scenario Pause a scenario when you possibly want it back. Delete a scenario when you do not need it any more. Even then, you can restore it, as the next section shows.","keywords":""},{"kind":"section","title":"Scenarios","heading":"Restore deleted scenarios","section":"Documentation","crumbs":["Core concepts","Scenarios","Restore deleted scenarios"],"url":"/docs/core-concepts/scenarios#restore-deleted-scenarios","text":"When you delete a scenario or a full topic, the deletion is not permanent. Genezio moves the removed topics and scenarios to a Recycle Bin. Genezio keeps them there, and you can restore them if you deleted them by mistake. Thus, you can clean your workspace, and you do not lose a scenario that you wrote carefully. If you remove an item that you still need, restore it from the Recycle Bin. Then you can use it again as before.","keywords":""},{"kind":"section","title":"Scenarios","heading":"Next steps","section":"Documentation","crumbs":["Core concepts","Scenarios","Next steps"],"url":"/docs/core-concepts/scenarios#next-steps","text":"To see how Genezio runs scenarios with answer engines and analyzes the answers, refer to: - Conversations","keywords":""},{"kind":"section","title":"Scenarios","heading":"Scenarios in the public API","section":"Documentation","crumbs":["Core concepts","Scenarios","Scenarios in the public API"],"url":"/docs/core-concepts/scenarios#scenarios-in-the-public-api","text":"To read and change scenarios, refer to List scenarios.","keywords":""},{"kind":"page","title":"Topic and scenario relevance","heading":"","section":"Documentation","crumbs":["Documentation","Core concepts","Topic and scenario relevance"],"url":"/docs/core-concepts/topic-scenario-relevance","text":"Topic / Scenario Relevance is the percentage of conversations that mention your brand or a competitor. Learn how to read the score and make a scenario better. In Genezio, Topic / Scenario Relevance measures how well a scenario shows a real decision context in which answer engines talk about brands. The score is the percentage of conversations in which your brand or one of your competitors shows. Each scenario and each topic has a Relevance score. This score answers an important question: When Genezio runs this scenario, do answer engines really talk about brands in this category? If the answers frequently mention your brand or your competitors, the scenario has a high relevance. If the answers almost never mention brands, the scenario is possibly not well written. Or the scenario does not agree with how users really ask questions.","keywords":""},{"kind":"section","title":"Topic and scenario relevance","heading":"What Relevance represents","section":"Documentation","crumbs":["Core concepts","Topic and scenario relevance","What Relevance represents"],"url":"/docs/core-concepts/topic-scenario-relevance#what-relevance-represents","text":"Relevance is the percentage of the conversations in which your brand or one of your competitors shows. Genezio uses this formula: Example: - Conversations that Genezio ran: 50 - Conversations in which one or more brands show: 40 Relevance: This means that in 80% of the conversations from that scenario, the answer engine mentioned one or more relevant brands.","keywords":"Topic / Scenario Relevance = (Conversations mentioning your brand or competitors / Total conversations) x 100 80%"},{"kind":"section","title":"Topic and scenario relevance","heading":"Why Topic / Scenario Relevance is important","section":"Documentation","crumbs":["Core concepts","Topic and scenario relevance","Why Topic / Scenario Relevance is important"],"url":"/docs/core-concepts/topic-scenario-relevance#why-topic-scenario-relevance-is-important","text":"A scenario simulates a realistic situation in which a user compares products or services. If the answers do not mention any brands, the scenario possibly does not show a realistic discovery question. For example, a scenario that is not clear or not well written can cause the answer engine to give general advice with no product recommendations. In these cases, the scenario has little value for the competitive analysis. Relevance helps you find these cases.","keywords":""},{"kind":"section","title":"Topic and scenario relevance","heading":"Interpret Relevance scores","section":"Documentation","crumbs":["Core concepts","Topic and scenario relevance","Interpret Relevance scores"],"url":"/docs/core-concepts/topic-scenario-relevance#interpret-relevance-scores","text":"","keywords":""},{"kind":"section","title":"Topic and scenario relevance","heading":"High relevance (near 100%)","section":"Documentation","crumbs":["Core concepts","Topic and scenario relevance","High relevance (near 100%)"],"url":"/docs/core-concepts/topic-scenario-relevance#high-relevance-near-100","text":"A relevance score near 100% shows that the scenario always causes the answer engine to talk about brands. This usually means that the scenario is well written. It shows correctly how users ask answer engines for recommendations. These scenarios are the best scenarios to measure Brand Visibility and the competitive position.","keywords":""},{"kind":"section","title":"Topic and scenario relevance","heading":"Moderate relevance","section":"Documentation","crumbs":["Core concepts","Topic and scenario relevance","Moderate relevance"],"url":"/docs/core-concepts/topic-scenario-relevance#moderate-relevance","text":"A relevance score in the middle range shows that brands show in some answers, but not in all answers. This can show that: - The scenario is not sufficiently clear. - The question is too general. - The answer engine sometimes replies with no product recommendations.","keywords":""},{"kind":"section","title":"Topic and scenario relevance","heading":"Low relevance","section":"Documentation","crumbs":["Core concepts","Topic and scenario relevance","Low relevance"],"url":"/docs/core-concepts/topic-scenario-relevance#low-relevance","text":"A low relevance score shows that answer engines almost never mention brands in the answers. This usually means that you must make the scenario better. Possible problems are: - The scenario is too abstract. - The question does not show that the user looks for a product. - The goal of the user is not clear. When you make the scenario better, it can become much more useful for the analysis.","keywords":""},{"kind":"section","title":"Topic and scenario relevance","heading":"Relevance of a scenario and relevance of a topic","section":"Documentation","crumbs":["Core concepts","Topic and scenario relevance","Relevance of a scenario and relevance of a topic"],"url":"/docs/core-concepts/topic-scenario-relevance#relevance-of-a-scenario-and-relevance-of-a-topic","text":"Genezio calculates Relevance at two levels.","keywords":""},{"kind":"section","title":"Topic and scenario relevance","heading":"Scenario Relevance","section":"Documentation","crumbs":["Core concepts","Topic and scenario relevance","Scenario Relevance"],"url":"/docs/core-concepts/topic-scenario-relevance#scenario-relevance","text":"Each scenario has its own relevance score. Genezio calculates it from the conversations of that scenario.","keywords":""},{"kind":"section","title":"Topic and scenario relevance","heading":"Topic Relevance","section":"Documentation","crumbs":["Core concepts","Topic and scenario relevance","Topic Relevance"],"url":"/docs/core-concepts/topic-scenario-relevance#topic-relevance","text":"The relevance of a topic is the average relevance of its scenarios. This helps users know if the topic gives useful conversations in which answer engines talk about brands.","keywords":""},{"kind":"section","title":"Topic and scenario relevance","heading":"Relevance and the quality of scenarios","section":"Documentation","crumbs":["Core concepts","Topic and scenario relevance","Relevance and the quality of scenarios"],"url":"/docs/core-concepts/topic-scenario-relevance#relevance-and-the-quality-of-scenarios","text":"Relevance is also a useful indicator of the quality of a scenario. A well-designed scenario must: - Clearly describe a situation of the user - Have a realistic goal - Cause the answer engine to recommend products When a scenario has these conditions, answer engines usually reply with brand suggestions. Thus, the relevance score is higher.","keywords":""},{"kind":"section","title":"Topic and scenario relevance","heading":"Make Relevance better","section":"Documentation","crumbs":["Core concepts","Topic and scenario relevance","Make Relevance better"],"url":"/docs/core-concepts/topic-scenario-relevance#make-relevance-better","text":"If a scenario has a low relevance, you can frequently make it better. Make the goal of the user more clear. Do not use a scenario that is not clear, such as: A stronger scenario describes a specific situation, such as: This type of situation more frequently causes brand recommendations. For more guidance, refer to Scenarios.","keywords":"Tell me about customer management. The persona runs a small startup and needs a tool to manage incoming sales leads and follow up with potential customers."},{"kind":"section","title":"Topic and scenario relevance","heading":"Why Relevance keeps the analysis correct","section":"Documentation","crumbs":["Core concepts","Topic and scenario relevance","Why Relevance keeps the analysis correct"],"url":"/docs/core-concepts/topic-scenario-relevance#why-relevance-keeps-the-analysis-correct","text":"With Relevance, the conversations that measure AI visibility show realistic decision situations. Without a relevance filter, the analysis can include conversations in which no brands show. Thus, the visibility metrics become less useful. When teams monitor Relevance, they can continuously make the scenarios better. Thus, the quality of the AI visibility analysis increases.","keywords":""},{"kind":"section","title":"Topic and scenario relevance","heading":"Next steps","section":"Documentation","crumbs":["Core concepts","Topic and scenario relevance","Next steps"],"url":"/docs/core-concepts/topic-scenario-relevance#next-steps","text":"To know how the brand mentions in relevant conversations count for the visibility metrics, refer to: - Brand Visibility - Scenarios","keywords":""},{"kind":"page","title":"Conversations","heading":"","section":"Documentation","crumbs":["Documentation","Core concepts","Conversations"],"url":"/docs/core-concepts/conversations","text":"A conversation is one run of a scenario with one answer engine. Learn how Genezio runs conversations, what data they give, and how Google AI Overview differs. In Genezio, a conversation is one run of a scenario with one answer engine. In a conversation, Genezio sends prompts to an answer engine and records the results for the analysis. Thus, each conversation is an instance of a scenario that Genezio runs with one answer engine.","keywords":""},{"kind":"section","title":"Conversations","heading":"Scenario and conversation","section":"Documentation","crumbs":["Core concepts","Conversations","Scenario and conversation"],"url":"/docs/core-concepts/conversations#scenario-and-conversation","text":"The relation between a scenario and a conversation is simple: - A scenario gives the situation and the goal of the persona. - A conversation occurs when Genezio runs that scenario with an answer engine. For example: - Scenario: \"A startup founder looking for a CRM to manage early leads\" - Answer engine: ChatGPT When Genezio runs this scenario with ChatGPT, it makes one conversation. When Genezio runs the same scenario with more than one answer engine, it makes more than one conversation. For example: CRM for startups scenario: ChatGPT Conversation A. CRM for startups scenario: Claude Conversation B. CRM for startups scenario: Gemini Conversation C Each of these conversations can give different responses, citations, and brand mentions.","keywords":""},{"kind":"section","title":"Conversations","heading":"Single-turn and multi-step conversations","section":"Documentation","crumbs":["Core concepts","Conversations","Single-turn and multi-step conversations"],"url":"/docs/core-concepts/conversations#single-turn-and-multi-step-conversations","text":"The structure of a conversation depends on the type of the Genezio agent.","keywords":""},{"kind":"section","title":"Conversations","heading":"Conversations of the Prompter Agent","section":"Documentation","crumbs":["Core concepts","Conversations","Conversations of the Prompter Agent"],"url":"/docs/core-concepts/conversations#conversations-of-the-prompter-agent","text":"For a topic of the Prompter Agent, the conversation has one message. Genezio sends the prompt that you wrote word for word to the answer engine. Example prompt: User query: What CRM should a startup use? Genezio records the response of the answer engine and analyzes it.","keywords":""},{"kind":"section","title":"Conversations","heading":"Multi-step conversations","section":"Documentation","crumbs":["Core concepts","Conversations","Multi-step conversations"],"url":"/docs/core-concepts/conversations#multi-step-conversations","text":"For a topic of the Recommender Agent, the Introspector Agent, or the Comparer Agent, the conversation is multi-step: 1. The scenario gives the initial situation. 2. Genezio makes the prompts during the conversation. 3. The conversation can have many turns in the two directions. These conversations simulate how a real user explores a topic with an AI assistant.","keywords":""},{"kind":"section","title":"Conversations","heading":"Conversations and answer engines","section":"Documentation","crumbs":["Core concepts","Conversations","Conversations and answer engines"],"url":"/docs/core-concepts/conversations#conversations-and-answer-engines","text":"Genezio runs conversations with many answer engines. Examples: - ChatGPT - Claude - Gemini - Perplexity - Google AI Overview When Genezio runs scenarios with different answer engines, it can measure how the visibility of a brand changes from one platform to another. To select the answer engines, refer to Selecting answer engines.","keywords":""},{"kind":"section","title":"Conversations","heading":"Special case: Google AI Overview","section":"Documentation","crumbs":["Core concepts","Conversations","Special case: Google AI Overview"],"url":"/docs/core-concepts/conversations#special-case-google-ai-overview","text":"Google AI Overview operates differently from a conversational AI assistant. AI Overview does not start from a conversational prompt. It starts from a text that is almost the same as a web search query. Because of this difference: - Genezio changes the prompts into search-style queries for AI Overview. - These queries are almost the same as the text that a user types in the Google search box. Example of a conversation prompt: User query: What CRM should a startup use? The AI Overview query for the same prompt: AI Overview query: best CRM for startups Genezio makes these queries automatically. But users can override the query that Genezio makes. Do this to test a specific search text.","keywords":""},{"kind":"section","title":"Conversations","heading":"Conversation data","section":"Documentation","crumbs":["Core concepts","Conversations","Conversation data"],"url":"/docs/core-concepts/conversations#conversation-data","text":"Each conversation gives a structured set of data that Genezio analyzes. This data includes: - The prompts that Genezio sent to the answer engine - The responses of the model - The brands that the response mentions - The cited sources - The follow-up searches (query fanouts) of the answer engine This data is the base for metrics such as AI Recommendations, AI Visibility, share of voice, and competitive analysis.","keywords":""},{"kind":"section","title":"Conversations","heading":"Conversation detail view","section":"Documentation","crumbs":["Core concepts","Conversations","Conversation detail view"],"url":"/docs/core-concepts/conversations#conversation-detail-view","text":"Genezio has a conversation detail view. In this view, users examine the results of one conversation. This view shows: - The full conversation between Genezio and the answer engine - The prompts of each turn - The responses of the model - The query fanouts that Genezio extracted - The citations that Genezio found - The brands and competitors that the response mentions - The products that the response mentions, tagged and highlighted in the text With this view, teams can see exactly why a brand appeared, or did not appear, in a response of an answer engine. The transcripts know the products. Genezio highlights the products that an answer names, and it shows tables and maps in the text. Thus, you can find the moment when an answer engine recommended a specific product or ignored it. Refer to Product visibility.","keywords":""},{"kind":"section","title":"Conversations","heading":"Why conversations are important","section":"Documentation","crumbs":["Core concepts","Conversations","Why conversations are important"],"url":"/docs/core-concepts/conversations#why-conversations-are-important","text":"A conversation is the core unit of analysis in Genezio. Genezio calculates each visibility metric, each insight, and each competitor analysis from the results of many conversations. These conversations cover many topics, personas, and answer engines. Genezio analyzes these conversations at scale. Thus, it gives a measurable view of how answer engines show brands.","keywords":""},{"kind":"section","title":"Conversations","heading":"Next steps","section":"Documentation","crumbs":["Core concepts","Conversations","Next steps"],"url":"/docs/core-concepts/conversations#next-steps","text":"To learn how Genezio structures and analyzes the information from conversations, refer to: - Citations - Perceptions - Running conversations","keywords":""},{"kind":"section","title":"Conversations","heading":"Conversations in the public API","section":"Documentation","crumbs":["Core concepts","Conversations","Conversations in the public API"],"url":"/docs/core-concepts/conversations#conversations-in-the-public-api","text":"The public API reads conversations. Refer to List conversations.","keywords":""},{"kind":"page","title":"Query fanouts","heading":"","section":"Documentation","crumbs":["Documentation","Core concepts","Query fanouts"],"url":"/docs/core-concepts/query-fanouts","text":"Query fanouts are the extra searches that an answer engine runs to answer one question. Learn why they exist and how Genezio shows them for each conversation. In Genezio, query fanouts are the additional searches that an answer engine does internally when it answers a question. Genezio extracts these searches from each conversation, so that you can see what the answer engine searched for and which sources it found. When a user asks a question, a modern answer engine almost never uses only one query. It expands the initial question into many related searches to get more information. These additional searches are the query fanouts, also known as follow-up searches. Query fanouts are important because they have a large effect on: - The sources that the answer engine gets - The brands that the answers mention - The citations that the answer engine makes For a longer introduction, refer to Query fanouts explained.","keywords":""},{"kind":"section","title":"Query fanouts","heading":"Why query fanouts exist","section":"Documentation","crumbs":["Core concepts","Query fanouts","Why query fanouts exist"],"url":"/docs/core-concepts/query-fanouts#why-query-fanouts-exist","text":"User questions are frequently broad or not clear. One search query can get too little information for a useful answer. To make the responses better, the answer engine expands the initial question into many related queries. Each query examines the topic from a different angle. For example, the question: User query: What CRM should a startup use? can expand internally into searches such as: - best CRM for startups - startup CRM comparison - HubSpot vs Pipedrive for startups - affordable CRM tools for small teams Each of these searches gets different documents. The answer engine can use these documents to make its response.","keywords":""},{"kind":"section","title":"Query fanouts","heading":"Where query fanouts come from","section":"Documentation","crumbs":["Core concepts","Query fanouts","Where query fanouts come from"],"url":"/docs/core-concepts/query-fanouts#where-query-fanouts-come-from","text":"Most answer engines do not publish their search infrastructure. Only the companies that make these systems know which search engines or indexes they use. But public statements, research papers, and observed behavior show a probable answer. Answer engines use a combination of sources, such as: - Large web search engines (such as Google or Bing) - Internal search tools of the AI provider - Databases of open web content Thus, the answer engine can send its fanout queries to many different search tools before the model gets the content. These systems change continuously. Thus, the sources can be different for each answer engine, and they can change over time. For your content strategy, this is the important point: the search channels must find your content. One channel is not sufficient.","keywords":""},{"kind":"section","title":"Query fanouts","heading":"How query fanouts change the answers","section":"Documentation","crumbs":["Core concepts","Query fanouts","How query fanouts change the answers"],"url":"/docs/core-concepts/query-fanouts#how-query-fanouts-change-the-answers","text":"Query fanouts control which documents the answer engine gets when it collects information. These documents have an effect on: - The brands that the answer mentions - The cited sources - The claims in the response For example, a fanout query can get comparison articles that mention specific products. Then, the probability increases that these products appear in the answer. Because of this, query fanouts are one of the strongest signals for AI Recommendations and Visibility.","keywords":""},{"kind":"section","title":"Query fanouts","heading":"Query fanouts in Genezio","section":"Documentation","crumbs":["Core concepts","Query fanouts","Query fanouts in Genezio"],"url":"/docs/core-concepts/query-fanouts#query-fanouts-in-genezio","text":"Genezio finds and extracts the query fanouts of the conversations with answer engines. For each conversation, Genezio analyzes: - The additional queries of the answer engine - The relation between these queries and the initial prompt - The sources that the answer engine got for each query Thus, teams can see what the answer engine searched for when it answered a question.","keywords":""},{"kind":"section","title":"Query fanouts","heading":"Example of query fanouts for one prompt","section":"Documentation","crumbs":["Core concepts","Query fanouts","Example of query fanouts for one prompt"],"url":"/docs/core-concepts/query-fanouts#example-of-query-fanouts-for-one-prompt","text":"Initial prompt: User query: What CRM should a startup use? The fanout queries that Genezio finds can include: - Query fanout: best CRM for startups - Query fanout: CRM tools for SaaS startups - Query fanout: HubSpot vs Pipedrive - Query fanout: startup CRM pricing comparison These queries show how the answer engine expands the initial question to make a response.","keywords":""},{"kind":"section","title":"Query fanouts","heading":"Why query fanouts are important","section":"Documentation","crumbs":["Core concepts","Query fanouts","Why query fanouts are important"],"url":"/docs/core-concepts/query-fanouts#why-query-fanouts-are-important","text":"The analysis of fanout queries helps organizations to understand: - How answer engines interpret a topic - Which questions have an effect on the answers - Which sources control the available information - Where opportunities for content exist For example, fanout queries can frequently include topics that the content of your brand does not cover. This can show an opportunity to increase your visibility. To change these gaps into content, refer to From data to content strategy.","keywords":""},{"kind":"section","title":"Query fanouts","heading":"Query fanouts and keywords","section":"Documentation","crumbs":["Core concepts","Query fanouts","Query fanouts and keywords"],"url":"/docs/core-concepts/query-fanouts#query-fanouts-and-keywords","text":"Traditional SEO optimizes pages for specific keywords. Query fanouts show a broader concept: how answer engines examine a topic. An answer engine does not use only one keyword. It examines many related queries before it makes an answer. When you know this set of queries, you understand better how answer engines make their responses.","keywords":""},{"kind":"section","title":"Query fanouts","heading":"Query fanouts in the conversation view","section":"Documentation","crumbs":["Core concepts","Query fanouts","Query fanouts in the conversation view"],"url":"/docs/core-concepts/query-fanouts#query-fanouts-in-the-conversation-view","text":"In Genezio, you can examine query fanouts in the conversation detail view. This view shows: - The initial prompt - The fanout queries that Genezio found - The sources that the answer engine got for these queries - The final response of the answer engine With this view, users can understand how an answer engine got to a specific answer.","keywords":""},{"kind":"section","title":"Query fanouts","heading":"Next steps","section":"Documentation","crumbs":["Core concepts","Query fanouts","Next steps"],"url":"/docs/core-concepts/query-fanouts#next-steps","text":"To learn how the sources from fanout queries change the answers, refer to: - Citations - Perceptions","keywords":""},{"kind":"section","title":"Query fanouts","heading":"Query fanouts in the public API","section":"Documentation","crumbs":["Core concepts","Query fanouts","Query fanouts in the public API"],"url":"/docs/core-concepts/query-fanouts#query-fanouts-in-the-public-api","text":"The public API reads query fanouts. Refer to List the searches.","keywords":""},{"kind":"page","title":"Citations","heading":"","section":"Documentation","crumbs":["Documentation","Core concepts","Citations"],"url":"/docs/core-concepts/citations","text":"Citations are the web pages that answer engines use as references. Learn how Genezio groups and classifies AI citations and links them to brand perceptions. In Genezio, citations are the external web pages that an answer engine uses as references for an answer. Genezio records each citation, classifies it, and shows which sources have an effect on how answer engines describe your brand. Answer engines such as ChatGPT, Claude, Gemini, Perplexity, and Google AI Overview frequently use information from the web in their responses. Some systems show these references as links or as numbered sources. Genezio records and analyzes these references as citations. Citations help to explain: - Where the information in the answer came from - Which websites have an effect on the answers of answer engines - Why specific brands appear in answers For the background, refer to How AI citations work.","keywords":""},{"kind":"section","title":"Citations","heading":"What a citation is","section":"Documentation","crumbs":["Core concepts","Citations","What a citation is"],"url":"/docs/core-concepts/citations#what-a-citation-is","text":"A citation is a specific web page or document that had an effect on an answer. Frequent types of citations are: - A product page or a service page - A comparison article - A review site - A documentation page - A regulatory source or a research source Genezio records each referenced URL as a citation.","keywords":""},{"kind":"section","title":"Citations","heading":"Citations and sources","section":"Documentation","crumbs":["Core concepts","Citations","Citations and sources"],"url":"/docs/core-concepts/citations#citations-and-sources","text":"Different answer engines use different terms. Some interfaces show links as citations, and others show them as sources. In Genezio, the two terms have the same meaning. They are the web pages that gave information to the answer engine for its answer.","keywords":""},{"kind":"section","title":"Citations","heading":"Domain grouping","section":"Documentation","crumbs":["Core concepts","Citations","Domain grouping"],"url":"/docs/core-concepts/citations#domain-grouping","text":"One website can have many pages that appear in many conversations. To make the analysis simpler, Genezio can group citations by domain. For example, you do not have to analyze many URLs from the same website. Genezio can put them together under one domain, such as: Domain grouping helps users to quickly find: - The websites that have the largest effect on the answers - The publications that control a topic - The domains that frequently mention competitors Thus, you can analyze the full influence of a website, and not only the influence of each page.","keywords":"example.com"},{"kind":"section","title":"Citations","heading":"Citation attributes","section":"Documentation","crumbs":["Core concepts","Citations","Citation attributes"],"url":"/docs/core-concepts/citations#citation-attributes","text":"Each citation in Genezio has attributes that help you to analyze its role in the answers.","keywords":""},{"kind":"section","title":"Citations","heading":"URL","section":"Documentation","crumbs":["Core concepts","Citations","URL"],"url":"/docs/core-concepts/citations#url","text":"The URL is the exact web page that the answer engine used as a reference.","keywords":""},{"kind":"section","title":"Citations","heading":"Occurrences","section":"Documentation","crumbs":["Core concepts","Citations","Occurrences"],"url":"/docs/core-concepts/citations#occurrences","text":"The occurrences are the number of conversations that have the citation. A high number of occurrences frequently shows that a source has a strong effect on the answers for that topic.","keywords":""},{"kind":"section","title":"Citations","heading":"Sentiment","section":"Documentation","crumbs":["Core concepts","Citations","Sentiment"],"url":"/docs/core-concepts/citations#sentiment","text":"Genezio analyzes how the citation describes the brand. The possible classifications are: - Positive - Neutral - Negative This helps you to find the sources that show a brand favorably or critically.","keywords":""},{"kind":"section","title":"Citations","heading":"Automatic citation classification","section":"Documentation","crumbs":["Core concepts","Citations","Automatic citation classification"],"url":"/docs/core-concepts/citations#automatic-citation-classification","text":"Genezio classifies citations automatically. This helps users to understand the type of source that has an effect on the answers. The citation interface shows these classifications.","keywords":""},{"kind":"section","title":"Citations","heading":"Source ownership","section":"Documentation","crumbs":["Core concepts","Citations","Source ownership"],"url":"/docs/core-concepts/citations#source-ownership","text":"Genezio finds if a citation is of the analyzed brand, of a competitor, or of a neutral third party. Examples: - First Party - Competitor Owned - Third-Party Editorial","keywords":""},{"kind":"section","title":"Citations","heading":"Source category","section":"Documentation","crumbs":["Core concepts","Citations","Source category"],"url":"/docs/core-concepts/citations#source-category","text":"Genezio finds the type of content that has the citation. Examples: - Blog / Article - Review (Editorial) - Review (User-Generated) - Comparison / Vs Page - Listicle / Best-Of - News Article - Press Release - Documentation / Knowledge Base - Case Study - Research Paper / Report","keywords":""},{"kind":"section","title":"Citations","heading":"Website type","section":"Documentation","crumbs":["Core concepts","Citations","Website type"],"url":"/docs/core-concepts/citations#website-type","text":"The website type is the broader category of the website that has the content. Examples: - Industry / Trade Publication - Government / Regulatory - Academic / Research - Aggregator / Marketplace - Social / UGC Platform - Other","keywords":""},{"kind":"section","title":"Citations","heading":"Auto-clarification of citations","section":"Documentation","crumbs":["Core concepts","Citations","Auto-clarification of citations"],"url":"/docs/core-concepts/citations#auto-clarification-of-citations","text":"Genezio automatically adds data to citations and makes them clear. It analyzes each source and gives it these classifications. This process helps users to quickly understand: - If a citation is of a competitor - If the source is editorial, a review, or a product page - If the source is first party or third party Users do not have to examine each link manually. The citation list shows the role and the type of each source immediately.","keywords":""},{"kind":"section","title":"Citations","heading":"Citations in the conversation view","section":"Documentation","crumbs":["Core concepts","Citations","Citations in the conversation view"],"url":"/docs/core-concepts/citations#citations-in-the-conversation-view","text":"You can examine citations in the conversation detail view. This view shows: - The response of the answer engine - The citations (sources) that the answer engine used as references - The related follow-up searches (query fanouts) - The brand mentions that Genezio found With this view, users can see exactly which sources gave information to a specific answer.","keywords":""},{"kind":"section","title":"Citations","heading":"Citations and perceptions","section":"Documentation","crumbs":["Core concepts","Citations","Citations and perceptions"],"url":"/docs/core-concepts/citations#citations-and-perceptions","text":"Citations are more than a list of links. They have a direct relation to what answer engines say about your brand. From each citation, you can see the perceptions that come from that source. These are the claims and the impressions that an answer engine made with information from that page. This relation operates in the two directions. You can also see perceptions by citation domain. Thus, you can easily see which websites control how answers describe your brand. To learn how Genezio records and analyzes what answer engines say, refer to Perceptions.","keywords":""},{"kind":"section","title":"Citations","heading":"Cited sources and mentioned sources","section":"Documentation","crumbs":["Core concepts","Citations","Cited sources and mentioned sources"],"url":"/docs/core-concepts/citations#cited-sources-and-mentioned-sources","text":"Not all sources have the same role in an answer. Genezio makes a distinction between: - Cited sources: the pages that the answer engine explicitly cited as references - Mentioned sources: the pages that the answer engine only referred to or showed, without an explicit citation This distinction helps you to focus on the sources that the answer engine really uses. The other sources only appeared during the process.","keywords":""},{"kind":"section","title":"Citations","heading":"Work with citations","section":"Documentation","crumbs":["Core concepts","Citations","Work with citations"],"url":"/docs/core-concepts/citations#work-with-citations","text":"The citation interface has these functions, which make the analysis of sources easier: - Group sources by domain: Put many URLs from the same website into one domain view. Thus, you can find the full influence of a site, and not the influence of each page. - Filter by ownership: Show only \"My citations\" (the first-party sources that you own) or only the third-party sources. Thus, you can quickly separate your content from all other content. - \"Competitors mentioned\" chip: A citation row shows a chip when the source mentions competitors. Thus, you can quickly see the competitive exposure.","keywords":""},{"kind":"section","title":"Citations","heading":"Why citation analysis is important","section":"Documentation","crumbs":["Core concepts","Citations","Why citation analysis is important"],"url":"/docs/core-concepts/citations#why-citation-analysis-is-important","text":"Citation analysis helps organizations to understand the information that has an effect on answer engines. When teams find the sources that appear most frequently, they can: - Understand which websites control the responses of answer engines - Find the important publications in their category - Find the sources that mention competitors - Find opportunities to increase visibility","keywords":""},{"kind":"section","title":"Citations","heading":"Next steps","section":"Documentation","crumbs":["Core concepts","Citations","Next steps"],"url":"/docs/core-concepts/citations#next-steps","text":"To learn how Genezio changes citations and the other extracted signals into measurable visibility insights, refer to: - Perceptions - Share of Voice","keywords":""},{"kind":"section","title":"Citations","heading":"Citations in the public API","section":"Documentation","crumbs":["Core concepts","Citations","Citations in the public API"],"url":"/docs/core-concepts/citations#citations-in-the-public-api","text":"The public API reads citations. Refer to Cited domains.","keywords":""},{"kind":"page","title":"Perceptions","heading":"","section":"Documentation","crumbs":["Documentation","Core concepts","Perceptions"],"url":"/docs/core-concepts/perceptions","text":"A perception is one claim that an answer engine makes about a brand. Learn how Genezio checks if perceptions are correct and traces them to their sources. In Genezio, a perception is one claim that Genezio extracts from an answer of an answer engine about a brand, a product, or a category. Perceptions show what answer engines say about your brand, if it is correct, and which sources caused it. The response of an answer engine usually has many claims. Genezio divides each response into smaller units, the perceptions. Thus, Genezio can evaluate each claim separately. Together, these claims show how answer engines perceive your brand and how they describe it to buyers. The old name of perceptions is Statements. The concept did not change. The new name shows better that these claims are how answer engines perceive your brand. Perceptions help to answer these questions: - What exactly did the answer engine say about a brand? - Is the claim correct or incorrect? - Which source supports the claim? - Which competitors does the same context mention?.","keywords":""},{"kind":"section","title":"Perceptions","heading":"Why perceptions are important","section":"Documentation","crumbs":["Core concepts","Perceptions","Why perceptions are important"],"url":"/docs/core-concepts/perceptions#why-perceptions-are-important","text":"The responses of answer engines are frequently long paragraphs with many ideas. If you do not divide them, it is difficult to find which claims are correct. For example, an AI assistant can give this answer: Popular CRM tools for startups include HubSpot, Pipedrive, and Zoho CRM. HubSpot is known for its strong marketing automation features, while Pipedrive focuses on simple sales pipeline management. This response has many different claims. Genezio extracts them as separate perceptions, such as: - HubSpot is a CRM tool used by startups - Pipedrive is a CRM tool used by startups - Zoho CRM is a CRM tool used by startups - HubSpot is known for marketing automation - Pipedrive focuses on pipeline management Each of these is one perception. Genezio can compare each perception with what your brand really does, and find if it is correct.","keywords":""},{"kind":"section","title":"Perceptions","heading":"Perception accuracy","section":"Documentation","crumbs":["Core concepts","Perceptions","Perception accuracy"],"url":"/docs/core-concepts/perceptions#perception-accuracy","text":"The most important property of a perception is if it is correct or incorrect. Answer engines make claims about brands all the time. These claims control how buyers perceive you. Some claims are correct. Other claims are old, misleading, or wrong. Examples: - \"Zara is a fast fashion brand\": Correct - \"Gucci is a fast fashion brand\": Incorrect - \"HubSpot offers a free CRM tier\": Correct - \"Pipedrive includes built-in marketing automation\": Incorrect An incorrect claim about your brand can mislead buyers, and you can lose recommendations. A correct claim makes your positioning stronger.","keywords":""},{"kind":"section","title":"Perceptions","heading":"How Genezio finds the accuracy","section":"Documentation","crumbs":["Core concepts","Perceptions","How Genezio finds the accuracy"],"url":"/docs/core-concepts/perceptions#how-genezio-finds-the-accuracy","text":"To find if a perception is correct, Genezio compares it with your Knowledge Base. The Knowledge Base is a source of truth for your brand. You configure it one time in Brand Settings. The Knowledge Base usually has: - Product descriptions and capabilities - Prices and plans - The target audience and the use cases - The competitive differentiators - All other official brand guidelines Genezio compares the claims of answer engines with your Knowledge Base. Thus, it can flag the perceptions that show your brand incorrectly. You get a clear view of where answer engines are correct and where they are wrong.","keywords":""},{"kind":"section","title":"Perceptions","heading":"Grounded perceptions","section":"Documentation","crumbs":["Core concepts","Perceptions","Grounded perceptions"],"url":"/docs/core-concepts/perceptions#grounded-perceptions","text":"When you have a Knowledge Base, each perception from a conversation gets a grounded badge. The badge shows if Genezio could verify the perception, and how. The badge has three states: - Grounded-correct: The perception agrees with your Knowledge Base. - Grounded-incorrect: The perception does not agree with your Knowledge Base. - Not grounded: Genezio could not verify the perception with your Knowledge Base. A tooltip on the badge tells what each state means in its context. The badge changes a list of perceptions into a triage queue: 1. Examine the grounded-incorrect perceptions first. In these perceptions, answer engines show your brand incorrectly. 2. Then, examine the not-grounded perceptions. They show gaps in your Knowledge Base or in the public documentation of your messages. The grounded-correct perceptions are your wins. They are the proof that the narrative of answer engines about you agrees with the truth. Each brand gets a Knowledge Base from day one. Genezio crawls the website of the brand automatically when you create the brand. Thus, the grounded layer is active immediately. The number of perceptions that Genezio can verify depends on how much truth your Knowledge Base has. To find what to add, refer to Knowledge Base. Perceptions observe what answer engines say without a direct question. To test specific claims with direct questions, use the Fact Checker agent.","keywords":""},{"kind":"section","title":"Perceptions","heading":"Perceptions and their sources","section":"Documentation","crumbs":["Core concepts","Perceptions","Perceptions and their sources"],"url":"/docs/core-concepts/perceptions#perceptions-and-their-sources","text":"You can trace each perception back to the sources that made it. This is one of the most useful functions of a perception: you can see why the answer engine said it. When you open a perception, Genezio shows the citation paragraphs that support the claim. These are the exact passages from the cited web pages, with the matching text highlighted. For example, the answer engine says that your product is hard to set up. You can go from this claim directly to the third-party page that gave the answer engine this idea. Thus, you know exactly where to put your effort for a correction or for outreach. The link operates in the two directions: - From a perception: See the sources (citation paragraphs) that support it, with the matched text highlighted. - From a citation: See all the perceptions that come from that source. - By citation domain: See the perceptions of a specific domain. Thus, you can find which websites control your narrative. Genezio also makes a distinction between sources that the answer engine explicitly cites and brands or pages that it only mentions. Thus, you can separate the references that the model really used from the mentions that are not important. To learn how Genezio records and classifies sources, refer to Citations.","keywords":""},{"kind":"section","title":"Perceptions","heading":"Other attributes of a perception","section":"Documentation","crumbs":["Core concepts","Perceptions","Other attributes of a perception"],"url":"/docs/core-concepts/perceptions#other-attributes-of-a-perception","text":"","keywords":""},{"kind":"section","title":"Perceptions","heading":"Text","section":"Documentation","crumbs":["Core concepts","Perceptions","Text"],"url":"/docs/core-concepts/perceptions#text","text":"The text is the claim in the response of the answer engine. Example:","keywords":"HubSpot is known for strong marketing automation features."},{"kind":"section","title":"Perceptions","heading":"Brand references","section":"Documentation","crumbs":["Core concepts","Perceptions","Brand references"],"url":"/docs/core-concepts/perceptions#brand-references","text":"The brand references are the brands that the perception mentions. One perception can mention one brand or more than one brand. Example:","keywords":"HubSpot and Salesforce are widely used CRM platforms."},{"kind":"section","title":"Perceptions","heading":"Supporting citations","section":"Documentation","crumbs":["Core concepts","Perceptions","Supporting citations"],"url":"/docs/core-concepts/perceptions#supporting-citations","text":"Each perception has a link to one or more citations (sources). The answer engine used these citations when it made the claim. Thus, you can trace a perception back to the web page that had an effect on it. You can also understand why the answer engine made a specific claim.","keywords":""},{"kind":"section","title":"Perceptions","heading":"Perceptions across conversations","section":"Documentation","crumbs":["Core concepts","Perceptions","Perceptions across conversations"],"url":"/docs/core-concepts/perceptions#perceptions-across-conversations","text":"Genezio extracts perceptions from each conversation that it runs with answer engines. Genezio writes the perceptions in the language of your brand. Thus, the reports are easy to read for local markets. Genezio aggregates the perceptions of many conversations. Thus, it can find patterns such as: - Claims about a brand that occur again and again, correct or incorrect - Frequent errors that you must correct - Competitors that the answers mention frequently - Differences in how answer engines describe the same brand This helps organizations to understand the narrative that answer engines connect to their brand. It also shows if that narrative is correct.","keywords":""},{"kind":"section","title":"Perceptions","heading":"Perceptions in the interface","section":"Documentation","crumbs":["Core concepts","Perceptions","Perceptions in the interface"],"url":"/docs/core-concepts/perceptions#perceptions-in-the-interface","text":"In Genezio, you can examine perceptions in the conversation detail view and in the aggregated analysis views. You can sort a list of perceptions to show the most relevant claims first. Users can examine: - The initial response of the answer engine - The extracted perceptions - The accuracy of each claim - The related citations, with the supporting paragraphs highlighted - The brands that each claim mentions With these views, teams can see exactly what answer engines say about their brand, and if it is true.","keywords":""},{"kind":"section","title":"Perceptions","heading":"Perceptions and insights","section":"Documentation","crumbs":["Core concepts","Perceptions","Perceptions and insights"],"url":"/docs/core-concepts/perceptions#perceptions-and-insights","text":"Perceptions are the base for many higher-level insights in Genezio. They have an effect on: - AI Recommendations: Incorrect claims can decrease your recommendation rate. - AI Visibility: Incorrect descriptions change how frequently answer engines show your brand. - Share of Voice: The accuracy of claims has an effect on the competitive position. - Competitive positioning: You can see which claims are better for you and which are better for competitors. - AI Perception Summary: The full narrative, made by AI, of how answer engines perceive your brand. Genezio puts the responses of answer engines into the structure of perceptions, and it evaluates their accuracy. Thus, Genezio changes unstructured answers into data that you can measure, verify, and use.","keywords":""},{"kind":"section","title":"Perceptions","heading":"Next steps","section":"Documentation","crumbs":["Core concepts","Perceptions","Next steps"],"url":"/docs/core-concepts/perceptions#next-steps","text":"To learn how Genezio changes perceptions and citations into measurable metrics of AI visibility, refer to: - Share of Voice - AI Perception Summary","keywords":""},{"kind":"section","title":"Perceptions","heading":"Perceptions in the public API","section":"Documentation","crumbs":["Core concepts","Perceptions","Perceptions in the public API"],"url":"/docs/core-concepts/perceptions#perceptions-in-the-public-api","text":"The public API reads perceptions. Refer to Perceptions.","keywords":""},{"kind":"page","title":"Competitors","heading":"","section":"Documentation","crumbs":["Documentation","Core concepts","Competitors"],"url":"/docs/core-concepts/competitors","text":"Genezio finds the competitors that answer engines mention with your brand. Learn how competitor detection works and how to promote, demote, or merge them. In Genezio, competitors are the brands that appear in the answers of answer engines together with the analyzed brand. Genezio finds them automatically, and you can change the list so that it shows your real market. When answer engines answer questions, they frequently recommend or compare many products or services of the same category. Genezio automatically finds these brands and classifies them as competitors. Genezio monitors competitors at the brand level. Each brand that you create in Genezio has its own competitor landscape. For a company with many products, the product model is thus important: - Each product has different competitors and its own brand presence: make a separate brand for each product. This gives the clearest view. - The parent brand is strong, but the products do not have independent brand recognition: make one brand with Master Filters. For the full decision, refer to Companies with many products in Brands. When you monitor competitors, you understand the position of your brand in relation to the alternatives in the answers of answer engines.","keywords":""},{"kind":"section","title":"Competitors","heading":"Automatic competitor detection","section":"Documentation","crumbs":["Core concepts","Competitors","Automatic competitor detection"],"url":"/docs/core-concepts/competitors#automatic-competitor-detection","text":"Genezio automatically finds competitor brands when it analyzes conversations. Genezio finds competitors when: - A response of an answer engine recommends many products - The same answer mentions many brands - The answer compares products with each other - The cited sources of the answer engine mention competing brands For example, a conversation asks about tools that manage customer relationships. The answer engine can mention many products in the same answer. Genezio automatically extracts these brands and monitors them as competitors. Genezio aggregates these mentions across many conversations. Thus, it makes a competitor landscape for each topic.","keywords":""},{"kind":"section","title":"Competitors","heading":"Competitors across topics","section":"Documentation","crumbs":["Core concepts","Competitors","Competitors across topics"],"url":"/docs/core-concepts/competitors#competitors-across-topics","text":"The competitors can change with the analyzed topic. For example, a brand can compete with different products for different user intents or scenarios. Examples: - Topic: CRM for startups - Topic: marketing automation platforms The set of competitors in the answers can be different for these topics. This helps organizations to understand how answer engines interpret the competitive landscape for different problems.","keywords":""},{"kind":"section","title":"Competitors","heading":"Manage competitors","section":"Documentation","crumbs":["Core concepts","Competitors","Manage competitors"],"url":"/docs/core-concepts/competitors#manage-competitors","text":"Competitor detection is automatic. But users can change the competitor list to show their market better. Genezio lets users do these operations.","keywords":""},{"kind":"section","title":"Competitors","heading":"Promote competitors","section":"Documentation","crumbs":["Core concepts","Competitors","Promote competitors"],"url":"/docs/core-concepts/competitors#promote-competitors","text":"Genezio can find a relevant competitor that has a weak presence in the data. In this case, users can promote the brand. Then, the analysis uses it as a primary competitor.","keywords":""},{"kind":"section","title":"Competitors","heading":"Demote competitors","section":"Documentation","crumbs":["Core concepts","Competitors","Demote competitors"],"url":"/docs/core-concepts/competitors#demote-competitors","text":"Sometimes answer engines mention brands that are not real competitors. Users can demote these brands. Then, the competitive analysis gives them less importance.","keywords":""},{"kind":"section","title":"Competitors","heading":"Merge competitors","section":"Documentation","crumbs":["Core concepts","Competitors","Merge competitors"],"url":"/docs/core-concepts/competitors#merge-competitors","text":"Different names or variants of a name can refer to the same brand. For example, a brand can appear with small differences in its name in different responses. Genezio lets users merge competitor entries. Then, all mentions are under one brand. This keeps the analysis consistent across conversations.","keywords":""},{"kind":"section","title":"Competitors","heading":"Attach competitors to your brand","section":"Documentation","crumbs":["Core concepts","Competitors","Attach competitors to your brand"],"url":"/docs/core-concepts/competitors#attach-competitors-to-your-brand","text":"Sometimes a competitor that Genezio found is really a part of the analyzed brand. Examples: - Sub-brands - Acquired companies - Product lines of the same organization Users can attach these entities to their brand. Then, Genezio uses them as a part of the same brand, and not as a competitor.","keywords":""},{"kind":"section","title":"Competitors","heading":"Competitors in the analysis","section":"Documentation","crumbs":["Core concepts","Competitors","Competitors in the analysis"],"url":"/docs/core-concepts/competitors#competitors-in-the-analysis","text":"After Genezio finds competitors, it uses them in all of the platform to make insights. Competitor data is the base for metrics such as: - Comparisons of AI Recommendations and Visibility - Share of Voice - Competitive positioning in AI answers - Brand presence for each topic Thus, teams see how frequently their brand appears. They also see which brands control the conversation.","keywords":""},{"kind":"section","title":"Competitors","heading":"Competitors in the interface","section":"Documentation","crumbs":["Core concepts","Competitors","Competitors in the interface"],"url":"/docs/core-concepts/competitors#competitors-in-the-interface","text":"Competitors appear in many parts of the Genezio interface, such as: - Topic analysis views - Visibility comparisons - Conversation details - Citation analysis Users can examine how competitors appear in specific conversations and which sources mention them. Where a competitor appears, you can open its SWOT drawer in the same place. The drawer compares your brand directly with that competitor. It has an overview of the comparison that AI wrote. It also shows the products of the competitor and the products that answers mention together with them. Refer to SWOT Analysis. To test how answer engines compare your brand with selected competitors, use the Comparer agent.","keywords":""},{"kind":"section","title":"Competitors","heading":"Why competitor analysis is important","section":"Documentation","crumbs":["Core concepts","Competitors","Why competitor analysis is important"],"url":"/docs/core-concepts/competitors#why-competitor-analysis-is-important","text":"AI assistants have a larger and larger effect on how users find and compare products. When you know which brands appear together with your brand, you can answer questions such as: - Which competitors does AI recommend most frequently? - Which competitors control specific topics? - Which sources mention competitors but not your brand? These insights help organizations to find opportunities to increase their AI Recommendations and Visibility.","keywords":""},{"kind":"section","title":"Competitors","heading":"Next steps","section":"Documentation","crumbs":["Core concepts","Competitors","Next steps"],"url":"/docs/core-concepts/competitors#next-steps","text":"To see how Genezio changes competitor data into measurable insights of AI visibility, refer to: - Share of Voice","keywords":""},{"kind":"section","title":"Competitors","heading":"Competitors in the public API","section":"Documentation","crumbs":["Core concepts","Competitors","Competitors in the public API"],"url":"/docs/core-concepts/competitors#competitors-in-the-public-api","text":"The public API reads and changes competitors. Refer to List competitors.","keywords":""},{"kind":"page","title":"Knowledge Base","heading":"","section":"Documentation","crumbs":["Documentation","Core concepts","Knowledge Base"],"url":"/docs/core-concepts/knowledge-base","text":"The Knowledge Base is the source of truth of your brand in Genezio. Learn what to add to it and how Genezio uses it to verify what answer engines say about you. The Knowledge Base is the source of truth of your brand in Genezio. Genezio compares what answer engines say about your brand with the Knowledge Base, and it shows which claims are correct. The Knowledge Base is a set of documents, pages, and notes that you keep current. These are the items that you show when a person asks \"what is really true about your brand?\".","keywords":""},{"kind":"section","title":"Knowledge Base","heading":"Why a Knowledge Base exists","section":"Documentation","crumbs":["Core concepts","Knowledge Base","Why a Knowledge Base exists"],"url":"/docs/core-concepts/knowledge-base#why-a-knowledge-base-exists","text":"Answer engines make claims about brands all the time. Some claims are correct. Some claims are old. Some claims are completely wrong. Before, a person had to read each perception and compare it with an internal source to find the difference. The Knowledge Base does this comparison automatically. You give the truth one time. Then, the platform uses it as the reference each time it extracts a perception about your brand from a conversation with an answer engine. This is the base for an accuracy audit loop: 1. The platform records what answer engines say about you. 2. The Knowledge Base gives what is really true. 3. The platform compares the two, and you see where they agree and where they do not agree.","keywords":""},{"kind":"section","title":"Knowledge Base","heading":"Where to configure the Knowledge Base","section":"Documentation","crumbs":["Core concepts","Knowledge Base","Where to configure the Knowledge Base"],"url":"/docs/core-concepts/knowledge-base#where-to-configure-the-knowledge-base","text":"The Knowledge Base is in Brand Settings → Knowledge Base. You configure it one time for each brand. After that, it operates in the background. The platform automatically compares each new perception from conversations with the Knowledge Base.","keywords":""},{"kind":"section","title":"Knowledge Base","heading":"You start with one item","section":"Documentation","crumbs":["Core concepts","Knowledge Base","You start with one item"],"url":"/docs/core-concepts/knowledge-base#you-start-with-one-item","text":"Your Knowledge Base is not empty at the start. When you create a brand, Genezio automatically crawls the website of the brand and adds it as the first item of the Knowledge Base. Thus, the grounded-perception checks operate from day one. They use your public site as the initial source of truth, also when you did no other configuration. After that, add to the Knowledge Base the documents, URLs, and free-text snippets that your website does not cover.","keywords":""},{"kind":"section","title":"Knowledge Base","heading":"What types of input you can add","section":"Documentation","crumbs":["Core concepts","Knowledge Base","What types of input you can add"],"url":"/docs/core-concepts/knowledge-base#what-types-of-input-you-can-add","text":"The Knowledge Base accepts three types of input. Thus, you can use the format that your truth has today: - Documents: Upload datasheets, FAQs, positioning documents, internal one-pagers, or other files that have the truth about the brand. - URLs: Add pages that already show your official messages (your product pages, your about page, your pricing page, partner pages). - Free-text snippets: Paste facts, definitions, or correction notes that are not yet in a document (\"our pricing starts at $X\", \"we do not offer a free tier\", \"our HQ is in Berlin, not San Francisco\"). You can use all three types together. A usual Knowledge Base for a mature brand has these items: - Some authoritative documents. - Links to some canonical pages. - A list of free-text corrections that increases over time.","keywords":""},{"kind":"section","title":"Knowledge Base","heading":"What to put in the Knowledge Base","section":"Documentation","crumbs":["Core concepts","Knowledge Base","What to put in the Knowledge Base"],"url":"/docs/core-concepts/knowledge-base#what-to-put-in-the-knowledge-base","text":"Think about the claims that an answer engine must make correctly when it describes your brand: - Products and capabilities: What you sell and what each product really does. - Prices and plans: The current tiers, what they include, and what they do not include. - Target audience and use cases: The customers that you are for and the customers that you are not for. - Competitive differentiators: The messages that make you different. - Facts that answer engines get wrong: The things that answer engines already said incorrectly about you. This is a very good signal. Put here each item that you corrected in a sales call. - Official statements and brand guidelines: Positioning documents, key messages, taglines, and do-not-say lists. Quality is more important than quantity. A small, correct, and current Knowledge Base is more useful than a large and old one.","keywords":""},{"kind":"section","title":"Knowledge Base","heading":"How the platform uses the Knowledge Base","section":"Documentation","crumbs":["Core concepts","Knowledge Base","How the platform uses the Knowledge Base"],"url":"/docs/core-concepts/knowledge-base#how-the-platform-uses-the-knowledge-base","text":"When you have a Knowledge Base, it supplies two functions in the platform today.","keywords":""},{"kind":"section","title":"Knowledge Base","heading":"Grounded perceptions","section":"Documentation","crumbs":["Core concepts","Knowledge Base","Grounded perceptions"],"url":"/docs/core-concepts/knowledge-base#grounded-perceptions","text":"The platform compares each perception from a conversation with an answer engine with your Knowledge Base. The perceptions that the platform can verify get a grounded badge with one of three states: - Grounded-correct: The perception agrees with your Knowledge Base. - Grounded-incorrect: The perception does not agree with your Knowledge Base. - Not grounded: The platform could not verify the perception with your Knowledge Base. This badge shows the accuracy audit. To see how grounded perceptions appear in the platform, refer to Perceptions.","keywords":""},{"kind":"section","title":"Knowledge Base","heading":"Reports and insights","section":"Documentation","crumbs":["Core concepts","Knowledge Base","Reports and insights"],"url":"/docs/core-concepts/knowledge-base#reports-and-insights","text":"The reports of the platform use the grounded perceptions. You can see the percentage of perceptions about your brand that are correct, incorrect, or that the platform cannot verify. You can also examine the specific perceptions that are wrong. Then, you can correct the source: your website, your documents, or the Knowledge Base. The Knowledge Base is also a source of claims for the Fact Checker agent. The Fact Checker tests if answer engines confirm these claims when a user asks about them directly.","keywords":""},{"kind":"section","title":"Knowledge Base","heading":"Strength of the Knowledge Base","section":"Documentation","crumbs":["Core concepts","Knowledge Base","Strength of the Knowledge Base"],"url":"/docs/core-concepts/knowledge-base#strength-of-the-knowledge-base","text":"Genezio crawls the website of each brand at the start. Thus, the grounded layer is active from day one. But the strength of the signal depends on how much truth your Knowledge Base has. If your website is detailed and current, the initial crawl gives much value. Your website can have little content or mostly marketing content. It can also have no prices, specifications, or positioning details. In these cases, the platform can verify fewer perceptions, and you see more \"not grounded\" badges. Then, you must extend the Knowledge Base. Add the documents, URLs, and snippets that fill the gaps of your website.","keywords":""},{"kind":"section","title":"Knowledge Base","heading":"Maintain the Knowledge Base","section":"Documentation","crumbs":["Core concepts","Knowledge Base","Maintain the Knowledge Base"],"url":"/docs/core-concepts/knowledge-base#maintain-the-knowledge-base","text":"The Knowledge Base is a living asset. Use it as a brand book that you edit: - Update it when your brand changes: For example, a new product, new prices, or a new positioning. - Add to it when you find an incorrect perception: Answer engines can say an untrue thing again and again. Write the correction in the Knowledge Base. Then, the platform can flag the next occurrences. - Remove old documents: Old truth is worse than no truth. When the Knowledge Base is more current and correct, you can trust the grounded signal more.","keywords":""},{"kind":"section","title":"Knowledge Base","heading":"Related pages","section":"Documentation","crumbs":["Core concepts","Knowledge Base","Related pages"],"url":"/docs/core-concepts/knowledge-base#related-pages","text":"- Perceptions - Fact Checker agent - Brands - Competitors","keywords":""},{"kind":"section","title":"Knowledge Base","heading":"The Knowledge Base in the public API","section":"Documentation","crumbs":["Core concepts","Knowledge Base","The Knowledge Base in the public API"],"url":"/docs/core-concepts/knowledge-base#the-knowledge-base-in-the-public-api","text":"The public API reads and writes the knowledge bases of a brand. Refer to List the knowledge bases.","keywords":""},{"kind":"page","title":"Master Filters","heading":"","section":"Documentation","crumbs":["Documentation","Core concepts","Master Filters"],"url":"/docs/core-concepts/master-filters","text":"Master Filters split the AI visibility data of one brand by product line. Learn how they work, how to configure them, and when to use separate brands instead. Master Filters divide the data of one brand by product line in Genezio. Thus, a company with many products can stay as one brand and also see the results of each product line separately. A master filter is a set of topics. You create master filters and select one in the picker. Then, each view in Genezio shows only the topics of that filter: dashboards, conversations, insights, citations, share of voice, perceptions, and all other views. The old name of Master Filters is Products. The name changed because users confused them with the products that AI mentions in shopping answers. That is a fully different feature. Refer to Products.","keywords":""},{"kind":"section","title":"Master Filters","heading":"Master Filters or separate brands","section":"Documentation","crumbs":["Core concepts","Master Filters","Master Filters or separate brands"],"url":"/docs/core-concepts/master-filters#master-filters-or-separate-brands","text":"Genezio supports two methods to model a company with many products. It is important to select the correct method.","keywords":""},{"kind":"section","title":"Master Filters","heading":"Use separate brands when","section":"Documentation","crumbs":["Core concepts","Master Filters","Use separate brands when"],"url":"/docs/core-concepts/master-filters#use-separate-brands-when","text":"Your products have their own brand presence in answer engines: - Answer engines know the products well by their product names. - Customers and buyers use the product names directly. - Each product has its own different competitors. In this case, make a separate brand for each product. For an answer engine, Microsoft Word, Microsoft Excel, and Microsoft Teams are different brands. Each one has its own competitors (Google Docs, Google Sheets, Slack), its own buyer personas, and its own visibility results. Refer to Companies with many products in Brands.","keywords":""},{"kind":"section","title":"Master Filters","heading":"Use Master Filters when","section":"Documentation","crumbs":["Core concepts","Master Filters","Use Master Filters when"],"url":"/docs/core-concepts/master-filters#use-master-filters-when","text":"Your brand is strong, but your product lines do not have their own brand recognition: - Customers know the name of the company, not the names of the product lines. - Answer engines describe the lines as offers of the parent brand, not as independent names. - The lines frequently have the same competitors, or the competitors are not very different for each line. - You want one source of truth at the brand level, with a view of each product line above it. In this case, keep all data at the brand level. Define Master Filters to filter the data for each product line.","keywords":""},{"kind":"section","title":"Master Filters","heading":"Quick decision","section":"Documentation","crumbs":["Core concepts","Master Filters","Quick decision"],"url":"/docs/core-concepts/master-filters#quick-decision","text":"Each product has its own brand recognition and different competitors: Separate brands. The parent brand is strong, and the product lines do not have an independent brand presence: Master Filters (in one brand). You are not sure: Start with one brand and Master Filters. It is easy to divide the brand later. It is more difficult to merge brands later.","keywords":""},{"kind":"section","title":"Master Filters","heading":"How Master Filters operate","section":"Documentation","crumbs":["Core concepts","Master Filters","How Master Filters operate"],"url":"/docs/core-concepts/master-filters#how-master-filters-operate","text":"A master filter is a named set of topics. That is the full definition. - More than one master filter can have the same topic. For example, the topic \"Customer support best practices\" can be in your Sales filter and in your Service filter at the same time. - A topic that is not in a master filter stays on the brand. It does not appear when a filter is active. - You manage master filters at the brand level. Each brand has its own list.","keywords":""},{"kind":"section","title":"Master Filters","heading":"Where to configure Master Filters","section":"Documentation","crumbs":["Core concepts","Master Filters","Where to configure Master Filters"],"url":"/docs/core-concepts/master-filters#where-to-configure-master-filters","text":"You define Master Filters in Settings. Give each master filter a name, and select its topics. When the brand has one or more master filters, a picker appears in the main header. With the picker, you change the full view from one filter to another, or back to the default view without a filter.","keywords":""},{"kind":"section","title":"Master Filters","heading":"Filter your reports by master filter and persona","section":"Documentation","crumbs":["Core concepts","Master Filters","Filter your reports by master filter and persona"],"url":"/docs/core-concepts/master-filters#filter-your-reports-by-master-filter-and-persona","text":"The filters are in the report header, and they stay active from one screen to the next. When you make a selection on one page, the next page keeps it. Thus, you do not have to apply it again on each page. The report header has two filters: - Master filter: Appears when the brand has one or more master filters. When you select one, each view shows only the topics of that filter. - Persona: Appears when the topics of the brand are for two or more personas. When you select one, the view shows only the data of that persona. Refer to Personas. If you did not define master filters, Genezio does not show that dropdown, and Genezio operates as before. When you select a master filter: - Dashboards show the metrics of the topics in that filter. - Conversations show only the topics of that filter. - Insights use only the data of that filter. - Citations, SOV, perceptions, and competitors: each view shows only the data of that filter. The default view has no filter. It shows the full brand, and no master filter has an effect on it. The picker is an optional lens that you can select.","keywords":""},{"kind":"section","title":"Master Filters","heading":"Typical configuration: one company with three product lines","section":"Documentation","crumbs":["Core concepts","Master Filters","Typical configuration: one company with three product lines"],"url":"/docs/core-concepts/master-filters#typical-configuration-one-company-with-three-product-lines","text":"For example, a consumer-electronics company has three product lines: The company creates one brand for itself. In Settings, the company defines three master filters: Headphones, Speakers, and Wearables. The company attaches the topics for all three lines (for example, \"Brand reputation\") to all three filters. Or, it keeps them at the brand level if the leadership wants to see them for the full company. The topics for one line (for example, \"Noise-cancelling headphones for travel\") are only in Headphones. The picker appears in the header. The marketing manager of Headphones uses it each day to see only that product line. The CMO keeps the view without a filter to see the full company.","keywords":""},{"kind":"section","title":"Master Filters","heading":"Master Filters and the other parts of the platform","section":"Documentation","crumbs":["Core concepts","Master Filters","Master Filters and the other parts of the platform"],"url":"/docs/core-concepts/master-filters#master-filters-and-the-other-parts-of-the-platform","text":"The picker filters the data at the data level. Thus, Master Filters operate correctly with the other parts of Genezio: - The Geo Assistant uses the current master filter when it answers questions. - Insights and Share of Voice use the active filter. - When you export reports for reviews with stakeholders, the export uses the filter context. - Topic Tags continue to operate independently. One topic can have Headphones (master filter) and Pricing (tag) at the same time.","keywords":""},{"kind":"section","title":"Master Filters","heading":"Master Filters and shopping products are different","section":"Documentation","crumbs":["Core concepts","Master Filters","Master Filters and shopping products are different"],"url":"/docs/core-concepts/master-filters#master-filters-and-shopping-products-are-different","text":"Master Filters put topics into groups to divide your reports. They are a reporting lens that you configure. The Shopping section is about a different thing. It shows the products that answer engines really mention in shopping answers. Genezio finds these products automatically in conversations. You do not configure these products.","keywords":""},{"kind":"section","title":"Master Filters","heading":"Related pages","section":"Documentation","crumbs":["Core concepts","Master Filters","Related pages"],"url":"/docs/core-concepts/master-filters#related-pages","text":"- Brands - Topics - Topic Tags - Competitors - Product Visibility","keywords":""},{"kind":"page","title":"Prompter agent","heading":"","section":"Documentation","crumbs":["Documentation","Agents","Prompter agent"],"url":"/docs/genezio-agents/prompter-agent","text":"The Prompter agent sends one prompt to an answer engine, exactly as you wrote it. Learn when to use this single-turn test and how it counts for visibility. The Prompter agent is the Genezio agent type that sends one prompt to an answer engine and records one answer. It tests direct discovery questions, in which a user asks one clear question and gets one answer.","keywords":""},{"kind":"section","title":"Prompter agent","heading":"How the Prompter agent works","section":"Documentation","crumbs":["Agents","Prompter agent","How the Prompter agent works"],"url":"/docs/genezio-agents/prompter-agent#how-the-prompter-agent-works","text":"For a Prompter agent topic: - Genezio sends the prompt to the answer engine exactly as you wrote it. - Genezio does not make follow-up prompts. Prompter agent topics use prompts, not scenarios. The other agent types use scenarios. A prompt is a direct question, of the type that a real user types into an AI assistant. Example prompt: Genezio sends this text directly to the answer engine as one message.","keywords":"What CRM should a startup with a small sales team use to track leads and follow up with customers?"},{"kind":"section","title":"Prompter agent","heading":"Typical use cases for single-turn prompts","section":"Documentation","crumbs":["Agents","Prompter agent","Typical use cases for single-turn prompts"],"url":"/docs/genezio-agents/prompter-agent#typical-use-cases-for-single-turn-prompts","text":"Prompter agent conversations are useful when you want to measure these items: - Which brands the answers mention for direct discovery questions - If your brand occurs in the first recommendations - How different answer engines answer the same prompt","keywords":""},{"kind":"section","title":"Prompter agent","heading":"Effect of the Prompter on Brand Visibility","section":"Documentation","crumbs":["Agents","Prompter agent","Effect of the Prompter on Brand Visibility"],"url":"/docs/genezio-agents/prompter-agent#effect-of-the-prompter-on-brand-visibility","text":"Genezio includes Prompter agent conversations in Brand Visibility and in the related visibility KPIs. The reason is that the answer engine must select brands organically. The prompt does not include brand names. Genezio does not use Prompter conversations for AI Recommendations, because the prompt does not ask for a recommendation. Refer to Brand Recommendation.","keywords":""},{"kind":"section","title":"Prompter agent","heading":"Why one turn gives a clean visibility baseline","section":"Documentation","crumbs":["Agents","Prompter agent","Why one turn gives a clean visibility baseline"],"url":"/docs/genezio-agents/prompter-agent#why-one-turn-gives-a-clean-visibility-baseline","text":"Prompter agent conversations have one turn. Thus: - You can easily compare them across answer engines. - They have less variability, because there are no follow-up prompts. - They give a clean baseline to track visibility over time.","keywords":""},{"kind":"section","title":"Prompter agent","heading":"Related pages","section":"Documentation","crumbs":["Agents","Prompter agent","Related pages"],"url":"/docs/genezio-agents/prompter-agent#related-pages","text":"- Scenarios - Brand visibility - Recommender agent","keywords":""},{"kind":"section","title":"Prompter agent","heading":"Prompter data in the public API","section":"Documentation","crumbs":["Agents","Prompter agent","Prompter data in the public API"],"url":"/docs/genezio-agents/prompter-agent#prompter-data-in-the-public-api","text":"The public API lists the topics of a brand: List topics.","keywords":""},{"kind":"page","title":"Recommender agent","heading":"","section":"Documentation","crumbs":["Documentation","Agents","Recommender agent"],"url":"/docs/genezio-agents/recommender-agent","text":"The Recommender agent simulates a buyer who asks answer engines for recommendations over many turns. Learn how it works and how it counts for the KPIs. The Recommender agent is the Genezio agent type that simulates a user who asks answer engines for recommendations in a multi-step conversation. Use it for scenarios where a persona looks for a solution that matches their needs and constraints.","keywords":""},{"kind":"section","title":"Recommender agent","heading":"How the Recommender agent works","section":"Documentation","crumbs":["Agents","Recommender agent","How the Recommender agent works"],"url":"/docs/genezio-agents/recommender-agent#how-the-recommender-agent-works","text":"For a Recommender agent topic: - The scenario defines the situation and the goal of the user. - Genezio makes a sequence of prompts from that scenario. - The AI conversation runs across many turns. Example scenario: From this scenario, Genezio makes conversational prompts such as these: User query: I run a small startup and we've been tracking leads in a spreadsheet, but it's not working anymore. What CRM tools would you recommend for a team of three? User query: We need something that integrates with Gmail and stays under $50 per user per month. Which of those would fit?","keywords":"John's team of three has been tracking leads in a spreadsheet but they are losing follow-ups as inbound volume grows. They need a simple CRM that integrates with Gmail and costs under $50/user/month."},{"kind":"section","title":"Recommender agent","heading":"Typical use cases for recommendation tests","section":"Documentation","crumbs":["Agents","Recommender agent","Typical use cases for recommendation tests"],"url":"/docs/genezio-agents/recommender-agent#typical-use-cases-for-recommendation-tests","text":"Recommender agent conversations are useful when you want to examine these items: - The recommendation frequency of your brand - The position of your brand, when compared with competitors - How the recommendations change after the user adds constraints in follow-up prompts","keywords":""},{"kind":"section","title":"Recommender agent","heading":"Effect of the Recommender on the KPIs","section":"Documentation","crumbs":["Agents","Recommender agent","Effect of the Recommender on the KPIs"],"url":"/docs/genezio-agents/recommender-agent#effect-of-the-recommender-on-the-kpis","text":"Recommender agent conversations are a core input to these KPIs: - AI Recommendations: of the Recommender conversations in which your brand is visible, the percentage in which the answer engine recommends your brand. - Brand Visibility: Genezio includes these conversations because the answer engine selects the brands. The words of the prompt do not force the brands.","keywords":""},{"kind":"section","title":"Recommender agent","heading":"Multi-step behavior and realistic buyer journeys","section":"Documentation","crumbs":["Agents","Recommender agent","Multi-step behavior and realistic buyer journeys"],"url":"/docs/genezio-agents/recommender-agent#multi-step-behavior-and-realistic-buyer-journeys","text":"The multi-step flow makes this agent useful for realistic buyer journeys: - A first, wide recommendation - A filter by budget, features, or team size - A shortlist of alternatives","keywords":""},{"kind":"section","title":"Recommender agent","heading":"Related pages","section":"Documentation","crumbs":["Agents","Recommender agent","Related pages"],"url":"/docs/genezio-agents/recommender-agent#related-pages","text":"- Conversations - Brand Recommendation - Prompter agent","keywords":""},{"kind":"section","title":"Recommender agent","heading":"Recommender data in the public API","section":"Documentation","crumbs":["Agents","Recommender agent","Recommender data in the public API"],"url":"/docs/genezio-agents/recommender-agent#recommender-data-in-the-public-api","text":"The public API lists the scenarios and the conversations of a brand: List scenarios and List conversations.","keywords":""},{"kind":"page","title":"Introspector agent","heading":"","section":"Documentation","crumbs":["Documentation","Agents","Introspector agent"],"url":"/docs/genezio-agents/introspector-agent","text":"The Introspector agent examines how answer engines describe a specific brand. Learn when to add a persona and why it does not count for AI visibility. The Introspector agent is the Genezio agent type that examines how answer engines describe a specific brand. Use it to analyze the explanation and the position of a brand. Do not use it to measure organic visibility.","keywords":""},{"kind":"section","title":"Introspector agent","heading":"How the Introspector agent works","section":"Documentation","crumbs":["Agents","Introspector agent","How the Introspector agent works"],"url":"/docs/genezio-agents/introspector-agent#how-the-introspector-agent-works","text":"For an Introspector agent topic: - The scenario explicitly includes the target brand. - A persona is optional. Refer to the next section. - Genezio runs a multi-step conversation about that brand. - The prompts examine the use cases, the fit, the strengths, and the limits of the brand.","keywords":""},{"kind":"section","title":"Introspector agent","heading":"The persona is optional","section":"Documentation","crumbs":["Agents","Introspector agent","The persona is optional"],"url":"/docs/genezio-agents/introspector-agent#the-persona-is-optional","text":"Introspector topics can run with or without a persona. The Introspector asks this question: \"What does the answer engine say about us when a user asks directly?\" The brand name is already in the prompt. Thus, you do not always need a persona to get a useful answer. Without a persona, the configuration has one step less. This is useful in the usual case, when you only want to know how answer engines describe your brand. Add a persona to an Introspector topic in these cases: - You want to know how the description changes for a specific region, language, or buyer profile. - You audit the position of the brand for a specific audience segment. Do not add a persona in these cases: - You want a baseline of how answer engines describe your brand, with no audience overlay. - You want to configure the Introspector quickly, and you do not have a relevant persona yet. Thus, the Introspector is the agent with the easiest configuration. You need only a topic and the brand name. Then you can run it. Example scenario: From this scenario, Genezio makes conversational prompts such as these: User query: What is HubSpot mainly used for? What kind of teams typically use it? User query: What are HubSpot's main strengths compared to other tools in its category? User query: Are there any common limitations or complaints about HubSpot?","keywords":"Sarah has heard about HubSpot from a colleague and wants to understand what it actually does before evaluating it. She wants to know what types of teams use it what its main strengths are and whether it has any notable limitations."},{"kind":"section","title":"Introspector agent","heading":"Typical use cases for brand description audits","section":"Documentation","crumbs":["Agents","Introspector agent","Typical use cases for brand description audits"],"url":"/docs/genezio-agents/introspector-agent#typical-use-cases-for-brand-description-audits","text":"Introspector agent conversations help teams analyze these items: - How AI systems explain the category and the value of a brand - The strengths or the weaknesses that occur again and again in brand narratives - The consistency of the brand description across answer engines","keywords":""},{"kind":"section","title":"Introspector agent","heading":"Effect of the Introspector on the visibility KPIs","section":"Documentation","crumbs":["Agents","Introspector agent","Effect of the Introspector on the visibility KPIs"],"url":"/docs/genezio-agents/introspector-agent#effect-of-the-introspector-on-the-visibility-kpis","text":"Genezio does not include Introspector agent conversations in the Brand Visibility calculations. The reason is that the brand name is already in the prompt. In the visibility calculations, these conversations make the results too high.","keywords":""},{"kind":"section","title":"Introspector agent","heading":"Why the Introspector is still important","section":"Documentation","crumbs":["Agents","Introspector agent","Why the Introspector is still important"],"url":"/docs/genezio-agents/introspector-agent#why-the-introspector-is-still-important","text":"The visibility KPIs do not include this agent. But the agent is valuable for these tasks: - The review of the narrative quality - Positioning audits - The analysis of the message consistency To see the claims that answer engines make in these conversations, refer to Perceptions. For a summary of the full narrative, refer to AI Perception Summary.","keywords":""},{"kind":"section","title":"Introspector agent","heading":"Related pages","section":"Documentation","crumbs":["Agents","Introspector agent","Related pages"],"url":"/docs/genezio-agents/introspector-agent#related-pages","text":"- Perceptions - Brand visibility - Conversations - Fact Checker agent","keywords":""},{"kind":"section","title":"Introspector agent","heading":"Introspector data in the public API","section":"Documentation","crumbs":["Agents","Introspector agent","Introspector data in the public API"],"url":"/docs/genezio-agents/introspector-agent#introspector-data-in-the-public-api","text":"The public API reads what answer engines say about a brand: Perceptions.","keywords":""},{"kind":"page","title":"Comparer agent","heading":"","section":"Documentation","crumbs":["Documentation","Agents","Comparer agent"],"url":"/docs/genezio-agents/comparer-agent","text":"The Comparer agent tests how answer engines compare your brand with specific competitors. Learn how it works, what SWOT data it gives, and its KPI effect. The Comparer agent is the Genezio agent type that examines how answer engines compare your brand with specific competitors in the same conversation. Use it to analyze your competitive position in AI answers.","keywords":""},{"kind":"section","title":"Comparer agent","heading":"How the Comparer agent works","section":"Documentation","crumbs":["Agents","Comparer agent","How the Comparer agent works"],"url":"/docs/genezio-agents/comparer-agent#how-the-comparer-agent-works","text":"For a Comparer agent topic: - You select the exact competitors for the comparison at the topic level. - The scenario names your brand and those competitors. - Genezio runs a multi-step comparison conversation. - The prompts ask for the differences, the tradeoffs, and the best-fit recommendations. You select the competitors for each topic. Thus, the Comparer scenarios that Genezio makes automatically target the brands that are important for that subject area. They do not target a random competitor that occurs in AI answers. To manage the competitor list, refer to Competitors. Example scenario: From this scenario, Genezio makes conversational prompts such as these: User query: I'm comparing HubSpot and Pipedrive for a sales team of five. What are the main differences between the two? User query: Which one is easier to onboard for a small team with no CRM experience? User query: How do they compare on reporting and email integration?","keywords":"Alex is evaluating HubSpot and Pipedrive for his sales team of five. He needs a CRM that is easy to onboard has good reporting and works well with their existing email workflow. He wants to understand the key differences before making a decision."},{"kind":"section","title":"Comparer agent","heading":"Typical use cases for brand comparisons","section":"Documentation","crumbs":["Agents","Comparer agent","Typical use cases for brand comparisons"],"url":"/docs/genezio-agents/comparer-agent#typical-use-cases-for-brand-comparisons","text":"Comparer agent conversations help teams understand these items: - Which competitor the AI prefers in direct comparisons - How AI systems describe the strengths and the weaknesses of each brand - Which objections or limits the AI mentions again and again","keywords":""},{"kind":"section","title":"Comparer agent","heading":"What you learn from the Comparer: SWOT and side-by-side KPIs","section":"Documentation","crumbs":["Agents","Comparer agent","What you learn from the Comparer: SWOT and side-by-side KPIs"],"url":"/docs/genezio-agents/comparer-agent#what-you-learn-from-the-comparer-swot-and-side-by-side-kpis","text":"With the Comparer, you can do these tasks: - See how answer engines compare your brand with specific competitors. - Compare the recommendation rates and the visibility rates side by side. - Identify the strengths and the weaknesses that the AI connects to each brand. - Make SWOT insights that guide your competitive content strategy. Genezio automatically extracts a SWOT analysis from each Comparer conversation. It shows the SWOT in the conversation drawer. Genezio also collects the same SWOT at the scenario level and at the topic level. To see the SWOT Comparison view, go to Competitors SWOT in the menu. This view shows the SWOTs of your brand and of all the tracked competitors side by side. For the full details, read SWOT analysis. The result is a clear view of where you win, where you lose, and which narratives cause the difference.","keywords":""},{"kind":"section","title":"Comparer agent","heading":"Effect of the Comparer on the KPIs","section":"Documentation","crumbs":["Agents","Comparer agent","Effect of the Comparer on the KPIs"],"url":"/docs/genezio-agents/comparer-agent#effect-of-the-comparer-on-the-kpis","text":"Comparer agent conversations do not count for AI Recommendations or AI Visibility. The brand names are in the prompt. Thus, in the metrics, these conversations give incorrect results. The Comparer is a competitive analysis tool. It uses the brand-level KPIs that Genezio calculated from the Prompter and Recommender conversations. It shows these KPIs together with direct AI conversations between brands. Thus, it shows how your brand compares with specific competitors.","keywords":""},{"kind":"section","title":"Comparer agent","heading":"Related pages","section":"Documentation","crumbs":["Agents","Comparer agent","Related pages"],"url":"/docs/genezio-agents/comparer-agent#related-pages","text":"- SWOT analysis - Competitors - Brand visibility - Share of Voice - Running conversations","keywords":""},{"kind":"section","title":"Comparer agent","heading":"Comparer data in the public API","section":"Documentation","crumbs":["Agents","Comparer agent","Comparer data in the public API"],"url":"/docs/genezio-agents/comparer-agent#comparer-data-in-the-public-api","text":"The public API reads the SWOT of the brand and its competitors: The SWOT of the brand and its competitors. It also lists the competitors: List competitors.","keywords":""},{"kind":"page","title":"Fact Checker agent","heading":"","section":"Documentation","crumbs":["Documentation","Agents","Fact Checker agent"],"url":"/docs/genezio-agents/fact-checker-agent","text":"The Fact Checker agent tests if answer engines confirm specific claims about your brand. Learn where claims come from and how to read the results. The Fact Checker agent tests if answer engines confirm specific claims about your brand, such as prices, features, or certifications. It is the third main axis of measurement in Genezio, together with Visibility and Recommendation. Visibility asks how frequently you occur. Recommendation asks how frequently the AI selects you. The Fact Checker asks a different question: When answer engines talk about your brand, are their facts correct? You give specific claims about your brand that a person can verify. Examples are pricing, features, founding year, customer counts, certifications, and all other items that can be true or false. The platform tests the claims on each supported answer engine.","keywords":""},{"kind":"section","title":"Fact Checker agent","heading":"Why AI fact checking is necessary","section":"Documentation","crumbs":["Agents","Fact Checker agent","Why AI fact checking is necessary"],"url":"/docs/genezio-agents/fact-checker-agent#why-ai-fact-checking-is-necessary","text":"Answer engines hallucinate. They mix old information with current information. They confuse a brand with similar brands. For most brands, this is a small problem. For some brands, it is a real risk. Examples are security companies, financial services, regulated industries, healthcare, and all companies whose product details must be accurate. Before the Fact Checker, you did not have a method to monitor in a systematic way what answer engines say about specific facts. You learned about a hallucinated claim only when a customer mentioned it, or when a person asked the correct question by chance. The Fact Checker changes this into a measurement. You make a list of the claims that are important to you. The platform gives a result for each claim on each answer engine. The answer engine confirms the claim, contradicts it, or does not give an answer about it.","keywords":""},{"kind":"section","title":"Fact Checker agent","heading":"How the Fact Checker agent works","section":"Documentation","crumbs":["Agents","Fact Checker agent","How the Fact Checker agent works"],"url":"/docs/genezio-agents/fact-checker-agent#how-the-fact-checker-agent-works","text":"A Fact Checker topic has a set of claims to verify. These are claims that you say are true about your brand. For each claim, the platform runs conversations with each supported answer engine. It puts the answer into one of three results: - True: the answer engine confirms the claim. - False: the answer engine contradicts the claim. - Indecisive: the answer engine does not take a clear position. The result applies to one claim and one answer engine. Thus, ChatGPT can confirm a claim, Perplexity can contradict it, and Claude can ignore it. The platform shows all three results.","keywords":""},{"kind":"section","title":"Fact Checker agent","heading":"Where the claims come from","section":"Documentation","crumbs":["Agents","Fact Checker agent","Where the claims come from"],"url":"/docs/genezio-agents/fact-checker-agent#where-the-claims-come-from","text":"You can add claims to a Fact Checker topic in two ways.","keywords":""},{"kind":"section","title":"Fact Checker agent","heading":"Claims from the Knowledge Base","section":"Documentation","crumbs":["Agents","Fact Checker agent","Claims from the Knowledge Base"],"url":"/docs/genezio-agents/fact-checker-agent#claims-from-the-knowledge-base","text":"Get claims from the Knowledge Base of your brand. The Knowledge Base has the documents, URLs, and snippets that are already your source of truth. This is the usual method after you add content to your Knowledge Base. The truth is already in the documents, and the Fact Checker only measures if answer engines show it.","keywords":""},{"kind":"section","title":"Fact Checker agent","heading":"Custom claims","section":"Documentation","crumbs":["Agents","Fact Checker agent","Custom claims"],"url":"/docs/genezio-agents/fact-checker-agent#custom-claims","text":"Type claims directly on the Fact Checker topic. Custom claims are useful in these cases: - One-time measurements, for example: \"does ChatGPT know we acquired Company X last quarter?\" - Tests of claims that are not in the Knowledge Base yet - The monitoring of specific high-risk facts, for example a certification, a regulatory status, or a pricing detail You can use the two methods on one topic. Some claims can come from the Knowledge Base, and you can type other claims directly.","keywords":""},{"kind":"section","title":"Fact Checker agent","heading":"Where you see the Fact Checker results","section":"Documentation","crumbs":["Agents","Fact Checker agent","Where you see the Fact Checker results"],"url":"/docs/genezio-agents/fact-checker-agent#where-you-see-the-fact-checker-results","text":"","keywords":""},{"kind":"section","title":"Fact Checker agent","heading":"On the conversation","section":"Documentation","crumbs":["Agents","Fact Checker agent","On the conversation"],"url":"/docs/genezio-agents/fact-checker-agent#on-the-conversation","text":"Each conversation has a Fact Checker tab. The tab shows which claims the answer engine stated and which claims it did not state. Each card has a clear CLAIMED or NOT CLAIMED label. Thus, you can quickly examine one conversation.","keywords":""},{"kind":"section","title":"Fact Checker agent","heading":"On the topic and scenario drawers","section":"Documentation","crumbs":["Agents","Fact Checker agent","On the topic and scenario drawers"],"url":"/docs/genezio-agents/fact-checker-agent#on-the-topic-and-scenario-drawers","text":"Open the View More drawer on a topic or on a scenario. The drawer shows the Fact Checker charts. One chart is \"Claimed metric by answer engine\". It shows which answer engines support your claims and which answer engines do not. Use this view for reports to management. It gives a snapshot of which facts the answer engines confirm.","keywords":""},{"kind":"section","title":"Fact Checker agent","heading":"Effect of the Fact Checker on the KPIs","section":"Documentation","crumbs":["Agents","Fact Checker agent","Effect of the Fact Checker on the KPIs"],"url":"/docs/genezio-agents/fact-checker-agent#effect-of-the-fact-checker-on-the-kpis","text":"Fact Checker conversations do not count for AI Visibility or AI Recommendations. The prompt names the brand and the claims explicitly. Thus, the conversations do not measure organic discovery. They measure accuracy. In the KPIs, these conversations make the metrics incorrect. The Fact Checker is a separate measurement tool, as are the Introspector and the Comparer. It is separate from visibility and recommendation, but it is equally useful for the correct brand.","keywords":""},{"kind":"section","title":"Fact Checker agent","heading":"The accuracy stack: Knowledge Base, perceptions, and Fact Checker","section":"Documentation","crumbs":["Agents","Fact Checker agent","The accuracy stack: Knowledge Base, perceptions, and Fact Checker"],"url":"/docs/genezio-agents/fact-checker-agent#the-accuracy-stack-knowledge-base-perceptions-and-fact-checker","text":"The Fact Checker is the last part of the accuracy stack of Genezio: 1. Knowledge Base: defines what is true about your brand. 2. Grounded Perceptions: show if the claims from organic conversations match your Knowledge Base. 3. Fact Checker: tests if answer engines confirm specific claims when a user asks about them. Together, these give you a complete view of accuracy: - The Knowledge Base is the source of truth. - Grounded Perceptions measure what answer engines say about you without a question about the claim. - The Fact Checker measures what answer engines say when a user asks directly. Grounded Perceptions observe. The Fact Checker probes. When you read them together, you know the narrative that answer engines give without a request. You also know the answers that they give when a user asks directly.","keywords":""},{"kind":"section","title":"Fact Checker agent","heading":"When to use the Fact Checker","section":"Documentation","crumbs":["Agents","Fact Checker agent","When to use the Fact Checker"],"url":"/docs/genezio-agents/fact-checker-agent#when-to-use-the-fact-checker","text":"The Fact Checker is most valuable for brands where factual accuracy has a direct effect on buyer trust or regulatory standing: - Security and financial services: incorrect statements about certifications, compliance, or features can decrease trust. - Healthcare and regulated industries: claims about products have legal weight. - Brands that answer engines frequently confuse with competitors: answer engines mix your facts with the facts of another brand. - Brands after a major change, for example new pricing, an acquisition, or a rebrand: old information stays in answer engines for months. - All brands that are concerned about hallucinated misinformation: the Fact Checker changes that concern into a measured signal.","keywords":""},{"kind":"section","title":"Fact Checker agent","heading":"Typical Fact Checker workflow","section":"Documentation","crumbs":["Agents","Fact Checker agent","Typical Fact Checker workflow"],"url":"/docs/genezio-agents/fact-checker-agent#typical-fact-checker-workflow","text":"Add content to the Knowledge Base, or make sure that it is ready. This is the source of truth for the measurement. Attach claims from the Knowledge Base, custom claims, or both. The conversations run on each supported answer engine. Use the conversation tab for the individual results. Use the charts in the topic drawer or the scenario drawer for summaries across answer engines. Correct the public content: your website, your docs, and third-party listings. Then answer engines can get the correct information the next time. Answer engines change. A claim that an answer engine confirms today can become a contradiction tomorrow.","keywords":""},{"kind":"section","title":"Fact Checker agent","heading":"Related pages","section":"Documentation","crumbs":["Agents","Fact Checker agent","Related pages"],"url":"/docs/genezio-agents/fact-checker-agent#related-pages","text":"- Knowledge Base - Perceptions - Prompter agent - Recommender agent - Introspector agent - Comparer agent","keywords":""},{"kind":"section","title":"Fact Checker agent","heading":"Fact check data in the public API","section":"Documentation","crumbs":["Agents","Fact Checker agent","Fact check data in the public API"],"url":"/docs/genezio-agents/fact-checker-agent#fact-check-data-in-the-public-api","text":"The public API reads the fact check results over time, by topic, and by answer engine: Fact check over time, Fact check by topic, and Fact check by engine.","keywords":""},{"kind":"page","title":"Analysis","heading":"","section":"Documentation","crumbs":["Documentation","Running conversations","Analysis"],"url":"/docs/analysis","text":"See how Genezio makes its data: the questions it asks, the answer engines it asks, and the citations, competitors, and sentiment it extracts from each answer. The Analysis section tells how Genezio makes its data: what Genezio asks, where Genezio asks it, and what Genezio reads from the answers. To start, set up scenarios and answer engines. Then run conversations. From each answer, Genezio extracts these items: - Citations - Competitors - Query fanouts - Sentiment Most users must read Understanding conversation results first. Scores tell you where to look. Answers tell you why, and the work starts with the reason.","keywords":""},{"kind":"page","title":"Create topics and scenarios","heading":"","section":"Documentation","crumbs":["Documentation","Running conversations","Create topics and scenarios"],"url":"/docs/analysis/creating-scenarios","text":"Learn how to create topics and scenarios in Genezio, select the agent type of each topic, and write realistic scenarios that answer engines respond to. This guide tells how to create and manage topics and scenarios in Genezio. Genezio uses the scenarios when it runs conversations with answer engines, and good scenarios make your AI visibility data more accurate. A good scenario simulates the situation of a real user. Thus, the answer engine gives the same answer that a real buyer gets.","keywords":""},{"kind":"section","title":"Create topics and scenarios","heading":"Open the topics and scenarios in the settings","section":"Documentation","crumbs":["Running conversations","Create topics and scenarios","Open the topics and scenarios in the settings"],"url":"/docs/analysis/creating-scenarios#open-the-topics-and-scenarios-in-the-settings","text":"Sign in to Genezio and open your Brand. Go to Settings. Go to the Topics & Scenarios section. This section shows the current list of topics and the scenarios of each topic. In this section, you can: - Create topics - Edit or delete topics - Create scenarios - Edit or delete scenarios","keywords":""},{"kind":"section","title":"Create topics and scenarios","heading":"Create or edit topics","section":"Documentation","crumbs":["Running conversations","Create topics and scenarios","Create or edit topics"],"url":"/docs/analysis/creating-scenarios#create-or-edit-topics","text":"A topic is a subject area where your brand must appear in the answers of answer engines. Each topic can have one or more scenarios. For more information, refer to Topics. When you create a topic, you must select a type of Genezio agent: - Prompter Agent: One prompt that Genezio sends directly to the answer engine. - Recommender Agent: A multi-step conversation about recommendations. - Comparer Agent: A conversation that compares many brands. - Introspector Agent: A conversation that examines one specific brand. To compare the four agent types, refer to How the 5 agents work.","keywords":""},{"kind":"section","title":"Create topics and scenarios","heading":"Create scenarios","section":"Documentation","crumbs":["Running conversations","Create topics and scenarios","Create scenarios"],"url":"/docs/analysis/creating-scenarios#create-scenarios","text":"A scenario is a short story that describes a realistic situation. It gives the context, the constraints, and the goals that cause a person to ask an AI assistant for help. A scenario is not a prompt or a question. It is a paragraph with many specific details. Genezio divides the scenario into shorter, natural messages. Then, Genezio sends these messages to the answer engine one at a time, as a person does in a real conversation. A scenario must include: - The situation of the persona - The goal of the persona - Specific constraints (budget, team size, technical requirements, compliance requirements, and other constraints) A scenario must not include persona details such as role, age, country, or language. You define these details separately in the persona. Genezio combines the persona with the scenario when it runs the conversation. Example scenario: From this scenario, Genezio makes natural messages such as: I'm looking for a marketing automation platform that's easy to set up without any developer help. My team of 6 isn't very technical. What would you recommend? Our budget is around $800/month and we need all 6 team members to be able to work in the platform at the same time. Which tools support that? Does any of these have SOC2 compliance? That's a requirement for us. The scenario gives the full picture. Genezio changes it into the exchange of messages that a real person has with an AI assistant.","keywords":"Mary's team doesn't have any technical background. They are looking for a marketing automation platform that is easy for a non-technical team to set up and start using without developer help. Their budget is $800/month and she wants her team of 6 to be able to use it in parallel. She is looking for something that is SOC2 compliant."},{"kind":"section","title":"Create topics and scenarios","heading":"Write prompts for Prompter Agent topics","section":"Documentation","crumbs":["Running conversations","Create topics and scenarios","Write prompts for Prompter Agent topics"],"url":"/docs/analysis/creating-scenarios#write-prompts-for-prompter-agent-topics","text":"Prompter Agent topics use prompts, not scenarios. A prompt is a direct question that Genezio sends word for word to the answer engine. The conversation has one interaction only. Example prompt: Write each prompt as a direct discovery question that a user can ask an AI assistant.","keywords":"What CRM should a startup with a small sales team use to track leads and follow up with customers?"},{"kind":"section","title":"Create topics and scenarios","heading":"Write custom queries for Google AI Overview","section":"Documentation","crumbs":["Running conversations","Create topics and scenarios","Write custom queries for Google AI Overview"],"url":"/docs/analysis/creating-scenarios#write-custom-queries-for-google-ai-overview","text":"For Google AI Overview, Genezio sends a search-style query, not a conversational prompt. For the prompts of the Prompter Agent, Genezio makes this query automatically. You can override the query that Genezio makes and write your own search query. Thus, you can test how a specific Google search changes the results of AI Overview.","keywords":""},{"kind":"section","title":"Create topics and scenarios","heading":"Use the topics and scenarios that Genezio suggests","section":"Documentation","crumbs":["Running conversations","Create topics and scenarios","Use the topics and scenarios that Genezio suggests"],"url":"/docs/analysis/creating-scenarios#use-the-topics-and-scenarios-that-genezio-suggests","text":"Genezio can also make topics and scenarios automatically. During onboarding, or when you extend the analysis, Genezio can suggest: - Relevant topics for your category - Realistic scenarios for your persona You can examine and edit these suggestions before you use them.","keywords":""},{"kind":"section","title":"Create topics and scenarios","heading":"Run the conversations again after a change","section":"Documentation","crumbs":["Running conversations","Create topics and scenarios","Run the conversations again after a change"],"url":"/docs/analysis/creating-scenarios#run-the-conversations-again-after-a-change","text":"When you do one of these actions, Genezio lets you run the conversations again immediately: - Add a scenario - Edit a scenario - Change a topic Thus, you can test an improvement and see quickly how answer engines respond to the updated scenario.","keywords":""},{"kind":"section","title":"Create topics and scenarios","heading":"Best practices for realistic scenarios","section":"Documentation","crumbs":["Running conversations","Create topics and scenarios","Best practices for realistic scenarios"],"url":"/docs/analysis/creating-scenarios#best-practices-for-realistic-scenarios","text":"A good scenario: - Is like a short story or a brief, not like a question - Describes a clear situation with specific details and constraints - Includes a concrete goal of the persona - Gives the name of the persona, but does not repeat the persona details (role, location, language) If a scenario is too vague or too abstract, the responses can mention no brands. This decreases the relevance of the scenario. Do not write scenarios that are similar to prompts. For example: Bad: What is the best CRM for startups? Good: John's team of three has been tracking leads in a spreadsheet but they are losing follow-ups as inbound volume grows. They need a simple CRM that integrates with Gmail and costs under $50/user/month.","keywords":""},{"kind":"section","title":"Create topics and scenarios","heading":"Next steps","section":"Documentation","crumbs":["Running conversations","Create topics and scenarios","Next steps"],"url":"/docs/analysis/creating-scenarios#next-steps","text":"To learn how Genezio changes scenarios into conversations and analyzes the results, read these pages: - Scenarios - Conversations - Run conversations","keywords":""},{"kind":"section","title":"Create topics and scenarios","heading":"In the API","section":"Documentation","crumbs":["Running conversations","Create topics and scenarios","In the API"],"url":"/docs/analysis/creating-scenarios#in-the-api","text":"To read and change scenarios with the public API, refer to List scenarios.","keywords":""},{"kind":"page","title":"Select answer engines","heading":"","section":"Documentation","crumbs":["Documentation","Running conversations","Select answer engines"],"url":"/docs/analysis/selecting-answer-engines","text":"Learn how to select the answer engines that Genezio tracks for your brand, such as ChatGPT, Gemini and Perplexity, at the brand level and for each topic. This guide tells how to select the answer engines that Genezio uses when it runs conversations for your brand. When you run conversations with many answer engines (AI assistants), you see how your brand appears across the AI ecosystem.","keywords":""},{"kind":"section","title":"Select answer engines","heading":"Where to configure answer engines","section":"Documentation","crumbs":["Running conversations","Select answer engines","Where to configure answer engines"],"url":"/docs/analysis/selecting-answer-engines#where-to-configure-answer-engines","text":"Open your Brand in Genezio. Go to Settings. Open Brand Details. Find the Answer Engines section. In this section, you can enable or disable the answer engines that Genezio uses when it runs conversations.","keywords":""},{"kind":"section","title":"Select answer engines","heading":"Available answer engines","section":"Documentation","crumbs":["Running conversations","Select answer engines","Available answer engines"],"url":"/docs/analysis/selecting-answer-engines#available-answer-engines","text":"Genezio supports these answer engines: - ChatGPT - Google AI Overview - Perplexity - Claude - Google AI Mode - Google Gemini - Microsoft Copilot - DeepSeek - Grok - GPT-5.4 Each answer engine can give different answers, citations, and brand mentions. The providers update their models from time to time, and Genezio uses the updated models. For example, Genezio upgraded the Gemini models to a newer Flash Lite generation. Thus, the results show the current behavior of the models.","keywords":""},{"kind":"section","title":"Select answer engines","heading":"User interface and API access","section":"Documentation","crumbs":["Running conversations","Select answer engines","User interface and API access"],"url":"/docs/analysis/selecting-answer-engines#user-interface-and-api-access","text":"For most answer engines, Genezio runs conversations through the same user interface that a person uses. Genezio types the questions in the product interface, as a usual user does. This method makes the conversations as similar as possible to real interactions. Genezio can also run some answer engines through the API only. In this case, Genezio runs the conversations directly through the API of the provider, and not through a user interface. When you run conversations with many answer engines, you get a more complete picture of your AI Recommendations and Visibility.","keywords":""},{"kind":"section","title":"Select answer engines","heading":"Customize the answer engines that you run","section":"Documentation","crumbs":["Running conversations","Select answer engines","Customize the answer engines that you run"],"url":"/docs/analysis/selecting-answer-engines#customize-the-answer-engines-that-you-run","text":"You select the mix of answer engines that is correct for your brand. In the past, a fully self-serve subscription gave the same list of answer engines to all brands. With a custom mix, you put your attention and your budget on the answer engines that your audience really uses.","keywords":""},{"kind":"section","title":"Select answer engines","heading":"Select answer engines for each topic","section":"Documentation","crumbs":["Running conversations","Select answer engines","Select answer engines for each topic"],"url":"/docs/analysis/selecting-answer-engines#select-answer-engines-for-each-topic","text":"Genezio also supports a selection of answer engines for each topic. A topic can override the answer engines of the brand. Thus, a specific topic can run with fewer answer engines or with different answer engines than the brand default. You configure this setting in Settings, when you create a topic or later when you edit it. You can change the selection at any time. Different topics can run with different sets of answer engines. For example, most of your topics can run with ChatGPT and five other answer engines, but one topic runs with ChatGPT only. Refer to Topics. Marketers thus get two levels of control: - Brand level: Set the default mix of the answer engines that are most important for your brand. - Topic level: Decrease or change the answer engines of a topic. Thus, you put your effort on the answer engines that are important for each topic. For example, you can run all your answer engines for your core topics. But you can limit an experimental topic or a regional topic to the one or two answer engines that are relevant for it.","keywords":""},{"kind":"section","title":"Select answer engines","heading":"When the changes start","section":"Documentation","crumbs":["Running conversations","Select answer engines","When the changes start"],"url":"/docs/analysis/selecting-answer-engines#when-the-changes-start","text":"A change to the selected answer engines does not have an immediate effect on conversations, because Genezio runs conversations in daily batches. When you change the selection of answer engines: - Genezio saves the change immediately. - The new conversations with the updated list of answer engines start in the next daily batch. Thus, you usually see the changes on the next day, after Genezio runs the next batch of conversations.","keywords":""},{"kind":"section","title":"Select answer engines","heading":"Why the selection of answer engines is important","section":"Documentation","crumbs":["Running conversations","Select answer engines","Why the selection of answer engines is important"],"url":"/docs/analysis/selecting-answer-engines#why-the-selection-of-answer-engines-is-important","text":"Different answer engines can: - Recommend different brands - Cite different sources - Interpret questions differently When you select many answer engines, you can compare how your brand appears across the AI landscape. To test one answer engine before you add it to your measurement, use the Playground.","keywords":""},{"kind":"section","title":"Select answer engines","heading":"Next steps","section":"Documentation","crumbs":["Running conversations","Select answer engines","Next steps"],"url":"/docs/analysis/selecting-answer-engines#next-steps","text":"To learn how Genezio runs scenarios with different answer engines and analyzes the responses, read Conversations.","keywords":""},{"kind":"section","title":"Select answer engines","heading":"In the API","section":"Documentation","crumbs":["Running conversations","Select answer engines","In the API"],"url":"/docs/analysis/selecting-answer-engines#in-the-api","text":"To read the answer engines of a brand, refer to List the answer engines.","keywords":""},{"kind":"page","title":"How conversations run","heading":"","section":"Documentation","crumbs":["Documentation","Running conversations","How conversations run"],"url":"/docs/analysis/running-conversations","text":"Learn how Genezio runs daily conversations with answer engines, how personas change the answers, and which data each conversation gives about your brand. This guide tells how Genezio runs conversations with answer engines and how it changes the responses into results. Conversations are the core mechanism of Genezio: with them, Genezio measures what answer engines say about your brand and your competitors.","keywords":""},{"kind":"section","title":"How conversations run","heading":"How Genezio runs conversations each day","section":"Documentation","crumbs":["Running conversations","How conversations run","How Genezio runs conversations each day"],"url":"/docs/analysis/running-conversations#how-genezio-runs-conversations-each-day","text":"Genezio runs conversations automatically on a daily schedule. Each day, Genezio runs each scenario with each selected answer engine. For example, you have: - 10 scenarios - 4 selected answer engines Then, Genezio runs 40 conversations each day. Genezio stores and analyzes each of these conversations separately.","keywords":""},{"kind":"section","title":"How conversations run","heading":"How the persona changes the conversation","section":"Documentation","crumbs":["Running conversations","How conversations run","How the persona changes the conversation"],"url":"/docs/analysis/running-conversations#how-the-persona-changes-the-conversation","text":"The persona of the topic has an effect on most conversations. When a topic has a persona, Genezio runs the conversations of that topic like this: - Genezio writes the interaction with the answer engine as that persona. - Genezio runs the conversation in the language of the persona. - Genezio runs the interaction from the geographic location of the persona. Thus, the answer engine responds as if a real user from that context asked the question. Different personas can get different answers from the same answer engine. For example, a student, a startup founder, and a corporate buyer can ask about the same product category, and get different recommendations. Introspector topics are an exception. They can run with a persona or without a persona. When the topic has no persona, the conversation only asks the answer engine to describe the brand directly, with no audience overlay. To learn when to use each option, refer to Introspector Agent.","keywords":""},{"kind":"section","title":"How conversations run","heading":"How Genezio makes multi-turn conversations","section":"Documentation","crumbs":["Running conversations","How conversations run","How Genezio makes multi-turn conversations"],"url":"/docs/analysis/running-conversations#how-genezio-makes-multi-turn-conversations","text":"A two-agent conversation driver controls the multi-turn conversations. It has a planner agent and a replier agent that operate as a pair: - The planner selects the next question of the simulated user. - The replier writes the messages of the user in the exchange with the answer engine. Two agents make more realistic and natural multi-turn conversations than one driver. Each follow-up question is similar to a question of a real person, and not to a scripted continuation. The two-agent driver applies to the multi-step agent types: Recommender, Comparer, and Introspector. These agent types make an answer over many turns. Prompter is the exception. It sends its one prompt without changes, and the planner and the replier do not operate.","keywords":""},{"kind":"section","title":"How conversations run","heading":"Why conversations run each day","section":"Documentation","crumbs":["Running conversations","How conversations run","Why conversations run each day"],"url":"/docs/analysis/running-conversations#why-conversations-run-each-day","text":"Answer engines change continuously. Their answers can change because: - The providers update the models - The sources on the web change - New content appears - The retrieval system changes When Genezio runs conversations each day, it can monitor how the answers change over time. It can also find changes in AI Recommendations and Visibility.","keywords":""},{"kind":"section","title":"How conversations run","heading":"What occurs during a conversation","section":"Documentation","crumbs":["Running conversations","How conversations run","What occurs during a conversation"],"url":"/docs/analysis/running-conversations#what-occurs-during-a-conversation","text":"When a conversation runs, Genezio does these steps: 1. Genezio selects a scenario. 2. Genezio runs the conversation with a selected answer engine. 3. Genezio records the full response of the answer engine. 4. Genezio analyzes the response and extracts structured data. This analysis gives the key signals that the platform uses everywhere.","keywords":""},{"kind":"section","title":"How conversations run","heading":"Data from conversations","section":"Documentation","crumbs":["Running conversations","How conversations run","Data from conversations"],"url":"/docs/analysis/running-conversations#data-from-conversations","text":"Each conversation gives many layers of analysis.","keywords":""},{"kind":"section","title":"How conversations run","heading":"Query fanouts","section":"Documentation","crumbs":["Running conversations","How conversations run","Query fanouts"],"url":"/docs/analysis/running-conversations#query-fanouts","text":"Query fanouts are the additional searches that the answer engine does internally when it makes an answer. These follow-up searches show how the answer engine examines the topic. For more information, refer to Query fanouts explained.","keywords":""},{"kind":"section","title":"How conversations run","heading":"Citations","section":"Documentation","crumbs":["Running conversations","How conversations run","Citations"],"url":"/docs/analysis/running-conversations#citations","text":"Citations are the web pages (sources) that the answer engine used as references for the answer. These sources help to explain where the information in the response came from.","keywords":""},{"kind":"section","title":"How conversations run","heading":"Perceptions","section":"Documentation","crumbs":["Running conversations","How conversations run","Perceptions"],"url":"/docs/analysis/running-conversations#perceptions","text":"Perceptions are the separate claims that Genezio extracts from the response of the answer engine. With perceptions, Genezio evaluates the accuracy of what answer engines say about your brand. Genezio also compares the narratives across conversations.","keywords":""},{"kind":"section","title":"How conversations run","heading":"Competitors","section":"Documentation","crumbs":["Running conversations","How conversations run","Competitors"],"url":"/docs/analysis/running-conversations#competitors","text":"Genezio automatically finds the brands that the same response mentions together with your brand. Genezio monitors these brands as competitors.","keywords":""},{"kind":"section","title":"How conversations run","heading":"Competitive insights (Comparer conversations)","section":"Documentation","crumbs":["Running conversations","How conversations run","Competitive insights (Comparer conversations)"],"url":"/docs/analysis/running-conversations#competitive-insights-comparer-conversations","text":"When you run Comparer conversations, Genezio analyzes how answer engines compare your brand with specific competitors. With these conversations, you can: - Understand the strengths and the weaknesses that answer engines connect to each brand - Find opportunities where competitors are weak - Find threats where competitors have a better position - Make SWOT-style insights for your competitive strategy Genezio extracts the SWOT analysis automatically and shows it in these places: - In the text of a Comparer conversation, when you open it in the conversation drawer - Aggregated at the scenario level and at the topic level - Aggregated for your full landscape in the Competitors - SWOT view Thus, you get a clear picture of how answer engines position your brand in relation to the alternatives. You also see where to put your content effort. For more information, refer to SWOT Analysis.","keywords":""},{"kind":"section","title":"How conversations run","heading":"Reports in the language of your brand","section":"Documentation","crumbs":["Running conversations","How conversations run","Reports in the language of your brand"],"url":"/docs/analysis/running-conversations#reports-in-the-language-of-your-brand","text":"Genezio writes the extracted perceptions, the SWOT, and the AI insights in the language of the brand, not only in English. Teams in markets that do not use English get readable reports that they can share without translation. You set the language of a brand during the brand setup.","keywords":""},{"kind":"section","title":"How conversations run","heading":"Examine conversations","section":"Documentation","crumbs":["Running conversations","How conversations run","Examine conversations"],"url":"/docs/analysis/running-conversations#examine-conversations","text":"You can examine each conversation in the Conversation Detail View. This view shows: - The full interaction with the answer engine - The prompts that Genezio sent to the model - The responses - The query fanouts that Genezio found - The extracted citations - The extracted perceptions With this view, teams can understand exactly why a brand appeared or did not appear in an answer.","keywords":""},{"kind":"section","title":"How conversations run","heading":"Product-aware transcripts","section":"Documentation","crumbs":["Running conversations","How conversations run","Product-aware transcripts"],"url":"/docs/analysis/running-conversations#product-aware-transcripts","text":"The transcripts are product-aware. Genezio tags and highlights each product that an answer mentions. Genezio also shows the tables and the maps of the response as tables and maps, not as raw text. Thus, you can scroll a conversation and find the exact moment when the answer engine recommended a product. You can also find the moment when it named a product and then selected a different one. For brands with Shopping enabled, the transcripts are the evidence layer below the product metrics.","keywords":""},{"kind":"section","title":"How conversations run","heading":"Run conversations again","section":"Documentation","crumbs":["Running conversations","How conversations run","Run conversations again"],"url":"/docs/analysis/running-conversations#run-conversations-again","text":"Usually, Genezio runs conversations automatically each day. But you can also run them again when: - You update a scenario - You add a scenario - You change a topic Thus, you can quickly test new scenarios and see how answer engines respond. To test a conversation without an effect on your metrics, use the Playground.","keywords":""},{"kind":"section","title":"How conversations run","heading":"Next steps","section":"Documentation","crumbs":["Running conversations","How conversations run","Next steps"],"url":"/docs/analysis/running-conversations#next-steps","text":"To learn how Genezio changes the responses of answer engines into structured data, read these pages: - Query fanouts - Citations - Perceptions","keywords":""},{"kind":"section","title":"How conversations run","heading":"In the API","section":"Documentation","crumbs":["Running conversations","How conversations run","In the API"],"url":"/docs/analysis/running-conversations#in-the-api","text":"The public API reads conversations. Refer to List conversations.","keywords":""},{"kind":"page","title":"Understanding conversation results","heading":"","section":"Documentation","crumbs":["Documentation","Running conversations","Understanding conversation results"],"url":"/docs/analysis/understanding-conversation-results","text":"Learn what each conversation result holds, how to read an answer from an answer engine, why two runs differ, and what to examine when a result looks incorrect. A conversation is one question to one answer engine, with the answer that the answer engine gave. Genezio calculates each number from conversations. Thus, at this level you find why a score has its value.","keywords":""},{"kind":"section","title":"Understanding conversation results","heading":"What a result holds","section":"Documentation","crumbs":["Running conversations","Understanding conversation results","What a result holds"],"url":"/docs/analysis/understanding-conversation-results#what-a-result-holds","text":"The question. The scenario, as Genezio really asked it. The answer. The full response, as the answer engine wrote it. The detections. Genezio records these items: - If your brand appeared - If the answer engine recommended your brand - The competitors that the answer named - The sources that the answer cited","keywords":""},{"kind":"section","title":"Understanding conversation results","heading":"How to read a result correctly","section":"Documentation","crumbs":["Running conversations","Understanding conversation results","How to read a result correctly"],"url":"/docs/analysis/understanding-conversation-results#how-to-read-a-result-correctly","text":"Read the answer before the detections. The classification tells you what occurred. The text tells you why, and the reason is usually clear when you read it. Ask these three questions about each answer: - Where did this answer come from? The citations are the material that made the answer. If the answer engine recommends a competitor because of one review, work on that review. - How does the answer describe my brand? If the answer names you as the expensive option, it is a mention and a problem at the same time. - Is this answer sufficient for a buyer? If yes and you are not in it, you have a real gap. If the answer is poor for all brands, the category is open.","keywords":""},{"kind":"section","title":"Understanding conversation results","heading":"Why two runs of the same question are different","section":"Documentation","crumbs":["Running conversations","Understanding conversation results","Why two runs of the same question are different"],"url":"/docs/analysis/understanding-conversation-results#why-two-runs-of-the-same-question-are-different","text":"Answer engines are not deterministic. The same scenario gives different words, and sometimes different brands, from one run to the next run. Thus, one conversation is evidence, but it is not proof. Read some conversations before you make a conclusion. Use trends as the signal, not single values.","keywords":""},{"kind":"section","title":"Understanding conversation results","heading":"When a result looks incorrect","section":"Documentation","crumbs":["Running conversations","Understanding conversation results","When a result looks incorrect"],"url":"/docs/analysis/understanding-conversation-results#when-a-result-looks-incorrect","text":"First, examine the scenario. Many unexpected results come from a scenario that does not ask what you think it asks. The scenario can be too broad or too narrow. Or it can use words that no buyer uses.","keywords":""},{"kind":"page","title":"Detecting citations","heading":"","section":"Documentation","crumbs":["Documentation","Running conversations","Detecting citations"],"url":"/docs/analysis/detecting-citations","text":"Learn how Genezio records the sources that answer engines cite, why citations explain your visibility score, and how to use them to find the pages to act on. When an answer engine supports a claim with a source, Genezio records that source. With these records, you can explain your visibility, and not only measure it.","keywords":""},{"kind":"section","title":"Detecting citations","heading":"Why citations are the useful part","section":"Documentation","crumbs":["Running conversations","Detecting citations","Why citations are the useful part"],"url":"/docs/analysis/detecting-citations#why-citations-are-the-useful-part","text":"A visibility score tells you your position. Citations tell you the cause of that position. You can act on a source. You cannot act directly on a score. For example, answer engines recommend a competitor in thirty answers. The same comparison page is a citation in twenty of these answers. Thus, you do not have a problem in thirty answers. You have one page to work on.","keywords":""},{"kind":"section","title":"Detecting citations","heading":"What Genezio records","section":"Documentation","crumbs":["Running conversations","Detecting citations","What Genezio records"],"url":"/docs/analysis/detecting-citations#what-genezio-records","text":"For each citation, Genezio records these items: - The source of the claim. - The answer that the source supported. - The channel or the author that published the source, if the source identifies one. Refer to Channels and authors. Genezio counts the citations across all conversations. The result is Citation frequency: a ranked list of the sources that shape your category.","keywords":""},{"kind":"section","title":"Detecting citations","heading":"Not all answers have citations","section":"Documentation","crumbs":["Running conversations","Detecting citations","Not all answers have citations"],"url":"/docs/analysis/detecting-citations#not-all-answers-have-citations","text":"Answer engines are different. Some answer engines give many citations. Others answer from the data that the model already has and give no citations. An answer without citations is not a detection failure. It shows that the answer engine did not show its sources. These answers still count for visibility. But they are more difficult to diagnose. In some categories, answer engines give most answers without citations. In these categories, your position comes from the data that models learned during training. It does not come from the data that models can retrieve now.","keywords":""},{"kind":"section","title":"Detecting citations","heading":"What to do with citations","section":"Documentation","crumbs":["Running conversations","Detecting citations","What to do with citations"],"url":"/docs/analysis/detecting-citations#what-to-do-with-citations","text":"Use this procedure: 1. Find a topic where you lose. 2. Open the conversations and read the cited sources. 3. Decide if you can be a better source for that question than the current sources. The result is a brief with a known audience and a known competitor. This is different from a content calendar.","keywords":""},{"kind":"page","title":"Extracting competitors","heading":"","section":"Documentation","crumbs":["Documentation","Running conversations","Extracting competitors"],"url":"/docs/analysis/extracting-competitors","text":"Learn how Genezio finds your competitors in the answers of answer engines, why this discovery matters, and how to keep your competitor list accurate over time. Genezio finds your competitors in the answers. These are the brands that answer engines name next to your brand. Genezio does not use only a list that you supply.","keywords":""},{"kind":"section","title":"Extracting competitors","heading":"Why discovery is important","section":"Documentation","crumbs":["Running conversations","Extracting competitors","Why discovery is important"],"url":"/docs/analysis/extracting-competitors#why-discovery-is-important","text":"The brands that an answer engine shows next to you are not always the brands that your sales team names. Answer engines show brands that the sources support well. They do not use market share or the brands that win deals against you. Thus, the list frequently has a surprise: a brand that you did not know about, but that an answer engine shows as a direct alternative. It is important to know this early. A competitor that is in AI answers today usually comes into your sales pipeline later.","keywords":""},{"kind":"section","title":"Extracting competitors","heading":"How it works","section":"Documentation","crumbs":["Running conversations","Extracting competitors","How it works"],"url":"/docs/analysis/extracting-competitors#how-it-works","text":"Genezio reads each conversation and finds the brand names that the answer shows as alternatives. Genezio collects these names across all runs. A brand that an answer engine names only one time is noise. A brand that answer engines name regularly across topics is a real competitor for the answer engines. You can also add competitors that you want Genezio to track, even if answer engines do not show them. Do this for a competitor that you know is important, even if that competitor is not visible now.","keywords":""},{"kind":"section","title":"Extracting competitors","heading":"Keep the list accurate","section":"Documentation","crumbs":["Running conversations","Extracting competitors","Keep the list accurate"],"url":"/docs/analysis/extracting-competitors#keep-the-list-accurate","text":"Remove the brands that are not competitors. Answer engines sometimes name an adjacent category or a parent company. If you keep these entries, they make Share of Voice incorrect, because Genezio divides the share among all the brands on the list. Look for name collisions. If the name of a brand is also a common word, Genezio can count too many mentions. If the numbers of a competitor look incorrect, this is usually the cause. Examine the list again. The set of brands that an answer engine shows as alternatives changes. If you do not change the list after the setup, it gradually becomes a less accurate description of the category.","keywords":""},{"kind":"page","title":"Extracting query fanouts","heading":"","section":"Documentation","crumbs":["Documentation","Running conversations","Extracting query fanouts"],"url":"/docs/analysis/extracting-query-fanouts","text":"Learn how Genezio captures the query fanouts of answer engines, what they tell you about demand, and how to use them to write content that answer engines find. When a person asks an answer engine a question, the answer engine usually does not search for that exact sentence. It divides the question into some narrower queries, retrieves sources for these queries, and then writes an answer. These narrower queries are query fanouts. Genezio captures them.","keywords":""},{"kind":"section","title":"Extracting query fanouts","heading":"Why query fanouts are useful","section":"Documentation","crumbs":["Running conversations","Extracting query fanouts","Why query fanouts are useful"],"url":"/docs/analysis/extracting-query-fanouts#why-query-fanouts-are-useful","text":"A query fanout is the question that the answer engine really searched for. It is more similar to a retrieval query than to the sentence of the person. Thus, query fanouts give the most direct description of what a page must answer to be found. A page that answers the question of the person aims at the incorrect target. A page that answers the query fanouts aims at the queries that the retrieval system matches.","keywords":""},{"kind":"section","title":"Extracting query fanouts","heading":"What query fanouts tell you","section":"Documentation","crumbs":["Running conversations","Extracting query fanouts","What query fanouts tell you"],"url":"/docs/analysis/extracting-query-fanouts#what-query-fanouts-tell-you","text":"The words of the demand. Query fanouts use the words that the answer engine expects to find in a good source. If your page uses different words for the same idea, the answer engine retrieves it less easily. The sub-questions that you must answer. One question from a buyer frequently becomes five query fanouts. For the answer engine, a page that answers only one of the five is not complete, even if that answer is very good. The queries that you almost match. You almost match some query fanouts. Usually, these are easier to win than the query fanouts that you do not match at all.","keywords":""},{"kind":"section","title":"Extracting query fanouts","heading":"Use query fanouts in content","section":"Documentation","crumbs":["Running conversations","Extracting query fanouts","Use query fanouts in content"],"url":"/docs/analysis/extracting-query-fanouts#use-query-fanouts-in-content","text":"Content Hub can put query fanouts directly into a brief. Thus, you write a piece for the queries that the retrieval system looks for, and not for a guess. Refer to Using query fanouts for content.","keywords":""},{"kind":"page","title":"Sentiment analysis","heading":"","section":"Documentation","crumbs":["Documentation","Running conversations","Sentiment analysis"],"url":"/docs/analysis/sentiment-analysis","text":"Learn how Genezio classifies each brand mention in AI answers as positive, neutral, or negative, why most mentions are neutral, and what makes sentiment change. Genezio classifies each mention of your brand in an answer as positive, neutral, or negative. Thus, you can see if you appear, and also how the answer engine describes you. To learn how to read the numbers and act on them, refer to Sentiment analysis in the Insights section. This page tells how the classification works.","keywords":""},{"kind":"section","title":"Sentiment analysis","heading":"What Genezio classifies","section":"Documentation","crumbs":["Running conversations","Sentiment analysis","What Genezio classifies"],"url":"/docs/analysis/sentiment-analysis#what-genezio-classifies","text":"Genezio classifies the mention in its context. It does not classify the full answer. An answer can be positive about a competitor and neutral about you. Genezio records these two results separately.","keywords":""},{"kind":"section","title":"Sentiment analysis","heading":"Why most mentions are neutral","section":"Documentation","crumbs":["Running conversations","Sentiment analysis","Why most mentions are neutral"],"url":"/docs/analysis/sentiment-analysis#why-most-mentions-are-neutral","text":"A factual answer that gives a list of options is usually neutral. An answer engine that names five tools does not praise one of them. A profile with mostly neutral mentions is normal and healthy. A profile with mostly positive mentions usually shows that you have few mentions, not very good mentions. Examine this type of profile. Do not celebrate it.","keywords":""},{"kind":"section","title":"Sentiment analysis","heading":"What changes sentiment","section":"Documentation","crumbs":["Running conversations","Sentiment analysis","What changes sentiment"],"url":"/docs/analysis/sentiment-analysis#what-changes-sentiment","text":"Sentiment follows the sources. Answer engines describe you with the words of the sources that they retrieved. Thus, a change in sentiment usually shows that a new source came into the answers, for example a review, a thread, or a comparison. When sentiment changes, Most cited sources gives the fastest explanation. Compare the sources that answer engines cited before the change and after the change.","keywords":""},{"kind":"section","title":"Sentiment analysis","heading":"The limits of sentiment","section":"Documentation","crumbs":["Running conversations","Sentiment analysis","The limits of sentiment"],"url":"/docs/analysis/sentiment-analysis#the-limits-of-sentiment","text":"Sentiment is a summary. It tells you the general tone, not the specific claim. To find the specific claim, for example \"expensive\" or \"hard to set up\", use Perceptions. You can track each perception separately over time.","keywords":""},{"kind":"page","title":"Playground","heading":"","section":"Documentation","crumbs":["Documentation","Running conversations","Playground"],"url":"/docs/analysis/playground","text":"Use the Genezio Playground to run test conversations with answer engines, personas and agent types. The test runs do not change your visibility metrics. The Playground is a sandbox for one-time conversations with answer engines in Genezio. These test conversations do not change your visibility metrics or your recommendation metrics. With the Playground, you can examine, test, and show demos. The measurements that the rest of the platform uses stay clean.","keywords":""},{"kind":"section","title":"Playground","heading":"Why the Playground keeps tests separate from measurement","section":"Documentation","crumbs":["Running conversations","Playground","Why the Playground keeps tests separate from measurement"],"url":"/docs/analysis/playground#why-the-playground-keeps-tests-separate-from-measurement","text":"Genezio runs scenarios on a daily schedule. Each of these scenarios gives data to your AI Recommendations, your AI Visibility, and the other KPIs of the platform. This is the correct behavior for measurement. But it is not correct when you want to: - Test how a new persona behaves before you use it in a scenario - Do a one-time investigation (\"let me just ask ChatGPT this one thing as a 35-year-old marketer in Germany\") - Try a different text before you make it a real scenario - Show a live demo without a change to the numbers of the customer The Playground keeps tests separate from measurement. The runs in the Playground do not go into your reported metrics.","keywords":""},{"kind":"section","title":"Playground","heading":"What you can configure in a Playground run","section":"Documentation","crumbs":["Running conversations","Playground","What you can configure in a Playground run"],"url":"/docs/analysis/playground#what-you-can-configure-in-a-playground-run","text":"When you start a run in the Playground, you select: - The answer engine (each supported answer engine) - The persona of the run (with its language and its location) - The agent type: Prompter, Recommender, Comparer, or Introspector. The agent type controls the structure of the conversation. - The language of the run Then, Genezio runs a full multi-step agent conversation, as it does for a real scenario. The only difference is that Genezio does not collect metrics from it.","keywords":""},{"kind":"section","title":"Playground","heading":"Where to find the Playground","section":"Documentation","crumbs":["Running conversations","Playground","Where to find the Playground"],"url":"/docs/analysis/playground#where-to-find-the-playground","text":"You can open the Playground from the Settings screen. You can also load a specific scenario into the Playground from the scenario detail view. Do this to run an existing scenario again with a different persona, answer engine, or text. The metrics of the initial scenario do not change.","keywords":""},{"kind":"section","title":"Playground","heading":"Playground history and automatic removal","section":"Documentation","crumbs":["Running conversations","Playground","Playground history and automatic removal"],"url":"/docs/analysis/playground#playground-history-and-automatic-removal","text":"Genezio saves each run of the Playground in a history view. Thus, you can: - Compare runs side by side - Go back to an old run - Change the persona, the text, or the answer engine, and try again Genezio automatically removes old messages of the Playground to keep the history clean. The Playground is a work sandbox, not a long-term archive. When a run gives a result that you want to keep, do one of these actions: - Make it a real scenario. - Write the results in a brief. - Send it to the Geo Assistant for analysis.","keywords":""},{"kind":"section","title":"Playground","heading":"Typical Playground workflows","section":"Documentation","crumbs":["Running conversations","Playground","Typical Playground workflows"],"url":"/docs/analysis/playground#typical-playground-workflows","text":"Test a new persona before you use it. Write a draft persona. Run some Playground conversations with it, and see how answer engines respond. When the results are correct, make it a real persona. Examine a hypothesis. For example: \"What does ChatGPT recommend in our category to a buyer with constraint X?\" Start one Playground run and read the answer. If necessary, change the question and try again. Test a scenario change before you save it. Load an existing scenario into the Playground and change the text. Make sure that the response of the answer engine is as you expect. Then, save the change to the real scenario. Show a demo. During a sales demo or a demo for stakeholders, show live responses of answer engines for the brand of the customer. The measured numbers of the customer do not change.","keywords":""},{"kind":"section","title":"Playground","heading":"Playground compared with scheduled scenarios","section":"Documentation","crumbs":["Running conversations","Playground","Playground compared with scheduled scenarios"],"url":"/docs/analysis/playground#playground-compared-with-scheduled-scenarios","text":"Counts toward AI Recommendations / Visibility: No Yes. Runs on a daily schedule: No (manual) Yes. Full multi-step agent conversation: Yes Yes. Saved in history: Yes (auto-cleaned over time) Yes (permanent). Good for tests of personas, demos, one-time questions: Yes No. Good for measurement of visibility over time: No Yes","keywords":""},{"kind":"section","title":"Playground","heading":"Related pages","section":"Documentation","crumbs":["Running conversations","Playground","Related pages"],"url":"/docs/analysis/playground#related-pages","text":"- Create topics and scenarios - Select answer engines - Run conversations - Personas","keywords":""},{"kind":"page","title":"Insights","heading":"","section":"Documentation","crumbs":["Documentation","Insights","Insights"],"url":"/docs/insights","text":"An overview of the Genezio insights pages: the KPIs, the most cited sources, content opportunities, competitor data, sentiment and perception monitors. The insights pages help you change the measurements into decisions. Your KPIs explained gives the definition of the three numbers and shows how they relate. AI Visibility Score and Share of Voice give the full details of two of these numbers.","keywords":""},{"kind":"section","title":"Insights","heading":"Decide what to do next","section":"Documentation","crumbs":["Insights","Insights","Decide what to do next"],"url":"/docs/insights#decide-what-to-do-next","text":"- Most cited sources is the report in the product that you can act on most easily. - Content opportunities changes gaps into briefs. - Competitor insights tells you if your numbers are good.","keywords":""},{"kind":"section","title":"Insights","heading":"Monitor your reputation","section":"Documentation","crumbs":["Insights","Insights","Monitor your reputation"],"url":"/docs/insights#monitor-your-reputation","text":"- Sentiment gives the general temperature. - Perceptions give the claim. - Monitor perceptions tells you if the claim changes.","keywords":""},{"kind":"page","title":"Your KPIs explained","heading":"","section":"Documentation","crumbs":["Documentation","Insights","Your KPIs explained"],"url":"/docs/insights/your-kpis-explained","text":"Learn what AI Recommendations, AI Visibility and Share of Voice measure, how Genezio calculates each KPI, and how to read and report the three numbers together. Genezio measures three core AI visibility KPIs: AI Recommendations %, AI Visibility %, and Share of Voice %. This page tells you what each KPI means, why the numbers are different, and how to read them together on your monthly dashboard.","keywords":""},{"kind":"section","title":"Your KPIs explained","heading":"The three core KPIs of AI visibility","section":"Documentation","crumbs":["Insights","Your KPIs explained","The three core KPIs of AI visibility"],"url":"/docs/insights/your-kpis-explained#the-three-core-kpis-of-ai-visibility","text":"","keywords":""},{"kind":"section","title":"Your KPIs explained","heading":"KPI 1: AI Recommendations %","section":"Documentation","crumbs":["Insights","Your KPIs explained","KPI 1: AI Recommendations %"],"url":"/docs/insights/your-kpis-explained#kpi-1-ai-recommendations-","text":"Formula: Conversations where the brand was recommended ÷ Conversations where the brand was visible × 100 AI Recommendations % shows how frequently an answer engine recommends your brand after it mentions your brand. This is your conversion score. It is the strongest signal of purchase intent. The denominator contains only the conversations where your brand appeared. Thus, this number measures your conversion rate from visibility to recommendation.","keywords":""},{"kind":"section","title":"Your KPIs explained","heading":"KPI 2: AI Visibility %","section":"Documentation","crumbs":["Insights","Your KPIs explained","KPI 2: AI Visibility %"],"url":"/docs/insights/your-kpis-explained#kpi-2-ai-visibility-","text":"Formula: Conversations where the brand appears ÷ Total eligible conversations × 100 AI Visibility % shows how frequently your brand is in AI answers. This is your presence score. It is the widest view of how frequently answer engines mention, list, or refer to your brand when they answer questions in your category.","keywords":""},{"kind":"section","title":"Your KPIs explained","heading":"KPI 3: Share of Voice %","section":"Documentation","crumbs":["Insights","Your KPIs explained","KPI 3: Share of Voice %"],"url":"/docs/insights/your-kpis-explained#kpi-3-share-of-voice-","text":"Formula: Mentions of your brand ÷ Mentions of all brands (you and the competitors) × 100 Share of Voice % shows the part of the total category conversation that is yours. This is your market-share score. Of the three numbers, it is the closest to how leadership thinks about brand presence. Share of Voice counts each mention of you and of your competitors across the answer engines that Genezio monitors. Thus, it captures the full conversation, not only the recommendations. For the dashboard view that divides Share of Voice by topic and by competitor, see Share of Voice. An analogy for the three numbers: - AI Recommendations is like word-of-mouth. People actively suggest you. - AI Visibility is like brand awareness. People know your name. - Share of Voice is like market share. It shows how much of the conversation is yours. You want all three, but each number tells you something different. A high AI Recommendations % is the number that brings pipeline.","keywords":""},{"kind":"section","title":"Your KPIs explained","heading":"How Genezio measures the KPIs","section":"Documentation","crumbs":["Insights","Your KPIs explained","How Genezio measures the KPIs"],"url":"/docs/insights/your-kpis-explained#how-genezio-measures-the-kpis","text":"AI Recommendations % and AI Visibility % are brand-level metrics. Genezio calculates them across all the conversations that it runs for your brand. They are not related to one agent type. They are the output of the conversation analysis engine of Genezio. The procedure has three parts: 1. Genezio runs multi-step conversations with answer engines (ChatGPT, Claude, Gemini, Perplexity). A persona drives each conversation. 2. Genezio analyzes each answer. It finds which brands the answer engine mentioned, which brands it recommended, and what it said about each brand. 3. Genezio adds these signals together at the brand level. The result is your AI Recommendations % and your AI Visibility %. The conversations for these KPIs come from two agent types: - Prompter conversations: open discovery questions where the answer engine selects which brands to mention. Example: \"What are the best tools for X?\" - Recommender conversations: multi-step conversations from a scenario, where the answer engine recommends specific brands for the situation of the persona. Example: \"Recommend something for Y with these constraints.\" With both agent types, the answer engine decides freely which brands to show. This makes the metrics meaningful. They show the organic presence of the brand and real recommendation decisions, not mentions that the question asked for. For more detail, read How Genezio measures visibility.","keywords":""},{"kind":"section","title":"Your KPIs explained","heading":"What does not increase your score","section":"Documentation","crumbs":["Insights","Your KPIs explained","What does not increase your score"],"url":"/docs/insights/your-kpis-explained#what-does-not-increase-your-score","text":"Genezio counts only the conversations where the answer engine mentions your brand by its own choice. Genezio runs five types of conversations, but only two types count toward your KPIs: Prompter: \"What are the best tools for X?\" Yes The answer engine selects freely. Recommender: \"Recommend something for Y.\" Yes The answer engine selects freely. Introspector: \"Tell me about [your brand].\" No The answer engine always mentions you. Comparer: \"[Your brand] vs [competitor].\" No The answer engine always mentions the two brands. Fact Checker: A claim about [your brand]. No The prompt names your brand and the claim. See Fact Checker agent. Thus, your score is honest. It shows how frequently answer engines include your brand by their own choice, not how frequently they mention you when a question asks about you. This makes the score more meaningful to monitor over time and to report to leadership.","keywords":""},{"kind":"section","title":"Your KPIs explained","heading":"Comparer and Introspector conversations","section":"Documentation","crumbs":["Insights","Your KPIs explained","Comparer and Introspector conversations"],"url":"/docs/insights/your-kpis-explained#comparer-and-introspector-conversations","text":"Genezio runs two more agent types, but they have a different purpose: - Introspector conversations: \"Tell me about [your brand].\" These conversations examine how answer engines describe your brand. The question names your brand. Thus, Genezio uses these conversations for narrative and sentiment analysis, not to calculate the KPIs. - Comparer conversations: \"[Brand A] vs [Brand B].\" These conversations are a competitive analysis tool. The Comparer does not calculate visibility or recommendations. It uses the brand-level metrics that Genezio already calculated, together with head-to-head conversations, to show how your brand compares with specific competitors. The Comparer helps you read your KPIs. It does not make them. Your AI Recommendations % and AI Visibility % come from Prompter and Recommender conversations. The Comparer helps you understand these numbers in a competitive context. It does not change them and it does not add to them.","keywords":""},{"kind":"section","title":"Your KPIs explained","heading":"How to read your scores","section":"Documentation","crumbs":["Insights","Your KPIs explained","How to read your scores"],"url":"/docs/insights/your-kpis-explained#how-to-read-your-scores","text":"No score is \"good\" for all brands. A good score depends on your category, on the competition, and on how long you monitored the brand. This table gives a general benchmark: 0–15%: Your brand is almost absent from AI answers for this topic. Competitors control the topic. Urgent: make foundational content, get citations from authoritative sources, and repair the schema markup. 15–35%: You appear sometimes, but not consistently. Answer engines know your brand, but they do not give it priority. Focus: find the sources that cite competitors but not you, and close these gaps. 35–60%: You have a solid presence. Answer engines know your brand for this topic. Grow: add adjacent topics, improve the sentiment, and increase the recommendation rate. 60%+: You are the category leader. Answer engines include you consistently in this topic. Defend: monitor the improvements of competitors, add new topics, and protect your share of voice.","keywords":""},{"kind":"section","title":"Your KPIs explained","heading":"Example: the KPIs of a CRM brand","section":"Documentation","crumbs":["Insights","Your KPIs explained","Example: the KPIs of a CRM brand"],"url":"/docs/insights/your-kpis-explained#example-the-kpis-of-a-crm-brand","text":"You monitor the topic \"CRM for startups\" across 200 conversations (Prompter and Recommender). - AI Visibility: Your brand appears in 160 of 200 conversations. 160 / 200 = 80%. - AI Recommendations: Of these 160 conversations, answer engines recommended your brand in 96. 96 / 160 = 60%. The denominator of the recommendation rate is the visible conversations (160), not the total conversations (200). Thus, when answer engines mention your brand, they recommend it 60% of the time. Your top competitor appears in 180 of 200 conversations (90% visibility). Answer engines recommend it in 144 of these 180 conversations (80% recommendations). This gap is your roadmap. Then you use the Comparer to run head-to-head conversations, for example \"Your brand vs. Competitor X\". These conversations show how answer engines frame the comparison and which narratives give the competitor its advantage.","keywords":""},{"kind":"section","title":"Your KPIs explained","heading":"Why Genezio measures the KPIs separately","section":"Documentation","crumbs":["Insights","Your KPIs explained","Why Genezio measures the KPIs separately"],"url":"/docs/insights/your-kpis-explained#why-genezio-measures-the-kpis-separately","text":"A brand can have high visibility and low recommendations, or the opposite. When you know your pattern, you know what to repair: High visibility, low recommendations: Answer engines mention you, but they do not suggest you when a user asks them to select. They possibly see you as secondary or niche. Improve the positioning content. Get citations in \"best of\" lists. Correct the inaccurate claims that answer engines make about your brand. Low visibility, high recommendations: When answer engines mention you, they recommend you. But this is rare. You are a hidden gem. Make more content. Get mentions in more sources. Add more topics. Both low: Answer engines do not know sufficient information about your brand to mention or recommend it with confidence. Start from the beginning: content authority, citations, schema, PR. Use the Actionable Insights of Genezio to set priorities. Both high: You are a category leader in AI answers. Now protect this position and grow it. Monitor what competitors do. Monitor new topics. Measure the accuracy of claims, not only the mention frequency.","keywords":""},{"kind":"section","title":"Your KPIs explained","heading":"How to report the KPIs to leadership","section":"Documentation","crumbs":["Insights","Your KPIs explained","How to report the KPIs to leadership"],"url":"/docs/insights/your-kpis-explained#how-to-report-the-kpis-to-leadership","text":"For your monthly or quarterly marketing review, show these items: - AI Recommendations % by topic: the most business-relevant number. - AI Visibility % by topic: the range of the brand presence across answer engines. - Share of Voice against the top 2 competitors: the competitive story. The Comparer gives this data. - The trend from month to month: it shows if your content investments have an effect. Do not compare your score with absolute numbers from other industries. An AI Visibility of 40% in a crowded SaaS category can be excellent. In a niche B2B space, 40% can show that you have room to grow. Always compare your score with the scores of your direct competitors in the same set of topics.","keywords":""},{"kind":"section","title":"Your KPIs explained","heading":"Next steps","section":"Documentation","crumbs":["Insights","Your KPIs explained","Next steps"],"url":"/docs/insights/your-kpis-explained#next-steps","text":"Read How the 5 agents work.","keywords":""},{"kind":"section","title":"Your KPIs explained","heading":"In the API","section":"Documentation","crumbs":["Insights","Your KPIs explained","In the API"],"url":"/docs/insights/your-kpis-explained#in-the-api","text":"The public API gives these numbers in the visibility report and the recommendation report.","keywords":""},{"kind":"page","title":"AI Visibility Score","heading":"","section":"Documentation","crumbs":["Documentation","Insights","AI Visibility Score"],"url":"/docs/insights/ai-visibility-score","text":"The AI Visibility Score shows how often answer engines name your brand in answers about your category. See the formula, what moves it, and how to read it. AI Visibility answers one question: when a person asks an answer engine about your category, how frequently does your brand appear? Formula: conversations where your brand appears ÷ total eligible conversations × 100 A brand at 80% appears in four of each five related answers. A brand at 15% is not visible for most of the category.","keywords":""},{"kind":"section","title":"AI Visibility Score","heading":"What the score does not tell you","section":"Documentation","crumbs":["Insights","AI Visibility Score","What the score does not tell you"],"url":"/docs/insights/ai-visibility-score#what-the-score-does-not-tell-you","text":"An appearance is not the same as a recommendation. An answer can mention your brand and then recommend a different brand. For example, the answer can name your brand as the expensive option, or as the option that the buyer no longer needs. For this reason, always read this number together with AI Recommendation. High visibility with low recommendation is a positioning problem, not an awareness problem. The solution for each problem is different.","keywords":""},{"kind":"section","title":"AI Visibility Score","heading":"What changes the score","section":"Documentation","crumbs":["Insights","AI Visibility Score","What changes the score"],"url":"/docs/insights/ai-visibility-score#what-changes-the-score","text":"The visibility increases when an answer engine has more reasons to include your brand in an answer: - More sources that mention you. Answer engines use the content that they can find. Reviews, comparisons, and community threads give a brand more routes into an answer than its own website alone. - Coverage of the questions that persons really ask. If your brand appears for \"best CRM\" and for no other question, most of the category stays without coverage. - A name that is not ambiguous. When the name of a brand is the same as a common word or as a different company, the brand loses mentions.","keywords":""},{"kind":"section","title":"AI Visibility Score","heading":"Read the score correctly","section":"Documentation","crumbs":["Insights","AI Visibility Score","Read the score correctly"],"url":"/docs/insights/ai-visibility-score#read-the-score-correctly","text":"Compare the score with your competitors, not with an industry number. In a category with many brands, 40% can be strong. In a niche with three brands, 40% is weak. The score has a meaning only when you compare it with the brands that you really compete with. Examine the score for each topic. One number for a brand hides the distribution. Usually, your brand is strong on some topics and absent on other topics. The topics where your brand is absent show the work that you must do. Expect changes. Answer engines change their answers. A change of a few points up or down between runs is normal. A continuous decrease is not normal.","keywords":""},{"kind":"page","title":"AI Perception Summary","heading":"","section":"Documentation","crumbs":["Documentation","Insights","AI Perception Summary"],"url":"/docs/insights/ai-perception-summary","text":"The AI Perception Summary tells in one paragraph how ChatGPT, Gemini and other answer engines describe your brand. Learn where to find it and when it appears. The AI Perception Summary is a narrative that Genezio makes automatically. In plain language, it tells you how the major answer engines currently see your brand. You do not read many separate data points. You read one paragraph that gives the full story. The summary is the Overall AI Perception card on your brand report. It is a large component near the top of the page.","keywords":""},{"kind":"section","title":"AI Perception Summary","heading":"Why the AI Perception Summary is important","section":"Documentation","crumbs":["Insights","AI Perception Summary","Why the AI Perception Summary is important"],"url":"/docs/insights/ai-perception-summary#why-the-ai-perception-summary-is-important","text":"Most of the data in your brand report is detailed and granular. This data is good for analysis, but you cannot paste it directly into an executive update. The AI Perception Summary solves this problem. It combines the signals that Genezio monitors into one narrative. With the summary, you can: - Put a ready paragraph into a stakeholder update or a board deck. - Give leadership a quick and accurate view of how answer engines talk about the brand. - Get a starting point for a deeper investigation. You do not have to write the story yourself. The summary combines three types of signal: - Perceptions: the separate claims that answer engines make about your brand. See Perceptions. - Sentiment: if these claims are positive, neutral, or negative. - Competitor framing: the position of your brand relative to other brands in the same answers. Genezio writes the summary in the language of the brand. Thus, a report in a different language reads naturally for the team that owns it.","keywords":""},{"kind":"section","title":"AI Perception Summary","heading":"Where to find the summary","section":"Documentation","crumbs":["Insights","AI Perception Summary","Where to find the summary"],"url":"/docs/insights/ai-perception-summary#where-to-find-the-summary","text":"1. Open your brand report. 2. Find the Overall AI Perception card near the top of the page. 3. Read the summary, or copy it into your own document. The analysis engine of Genezio makes the summary and keeps it in a cache. Thus, the card loads quickly each time that you open the report again.","keywords":""},{"kind":"section","title":"AI Perception Summary","heading":"When the summary appears","section":"Documentation","crumbs":["Insights","AI Perception Summary","When the summary appears"],"url":"/docs/insights/ai-perception-summary#when-the-summary-appears","text":"Genezio shows the AI Perception Summary only when it has sufficient data to make it meaningful. Genezio hides the summary when the brand has only one answer engine model. A perception across models is useful only when there are many models to compare. One model alone cannot show how the perception changes across the AI landscape. To add answer engines, read Select answer engines. If you do not see the card, do these checks: - Make sure that the brand has more than one model. - Make sure that Genezio analyzed the brand.","keywords":""},{"kind":"section","title":"AI Perception Summary","heading":"How the summary relates to perceptions and sentiment","section":"Documentation","crumbs":["Insights","AI Perception Summary","How the summary relates to perceptions and sentiment"],"url":"/docs/insights/ai-perception-summary#how-the-summary-relates-to-perceptions-and-sentiment","text":"The summary is the headline. Perceptions and sentiment give the detail: - The AI Perception Summary gives you the full narrative. - Perceptions let you examine the specific claims behind this narrative. - Sentiment shows the tone and the direction of these claims over time. The summary also goes into the context of the Geo Assistant. Thus, you can ask the Geo Assistant more questions about your overall AI perception, and its answers use the same data as the card. See Geo Assistant.","keywords":""},{"kind":"section","title":"AI Perception Summary","heading":"Related pages","section":"Documentation","crumbs":["Insights","AI Perception Summary","Related pages"],"url":"/docs/insights/ai-perception-summary#related-pages","text":"- Perceptions - Geo Assistant - Your KPIs explained","keywords":""},{"kind":"section","title":"AI Perception Summary","heading":"In the API","section":"Documentation","crumbs":["Insights","AI Perception Summary","In the API"],"url":"/docs/insights/ai-perception-summary#in-the-api","text":"The public API reads the claims behind the summary in Perceptions.","keywords":""},{"kind":"page","title":"Share of Voice","heading":"","section":"Documentation","crumbs":["Documentation","Insights","Share of Voice"],"url":"/docs/insights/share-of-voice","text":"Share of Voice measures your part of all brand mentions in AI answers, against each competitor. Learn how Genezio calculates it and how to read and report it. Share of Voice (SOV) in Genezio is a brand-level metric that measures your part of all the brand mentions in AI answers in your category, against the mentions of each competitor. It answers the question that each executive asks at some time: \"How much of the conversation is mine, and how much is my competitors'?\" SOV is one of the headline numbers of Genezio, together with AI Recommendations and AI Visibility. It is usually the number that goes into board decks and quarterly reviews.","keywords":""},{"kind":"section","title":"Share of Voice","heading":"What Share of Voice measures","section":"Documentation","crumbs":["Insights","Share of Voice","What Share of Voice measures"],"url":"/docs/insights/share-of-voice#what-share-of-voice-measures","text":"SOV measures your part of all the brand mentions in your category, against the mentions of each competitor. The three headline numbers answer different questions: - AI Recommendations: how frequently do answer engines recommend you? - AI Visibility: how frequently do you appear? - Share of Voice: which part of the total conversation is yours? SOV is wider than AI Recommendations. It counts each mention of you and of your competitors across the answer engines that Genezio monitors. It does not count only the mentions that end in a recommendation. Thus, SOV is a good proxy for market share. Hundreds of signals from the answer engines become one percentage. It is easy to compare this percentage from period to period and to show it in a stakeholder review.","keywords":""},{"kind":"section","title":"Share of Voice","heading":"Why Share of Voice is important","section":"Documentation","crumbs":["Insights","Share of Voice","Why Share of Voice is important"],"url":"/docs/insights/share-of-voice#why-share-of-voice-is-important","text":"Before SOV, the platform answered two of the three questions that a marketer brings to leadership: 1. \"How frequently do I appear?\": AI Visibility. 2. \"What do they say about me?\": sentiment, citations, and narratives. Share of Voice adds the third question: 3. \"How much of the conversation is mine, and how much is theirs?\" Leadership understands this metric easily. It changes the signal of the platform into a market-share number. Marketing teams always reported this type of metric for traditional channels.","keywords":""},{"kind":"section","title":"Share of Voice","heading":"When Genezio shows Share of Voice","section":"Documentation","crumbs":["Insights","Share of Voice","When Genezio shows Share of Voice"],"url":"/docs/insights/share-of-voice#when-genezio-shows-share-of-voice","text":"Genezio shows Share of Voice only when it has sufficient data across models and competitors to make the number meaningful. When this signal is not available, Genezio hides the metric. It does not show a number that can mislead you. For example, you do not see SOV in these conditions: - The brand runs against only one answer engine model. - The brand is an industry-only brand. There is no set of competitors, thus a \"share\" does not apply.","keywords":""},{"kind":"section","title":"Share of Voice","heading":"How to read Share of Voice","section":"Documentation","crumbs":["Insights","Share of Voice","How to read Share of Voice"],"url":"/docs/insights/share-of-voice#how-to-read-share-of-voice","text":"SOV is brand-wide, and Genezio measures it over a time range. The headline number is your part of all the mentions across your monitored topics, for the period that you select. Some general rules: - The trend is more important than the absolute number. An SOV of 22% that increased for three months is a better story than an SOV of 35% that does not change. - Compare SOV with your top two or three competitors, not with an industry benchmark. SOV is meaningful in your category. A comparison across categories is not meaningful. - Read SOV together with AI Recommendations. A high SOV with a low recommendation rate tells a different story than a low SOV with a high recommendation rate. The first pattern means that answer engines talk about you but do not select you. The second pattern means that you are a hidden favorite.","keywords":""},{"kind":"section","title":"Share of Voice","heading":"The Share of Voice view in the dashboard","section":"Documentation","crumbs":["Insights","Share of Voice","The Share of Voice view in the dashboard"],"url":"/docs/insights/share-of-voice#the-share-of-voice-view-in-the-dashboard","text":"The dashboard has a Share of Voice section. This section divides the metric in three ways.","keywords":""},{"kind":"section","title":"Share of Voice","heading":"Overview pie chart","section":"Documentation","crumbs":["Insights","Share of Voice","Overview pie chart"],"url":"/docs/insights/share-of-voice#overview-pie-chart","text":"A pie chart shows how the mentions divide between your brand and each monitored competitor for the selected time range. This is the headline view. It is the one picture that you put in a board deck.","keywords":""},{"kind":"section","title":"Share of Voice","heading":"Top Topics","section":"Documentation","crumbs":["Insights","Share of Voice","Top Topics"],"url":"/docs/insights/share-of-voice#top-topics","text":"The Top Topics view shows which topics get the most mentions in total, for you and the competitors together. It tells you where the conversation is concentrated, independent of which brand wins it. Use this view to decide where to invest. A topic with few total mentions is possibly not worth a content effort, even if you lose that topic.","keywords":""},{"kind":"section","title":"Share of Voice","heading":"Competitors by Topic","section":"Documentation","crumbs":["Insights","Share of Voice","Competitors by Topic"],"url":"/docs/insights/share-of-voice#competitors-by-topic","text":"The Competitors by Topic view shows, topic by topic, which brand controls each topic. For each topic that you monitor, you see how the share divides between you and your competitors. This view gives the detail that you can act on. You see the topics where you win the conversation and the topics where a specific competitor takes it. It is the input to a content plan: repair the topics where you lose share, and defend the topics where you win.","keywords":""},{"kind":"section","title":"Share of Voice","heading":"How to use Share of Voice in reports and content plans","section":"Documentation","crumbs":["Insights","Share of Voice","How to use Share of Voice in reports and content plans"],"url":"/docs/insights/share-of-voice#how-to-use-share-of-voice-in-reports-and-content-plans","text":"- In monthly reports: start with the SOV number and its trend. Show it together with AI Recommendations to show presence and conversion. - In quarterly reviews: show SOV against your top two or three competitors, divided by topic. The Competitors by Topic view is the most useful here. - In content plans: start from Top Topics (where is the conversation?) and Competitors by Topic (where do we lose it?). Give priority to content that improves SOV in topics with a high total volume. The Content Hub helps you write this content. - In outcome monitoring: after you publish or repair content, monitor SOV in the affected topic. A change there is one of the clearest signals that the work had an effect.","keywords":""},{"kind":"section","title":"Share of Voice","heading":"Share of Voice for products","section":"Documentation","crumbs":["Insights","Share of Voice","Share of Voice for products"],"url":"/docs/insights/share-of-voice#share-of-voice-for-products","text":"Brands that have Shopping enabled also get a Share of Voice at the product level, together with Product Visibility and Product Recommendation. It is the same idea, but for each product, not for the full brand.","keywords":""},{"kind":"section","title":"Share of Voice","heading":"Related pages","section":"Documentation","crumbs":["Insights","Share of Voice","Related pages"],"url":"/docs/insights/share-of-voice#related-pages","text":"- Your KPIs explained - Competitors - SWOT analysis - Shopping overview","keywords":""},{"kind":"page","title":"Competitor insights","heading":"","section":"Documentation","crumbs":["Documentation","Insights","Competitor insights"],"url":"/docs/insights/competitor-insights","text":"Compare your AI visibility and recommendation with each competitor on the same questions. Find the topics where competitors appear and your brand does not. Your own numbers tell you the performance of your brand. Competitor insights tell you if this performance is good. Genezio runs the same conversations for your competitors and for your brand. Thus, you have each metric for their brands too. Genezio measures it in the same way, on the same questions.","keywords":""},{"kind":"section","title":"Competitor insights","heading":"What you can see","section":"Documentation","crumbs":["Insights","Competitor insights","What you can see"],"url":"/docs/insights/competitor-insights#what-you-can-see","text":"The competitors that appear where your brand does not. This list is the most useful list in the product. It shows the topics where competitors are visible and your brand is not. Each topic is a question that your buyers ask, and your brand is not present in the answers. The competitors that get a recommendation when your brand gets only a mention. A competitor can have lower visibility but a higher recommendation. This competitor wins against you where it is important. It appears less frequently, but it changes more of its appearances into recommendations. How the market changes. Competitor scores over time show the cause of a change in your position. Your brand can improve, or the category can change around it.","keywords":""},{"kind":"section","title":"Competitor insights","heading":"Where the competitors come from","section":"Documentation","crumbs":["Insights","Competitor insights","Where the competitors come from"],"url":"/docs/insights/competitor-insights#where-the-competitors-come-from","text":"Genezio extracts the competitors from the answers. These are the brands that answer engines name together with your brand. Genezio does not use a list that you supply. Refer to Extracting competitors. This is important, because the competitors that an answer engine names are not always the competitors that your sales team names. Frequently, you find a brand that you did not know about. This result is useful. You can also add competitors that you want Genezio to track in all conditions.","keywords":""},{"kind":"section","title":"Competitor insights","heading":"Use the competitor insights","section":"Documentation","crumbs":["Insights","Competitor insights","Use the competitor insights"],"url":"/docs/insights/competitor-insights#use-the-competitor-insights","text":"The practical procedure is short: 1. Find a topic where a competitor is visible and your brand is not. 2. Open the conversations and read what the answer engine said about the competitor. 3. Look at the citations. These sources are the material that gives the result. 4. Decide if your brand can be a better source for that question. This brief is much more specific than \"improve our content\".","keywords":""},{"kind":"page","title":"SWOT analysis","heading":"","section":"Documentation","crumbs":["Documentation","Insights","SWOT analysis"],"url":"/docs/insights/swot-analysis","text":"Genezio extracts a SWOT analysis from Comparer conversations to show how answer engines position your brand against competitors, with the source of each point. The Genezio SWOT analysis (Strengths, Weaknesses, Opportunities, Threats) shows how answer engines position your brand against the specific competitors that you compete with. Genezio extracts the SWOT automatically from Comparer conversations. Genezio makes the SWOT from the language that answer engines use when they compare brands head-to-head. The SWOT shows these items: - The strengths that answer engines highlight. - The weaknesses that they mention. - The areas where competitors look more favorable. - The areas where your brand can gain ground.","keywords":""},{"kind":"section","title":"SWOT analysis","heading":"Select the competitors for the comparison","section":"Documentation","crumbs":["Insights","SWOT analysis","Select the competitors for the comparison"],"url":"/docs/insights/swot-analysis#select-the-competitors-for-the-comparison","text":"For each topic, you can select the exact set of competitors that the Comparer Agent compares with your brand. You make this selection at the topic level. Genezio uses it when it automatically makes Comparer scenarios for this topic. Thus, Comparer scenarios do not compare your brand with each brand that appears in AI answers. They compare your brand with the specific competitors that are important to you. When you select the correct competitors for each topic, you get a focused SWOT that helps decisions. The SWOT then shows the real market where you compete for this subject.","keywords":""},{"kind":"section","title":"SWOT analysis","heading":"SWOT in the conversation drawer","section":"Documentation","crumbs":["Insights","SWOT analysis","SWOT in the conversation drawer"],"url":"/docs/insights/swot-analysis#swot-in-the-conversation-drawer","text":"When you open a Comparer conversation in the conversation drawer, Genezio shows a SWOT analysis directly in the view. The SWOT shows: - Strengths: what answer engines say your brand does well against the competitors in the conversation. - Weaknesses: the limits or objections that answer engines mention about your brand. - Opportunities: the areas where competitors are weak or where the AI positioning is not stable. - Threats: the areas where answer engines frame competitors more favorably than your brand. Genezio also combines the same SWOT at the scenario level and at the topic level. Thus, you can go from one conversation to the full competitive picture, and you keep the evidence below it.","keywords":""},{"kind":"section","title":"SWOT analysis","heading":"The SWOT Comparison view","section":"Documentation","crumbs":["Insights","SWOT analysis","The SWOT Comparison view"],"url":"/docs/insights/swot-analysis#the-swot-comparison-view","text":"For a view across your full competitive landscape, Genezio has a dedicated SWOT Comparison view. To open it, select Competitors SWOT in the menu. The view has a brand-vs-competitor picker. Select your brand and the competitor that you want to examine. The comparison then shows that pair. Thus, you do not read all the SWOTs at the same time. You examine only the comparisons that are important to you. Each comparison starts with an AI-written overview. This overview is a short summary of how the two brands compare. You get the conclusion first, and then you examine the four quadrants for the detail. It is the fastest way to answer the question \"How do we compare with them?\" You do not have to interpret a grid yourself. The view shows a SWOT analysis for: - your brand - each of your monitored competitors When you see the SWOTs side by side, you can find patterns: - Strengths that many competitors share. - Weaknesses that only your brand has. - Opportunities where no competitor has a good position at this time. The SWOT views show competitor logos. Thus, you recognize each brand immediately, and you do not have to read a list of names.","keywords":""},{"kind":"section","title":"SWOT analysis","heading":"Open a SWOT from all locations","section":"Documentation","crumbs":["Insights","SWOT analysis","Open a SWOT from all locations"],"url":"/docs/insights/swot-analysis#open-a-swot-from-all-locations","text":"Each time that a competitor appears in Genezio, you can open its SWOT as a drawer in the same location. You do not lose the view that you work in. Together with the SWOT, the drawer shows: - The products of the competitor: the products that answer engines relate to this competitor. - Co-mentioned products: the products that answer engines name together with the competitor. Your products can be in this list. Thus, a competitor is not only a name in a list. You get the full picture: how answer engines frame the competitor, which products it is known for, and which products answer engines compare with it. This data connects to the product-level data in Shopping.","keywords":""},{"kind":"section","title":"SWOT analysis","heading":"The sources behind each SWOT point","section":"Documentation","crumbs":["Insights","SWOT analysis","The sources behind each SWOT point"],"url":"/docs/insights/swot-analysis#the-sources-behind-each-swot-point","text":"You can trace each SWOT point to the evidence that made it. For each strength, weakness, opportunity, or threat, Genezio can show the citation paragraphs (sources) behind it. These paragraphs are the real passages from the webpages that answer engines used. Genezio highlights the matching text. Thus, you can go from a high-level claim, such as \"competitors are framed as more affordable\", directly to the webpage passage that made it. You do not have to guess where the framing came from. Perceptions have the same source tracing, and Genezio applies it to your competitive SWOT too. Thus, you can audit each strength, weakness, opportunity, and threat down to the source paragraph.","keywords":""},{"kind":"section","title":"SWOT analysis","heading":"How Genezio makes the SWOT","section":"Documentation","crumbs":["Insights","SWOT analysis","How Genezio makes the SWOT"],"url":"/docs/insights/swot-analysis#how-genezio-makes-the-swot","text":"Genezio extracts SWOT insights from Comparer Agent conversations only. Comparer prompts name the brands. Thus, answer engines give a direct head-to-head comparison. Genezio needs this signal to extract the Strengths, Weaknesses, Opportunities, and Threats. Prompter, Recommender, and Introspector conversations do not add to the SWOT. These agents measure unbranded discovery and brand perception, not direct comparison.","keywords":""},{"kind":"section","title":"SWOT analysis","heading":"Related pages","section":"Documentation","crumbs":["Insights","SWOT analysis","Related pages"],"url":"/docs/insights/swot-analysis#related-pages","text":"- Comparer Agent - Competitors - Share of Voice","keywords":""},{"kind":"section","title":"SWOT analysis","heading":"In the API","section":"Documentation","crumbs":["Insights","SWOT analysis","In the API"],"url":"/docs/insights/swot-analysis#in-the-api","text":"The public API reads the SWOT in SWOT report and SWOT statement.","keywords":""},{"kind":"page","title":"Sentiment analysis","heading":"","section":"Documentation","crumbs":["Documentation","Insights","Sentiment analysis"],"url":"/docs/insights/sentiment-analysis","text":"Learn how to read the sentiment of AI answers about your brand: why neutral is normal, why negative mentions weigh more, and how to find the cause of a change. An appearance in an answer is not always good. Sentiment tells you how answer engines describe your brand when it appears.","keywords":""},{"kind":"section","title":"Sentiment analysis","heading":"What Genezio measures","section":"Documentation","crumbs":["Insights","Sentiment analysis","What Genezio measures"],"url":"/docs/insights/sentiment-analysis#what-genezio-measures","text":"Genezio reads each mention of your brand and classifies it as positive, neutral, or negative. Then, it tracks the balance over time. Most mentions are neutral. When an answer engine gives a list of options, it does not praise them. The signal is in the ratio and in the direction of its change, not in a single answer.","keywords":""},{"kind":"section","title":"Sentiment analysis","heading":"Read the sentiment","section":"Documentation","crumbs":["Insights","Sentiment analysis","Read the sentiment"],"url":"/docs/insights/sentiment-analysis#read-the-sentiment","text":"Neutral is normal, not a failure. A category list is neutral by nature. Do not try to get a profile with only positive mentions. A factual answer does not look like that. Negative mentions are more important than their share shows. One answer that calls your brand hard to set up appears exactly where a buyer is not sure. It can have more effect than ten neutral mentions in a list. Look at the trend, not at one reading. Sentiment changes when the sources change. A decrease usually means that answer engines started to cite a new source. Find this source in Most cited sources.","keywords":""},{"kind":"section","title":"Sentiment analysis","heading":"From sentiment to a specific claim","section":"Documentation","crumbs":["Insights","Sentiment analysis","From sentiment to a specific claim"],"url":"/docs/insights/sentiment-analysis#from-sentiment-to-a-specific-claim","text":"Sentiment gives you the temperature. To act, you must know the claim. Perceptions give you this claim. A perception is a specific thing that answer engines say about your brand, for example \"expensive\", \"best for small teams\", or \"hard to integrate\". You can track a perception over time and get a verdict that tells you if the claim is still true. Refer to Monitor perceptions. Use this sequence: 1. Sentiment tells you that something changed. 2. Perceptions tell you what changed. 3. Citations tell you where the change came from.","keywords":""},{"kind":"page","title":"Most cited sources","heading":"","section":"Documentation","crumbs":["Documentation","Insights","Most cited sources"],"url":"/docs/insights/most-cited-sources","text":"See the sources that answer engines cite most frequently in your category, ranked by count, and learn which of them you can influence, earn, or must accept. When an answer engine makes a claim about your category, it uses sources. This report is the list of these sources, ranked by how frequently they appear.","keywords":""},{"kind":"section","title":"Most cited sources","heading":"Why you can act on this report most easily","section":"Documentation","crumbs":["Insights","Most cited sources","Why you can act on this report most easily"],"url":"/docs/insights/most-cited-sources#why-you-can-act-on-this-report-most-easily","text":"Visibility tells you that your brand loses. Sources tell you where the answer came from. You can act on a source. A review site with forty citations in your category gives you forty opportunities. Your profile on this site can be thin, old, or missing. In each case, the task is specific and finite, and you know the result.","keywords":""},{"kind":"section","title":"Most cited sources","heading":"What to look for","section":"Documentation","crumbs":["Insights","Most cited sources","What to look for"],"url":"/docs/insights/most-cited-sources#what-to-look-for","text":"Sources that you can change directly. These sources are your own site, your documentation, and your profiles on directories and review platforms. If answer engines cite the pricing page of your competitor and not your pricing page, you must correct that page. Sources that you can earn. These sources are industry publications, comparison sites, and community threads. They take more time, but they have more weight than your own site. The reason is that an answer engine sees an independent source as better evidence. Sources that you cannot change. Some citations come from locations where no outreach can make a change. This information is useful too. It prevents work on a target where the work cannot succeed.","keywords":""},{"kind":"section","title":"Most cited sources","heading":"More detail than the domain","section":"Documentation","crumbs":["Insights","Most cited sources","More detail than the domain"],"url":"/docs/insights/most-cited-sources#more-detail-than-the-domain","text":"Frequently, a domain does not give sufficient detail for an action. \"YouTube was cited\" does not tell you what to do. When Genezio can identify the publisher of the cited material, it resolves the citation to the channel or to the author. This can be the specific YouTube channel, the specific subreddit, or the named writer. Refer to Channels and authors.","keywords":""},{"kind":"section","title":"Most cited sources","heading":"Citations from your own site","section":"Documentation","crumbs":["Insights","Most cited sources","Citations from your own site"],"url":"/docs/insights/most-cited-sources#citations-from-your-own-site","text":"Citations from your own site are good, but they have limits. Answer engines give more weight to independent sources, because the description of a brand by the brand itself is not evidence. A citation profile that has only your own pages is not stable. One competitor that gets third-party coverage can replace it.","keywords":""},{"kind":"page","title":"Channels and authors","heading":"","section":"Documentation","crumbs":["Documentation","Insights","Channels and authors"],"url":"/docs/insights/channels-and-authors","text":"Genezio resolves each citation to the YouTube channel, the subreddit, or the author behind it, so you know which creators and communities shape AI answers. A citation tells you which site an answer engine used. Frequently, this information is not sufficiently specific for an action. \"YouTube was cited\" does not help you. \" This channel was cited eleven times in answers about your category\" helps you, because a channel has an owner that you can contact. The same is true for Reddit. The site tells you nothing, but the subreddit tells you where persons discuss your brand. Genezio resolves citations to the channel and to the author behind them.","keywords":""},{"kind":"section","title":"Channels and authors","heading":"Where to find the data","section":"Documentation","crumbs":["Insights","Channels and authors","Where to find the data"],"url":"/docs/insights/channels-and-authors#where-to-find-the-data","text":"Open a citation and look at the domain. When Genezio can identify the publisher of the cited material, the drawer shows Channels & authors. It also shows how frequently each channel and each author appeared. You get these items: - Channels: the YouTube channel or the subreddit, not only the domain. - Authors: the person who wrote the piece, when the source gives a name. Genezio counts each channel and each author. Thus, you can see the difference between a single mention and a pattern.","keywords":""},{"kind":"section","title":"Channels and authors","heading":"What to do with the data","section":"Documentation","crumbs":["Insights","Channels and authors","What to do with the data"],"url":"/docs/insights/channels-and-authors#what-to-do-with-the-data","text":"Find the creators who have an effect on your category. Three answers about your brand can cite the same YouTube channel. In this case, that channel is part of how AI describes your market. This target is much more precise than \"do some YouTube marketing\". Find the communities that are important. When a subreddit appears again and again, your buyers already discuss there the problem that you solve. When you know the subreddit, you can be present in the correct location. Without this data, you have only a general social plan. Find a single author with a large effect. One author can be behind many citations. In this case, your reputation in AI answers is partly in the control of this author. This is important to know, for positive and for negative content. Use the evidence to set the priority of your outreach. A channel with eleven citations is a better use of a week than a channel with one citation.","keywords":""},{"kind":"section","title":"Channels and authors","heading":"Coverage","section":"Documentation","crumbs":["Insights","Channels and authors","Coverage"],"url":"/docs/insights/channels-and-authors#coverage","text":"Genezio cannot attribute each citation. To resolve a channel or an author, the source must identify it. Thus, the panel appears only where this information exists. Where the information does not exist, the panel does not appear. This view gives more detail about a part of your citations. It does not replace the full citation reports.","keywords":""},{"kind":"page","title":"Monitor perceptions","heading":"","section":"Documentation","crumbs":["Documentation","Insights","Monitor perceptions"],"url":"/docs/insights/monitor-perceptions","text":"Track a claim that answer engines make about your brand, such as 'expensive' or 'hard to set up', and get a verdict and a history each time Genezio checks it. A perception is a claim that answer engines make about your brand, for example \"expensive\", \"best for small teams\", or \"hard to set up\". The AI Perception Summary shows what they say now. The monitor lets you manage this claim over time. You do not only read the summary and hope that a claim disappears. You put the claim under observation, and Genezio continues to check it.","keywords":""},{"kind":"section","title":"Monitor perceptions","heading":"Why you monitor a perception","section":"Documentation","crumbs":["Insights","Monitor perceptions","Why you monitor a perception"],"url":"/docs/insights/monitor-perceptions#why-you-monitor-a-perception","text":"Some claims are much more important than their frequency shows. A single answer that calls your brand \"hard to integrate\" can cost more than ten answers that call it \"reliable\". The reason is that this claim appears exactly where a buyer is not sure. The monitor tells you if your work really changed what the answer engines say. For example, this work can be the documentation that you wrote again or the comparison page that you published. A claim that you want to be true is also worth a monitor. For example, you want your brand to be the option for regulated industries. In this case, you want to know when answer engines start to say this without help.","keywords":""},{"kind":"section","title":"Monitor perceptions","heading":"How to monitor a perception","section":"Documentation","crumbs":["Insights","Monitor perceptions","How to monitor a perception"],"url":"/docs/insights/monitor-perceptions#how-to-monitor-a-perception","text":"1. Open Perceptions for your brand. 2. Track the claim that is important to you. 3. Come back after your next content release, or select Refresh to check again immediately. Genezio keeps each check. Thus, a tracked perception has a history: what the answer engines said, and when it changed.","keywords":""},{"kind":"section","title":"Monitor perceptions","heading":"Read the result","section":"Documentation","crumbs":["Insights","Monitor perceptions","Read the result"],"url":"/docs/insights/monitor-perceptions#read-the-result","text":"A tracked perception gives you a verdict, not a general impression. The verdict tells you clearly if the claim agrees with what the answer engines say about your brand now. A summary alone cannot give you these two results: - Proof that your work had an effect. For example: \"We wrote the onboarding again in March, and the 'hard to set up' claim stopped in April.\" You can show this sentence to a board. - An early alert about a reversal. A claim that you corrected can come back. Usually, the cause is a new source that answer engines start to cite. With the monitor, you find the problem in weeks, not at the next quarterly review.","keywords":""},{"kind":"section","title":"Monitor perceptions","heading":"Read the perceptions from a different tool","section":"Documentation","crumbs":["Insights","Monitor perceptions","Read the perceptions from a different tool"],"url":"/docs/insights/monitor-perceptions#read-the-perceptions-from-a-different-tool","text":"Tracked perceptions and their verdicts are available over MCP and through the API. Thus, a perception check can be part of a larger report. Nobody must remember to open it.","keywords":""},{"kind":"page","title":"Content opportunities","heading":"","section":"Documentation","crumbs":["Documentation","Insights","Content opportunities"],"url":"/docs/insights/content-opportunities","text":"A content opportunity is a buyer question where answer engines cite a competitor and not your brand. Learn how Genezio finds them and how to use them. A content opportunity is a question that your buyers ask answer engines, where a different brand is present and your brand is not.","keywords":""},{"kind":"section","title":"Content opportunities","heading":"How Genezio finds opportunities","section":"Documentation","crumbs":["Insights","Content opportunities","How Genezio finds opportunities"],"url":"/docs/insights/content-opportunities#how-genezio-finds-opportunities","text":"Genezio already knows three things: - The questions that persons ask in your category. - The questions where your brand appears. - The sources that answer engines cite when they answer. An opportunity is a location where these three things do not agree. There is real demand, a competitor is present, and your brand is absent. This is a better start than a keyword list. It starts from answers that answer engines already give, not from search volume.","keywords":""},{"kind":"section","title":"Content opportunities","heading":"When an opportunity is worth the work","section":"Documentation","crumbs":["Insights","Content opportunities","When an opportunity is worth the work"],"url":"/docs/insights/content-opportunities#when-an-opportunity-is-worth-the-work","text":"Persons really ask the question. The question comes from real conversations, not from a guess about what persons possibly search. A competitor wins the question. When an answer engine recommends a brand, it shows that the engine is ready to recommend a brand here. Your brand can be a credible answer. This is the honest filter. If the question is about a capability that your product does not have, visibility does not help you. It can also cause damage.","keywords":""},{"kind":"section","title":"Content opportunities","heading":"Change an opportunity into work","section":"Documentation","crumbs":["Insights","Content opportunities","Change an opportunity into work"],"url":"/docs/insights/content-opportunities#change-an-opportunity-into-work","text":"Open the conversations behind the opportunity and read what the answer engine said. Look at the citations. These sources do the work now, and they show the shape of the content that can win against them. Then, Content Hub can change a topic directly into a brief, with the fanouts and the sources attached. Refer to From data to content strategy.","keywords":""},{"kind":"section","title":"Content opportunities","heading":"What not to do","section":"Documentation","crumbs":["Insights","Content opportunities","What not to do"],"url":"/docs/insights/content-opportunities#what-not-to-do","text":"Do not write a page for each opportunity. Answer engines take passages from pages. A thin page that you write only to fill a gap gives them nothing to take. One large piece that fully answers a group of related questions is better than five pages that each answer one question partially.","keywords":""},{"kind":"page","title":"Actionable Insights","heading":"","section":"Documentation","crumbs":["Documentation","Insights","Actionable Insights"],"url":"/docs/insights/actionable-insights","text":"Learn how Genezio Actionable Insights turn conversations, citations and perceptions into prioritized steps that improve your AI visibility and recommendations. Actionable Insights are the strategic recommendations that Genezio makes from your AI visibility data: conversations, citations, perceptions, competitors, and visibility metrics. Each insight tells you which action can improve the AI Recommendations and the AI Visibility of your brand. You do not have to analyze conversations and sources manually. Genezio continuously analyzes all the collected data and makes recommendations that you can act on.","keywords":""},{"kind":"section","title":"Actionable Insights","heading":"How Genezio makes insights","section":"Documentation","crumbs":["Insights","Actionable Insights","How Genezio makes insights"],"url":"/docs/insights/actionable-insights#how-genezio-makes-insights","text":"Genezio analyzes all the data that it collected for a brand. This data includes: - conversations - follow-up searches (query fanouts) - citations (sources) - perceptions - competitors - visibility metrics - signals about the accuracy of perceptions - the relevance of topics and scenarios This analysis runs automatically each day. When Genezio runs new conversations and collects new data, it updates the insights. Thus, the insights show the most recent state of the AI ecosystem.","keywords":""},{"kind":"section","title":"Actionable Insights","heading":"Types of insights","section":"Documentation","crumbs":["Insights","Actionable Insights","Types of insights"],"url":"/docs/insights/actionable-insights#types-of-insights","text":"Genezio puts insights into four categories. Each category shows a different type of improvement opportunity.","keywords":""},{"kind":"section","title":"Actionable Insights","heading":"Growth Opportunities","section":"Documentation","crumbs":["Insights","Actionable Insights","Growth Opportunities"],"url":"/docs/insights/actionable-insights#growth-opportunities","text":"Growth Opportunities are insights that show areas where your brand can get more visibility. Examples: - topics where competitors are dominant - scenarios where answer engines frequently recommend brands, but not your brand - new query patterns that Genezio found in query fanouts These insights help you find new opportunities to grow.","keywords":""},{"kind":"section","title":"Actionable Insights","heading":"Critical Visibility Gaps","section":"Documentation","crumbs":["Insights","Actionable Insights","Critical Visibility Gaps"],"url":"/docs/insights/actionable-insights#critical-visibility-gaps","text":"Critical Visibility Gaps are insights that show areas where competitors have much better results than your brand. Examples: - scenarios where competitors appear consistently but your brand does not - decision moments where your brand is absent - topics where competitors are dominant in AI recommendations These insights frequently have the label high priority, because they have a direct effect on visibility.","keywords":""},{"kind":"section","title":"Actionable Insights","heading":"Citation & Authority Leverage","section":"Documentation","crumbs":["Insights","Actionable Insights","Citation & Authority Leverage"],"url":"/docs/insights/actionable-insights#citation-authority-leverage","text":"Citation & Authority Leverage insights are about the sources that have an effect on AI answers. Examples: - influential websites that cite competitors but not your brand - authoritative sources that shape AI answers - opportunities to improve the citation coverage When your brand is present in these sources, answer engines are more likely to refer to your brand. To learn why, read How LLMs select sources.","keywords":""},{"kind":"section","title":"Actionable Insights","heading":"Positioning & Structural Optimization","section":"Documentation","crumbs":["Insights","Actionable Insights","Positioning & Structural Optimization"],"url":"/docs/insights/actionable-insights#positioning-structural-optimization","text":"Positioning & Structural Optimization insights analyze how answer engines describe your brand. Examples: - narratives in AI answers - strengths or weaknesses that answer engines mention many times - gaps in your messages, compared with competitors These insights help teams improve the position of their brand in the information ecosystem.","keywords":""},{"kind":"section","title":"Actionable Insights","heading":"Automatic insights","section":"Documentation","crumbs":["Insights","Actionable Insights","Automatic insights"],"url":"/docs/insights/actionable-insights#automatic-insights","text":"Genezio makes insights automatically each day. The system analyzes the full dataset of conversations and extracts patterns that can show: - competitive gaps - citation weaknesses - narrative problems - growth opportunities New insights can appear when Genezio runs new conversations and when AI answers change.","keywords":""},{"kind":"section","title":"Actionable Insights","heading":"Make insights manually","section":"Documentation","crumbs":["Insights","Actionable Insights","Make insights manually"],"url":"/docs/insights/actionable-insights#make-insights-manually","text":"You can also make insights on demand. On the Insights page, click one of the insight category cards: - Growth Opportunities - Critical Visibility Gaps - Citation & Authority Leverage - Positioning & Structural Optimization When you make insights manually, you can give more context. This context is optional. For example: - a specific topic - a specific scenario - more instructions or a strategic focus For example, you can ask for insights about: - the improvement of visibility in a specific product category - the actions with the highest effect - quick wins Thus, teams can adapt insights to their specific strategic needs.","keywords":""},{"kind":"section","title":"Actionable Insights","heading":"The parts of an insight","section":"Documentation","crumbs":["Insights","Actionable Insights","The parts of an insight"],"url":"/docs/insights/actionable-insights#the-parts-of-an-insight","text":"Each insight has: - a summary of the problem or opportunity - the reason that Genezio found it - the related topics or scenarios - the priority level (for example, High or Medium) Insights are actionable. They describe specific improvements that can have an effect on AI Recommendations and AI Visibility.","keywords":""},{"kind":"section","title":"Actionable Insights","heading":"How Genezio shows insights","section":"Documentation","crumbs":["Insights","Actionable Insights","How Genezio shows insights"],"url":"/docs/insights/actionable-insights#how-genezio-shows-insights","text":"Each insight has a structure that you can scan, and you can share it with stakeholders without a rewrite. You do not have to read it paragraph by paragraph.","keywords":""},{"kind":"section","title":"Actionable Insights","heading":"Diagrams where they help","section":"Documentation","crumbs":["Insights","Actionable Insights","Diagrams where they help"],"url":"/docs/insights/actionable-insights#diagrams-where-they-help","text":"Some insights are about a process, a comparison, or a chain of cause and effect. For these insights, Genezio shows a diagram in the insight. You do not read a long text about how three items connect. You see the connection. Thus, you understand the situation faster. You can also put the insight into a stakeholder deck or an executive brief, and you do not have to draw it again.","keywords":""},{"kind":"section","title":"Actionable Insights","heading":"Recommended Actions, formatted for scanning","section":"Documentation","crumbs":["Insights","Actionable Insights","Recommended Actions, formatted for scanning"],"url":"/docs/insights/actionable-insights#recommended-actions-formatted-for-scanning","text":"Each insight ends with Recommended Actions. These are the specific things to do next. They have this format: - Icons that show the type of action (content, citation outreach, configuration change, and others). - Answer engine badges that show on which answer engines Genezio saw the signal. Thus, you see immediately what to do and where the evidence came from. For example, Genezio saw the behavior behind an insight on ChatGPT and Perplexity. The insight then shows these two badges, and you know on which answer engines the recommendation is based.","keywords":""},{"kind":"section","title":"Actionable Insights","heading":"Answer engine icons in all locations","section":"Documentation","crumbs":["Insights","Actionable Insights","Answer engine icons in all locations"],"url":"/docs/insights/actionable-insights#answer-engine-icons-in-all-locations","text":"Answer engine icons also appear in insight tooltips and in the charts near the insights. When you put the pointer on a number or a trend, you see which answer engines contributed. You do not have to look at a legend.","keywords":""},{"kind":"section","title":"Actionable Insights","heading":"Why the structure is important","section":"Documentation","crumbs":["Insights","Actionable Insights","Why the structure is important"],"url":"/docs/insights/actionable-insights#why-the-structure-is-important","text":"Insights is the part of the platform that tells you what to do. Diagrams, action icons, and answer engine badges give each insight a clear structure. Thus, an insight is not a paragraph of analyst notes. It is more like a takeaway slide that you can scan, share, and act on.","keywords":""},{"kind":"section","title":"Actionable Insights","heading":"Manage insights","section":"Documentation","crumbs":["Insights","Actionable Insights","Manage insights"],"url":"/docs/insights/actionable-insights#manage-insights","text":"You can manage insights directly in the Insights interface. You can mark an insight as done or you can dismiss it.","keywords":""},{"kind":"section","title":"Actionable Insights","heading":"Mark as Done","section":"Documentation","crumbs":["Insights","Actionable Insights","Mark as Done"],"url":"/docs/insights/actionable-insights#mark-as-done","text":"When you complete the recommended action, mark the insight as Done. Thus, the team knows which opportunities it already addressed.","keywords":""},{"kind":"section","title":"Actionable Insights","heading":"Dismiss","section":"Documentation","crumbs":["Insights","Actionable Insights","Dismiss"],"url":"/docs/insights/actionable-insights#dismiss","text":"If an insight is not relevant, or if you cannot act on it, dismiss it. For example, you can dismiss an insight in these conditions: - You cannot do the recommendation. - The insight does not apply to the strategy of the brand. - The problem is outside the control of the organization. When you dismiss an insight, Genezio removes it from the active list. Thus, teams can put their attention on more relevant opportunities. When you manage insights in this way, the insights become a workflow of improvements, not a static report. To change an insight into content, read From data to content strategy.","keywords":""},{"kind":"section","title":"Actionable Insights","heading":"Why Actionable Insights are important","section":"Documentation","crumbs":["Insights","Actionable Insights","Why Actionable Insights are important"],"url":"/docs/insights/actionable-insights#why-actionable-insights-are-important","text":"The AI visibility landscape is complex. Conversations, citations, competitors, and perceptions all have an effect on each other, and together they shape AI answers. Insights change this complex data into clear recommendations. Teams do not have to analyze hundreds of conversations manually. They can put their effort on the actions that are most likely to improve their visibility in answer engines.","keywords":""},{"kind":"section","title":"Actionable Insights","heading":"Next steps","section":"Documentation","crumbs":["Insights","Actionable Insights","Next steps"],"url":"/docs/insights/actionable-insights#next-steps","text":"To understand how Genezio collects and structures the data behind insights, read these pages: - Conversations - Citations - Perceptions - Competitors","keywords":""},{"kind":"page","title":"Content Hub","heading":"","section":"Documentation","crumbs":["Documentation","Content Hub","Content Hub"],"url":"/docs/content-hub","text":"An overview of the Content Hub guides: write content that answer engines cite, and analyze the pages that you already have against the questions of your buyers. The Content Hub helps you make content that answer engines cite. It also helps you analyze the content that you already have.","keywords":""},{"kind":"section","title":"Content Hub","heading":"Write content","section":"Documentation","crumbs":["Content Hub","Content Hub","Write content"],"url":"/docs/content-hub#write-content","text":"Use these guides in this sequence: 1. Start from the data. 2. Make a brief from the query fanouts. 3. Generate the article. 4. Edit the article. 5. Publish the content.","keywords":""},{"kind":"section","title":"Content Hub","heading":"Analyze content","section":"Documentation","crumbs":["Content Hub","Content Hub","Analyze content"],"url":"/docs/content-hub#analyze-content","text":"The Content Analyzer examines a page against the questions that your buyers ask. It also examines these items: - Landing pages. The analyzer reads them in a different way. - Full lists of URLs.","keywords":""},{"kind":"section","title":"Content Hub","heading":"The test for each page","section":"Documentation","crumbs":["Content Hub","Content Hub","The test for each page"],"url":"/docs/content-hub#the-test-for-each-page","text":"One test applies to all of this section: can one paragraph of the page answer the question alone? Engines lift passages from pages. If a page has no passage that an engine can lift, the engine does not cite it. This is true even when the page is easy to read.","keywords":""},{"kind":"page","title":"Content Hub","heading":"","section":"Documentation","crumbs":["Documentation","Content Hub","Content Hub"],"url":"/docs/content-hub/content-hub","text":"Use the Genezio Content Hub to write AI-optimized articles and briefs from your topics and scenarios, then edit, export, publish and track them in AI citations. The Content Hub is the part of Genezio where you make AI-assisted articles that improve the visibility of your brand in answer engines. You generate, edit, refine, export, and publish these articles in one workspace. Usually, you make these articles to address the opportunities that Actionable Insights found. Frequently, an insight recommends specific content that makes your brand stronger in AI answers. For example, an insight can recommend: - an article that explains a specific product - a comparison page - a guide about a scenario that users frequently search for From the same entry point, you can also make briefs. A brief is a structured outline that you can give to a writer or an agency. See Briefs.","keywords":""},{"kind":"section","title":"Content Hub","heading":"Why content is important for AI recommendations","section":"Documentation","crumbs":["Content Hub","Content Hub","Why content is important for AI recommendations"],"url":"/docs/content-hub/content-hub#why-content-is-important-for-ai-recommendations","text":"Answer engines make their answers from information on the web. When authoritative content clearly explains a product, compares options, or answers frequent questions, answer engines are more likely to refer to it. The Content Hub helps you make content that is specially designed to have an effect on AI answers.","keywords":""},{"kind":"section","title":"Content Hub","heading":"The writing workspace","section":"Documentation","crumbs":["Content Hub","Content Hub","The writing workspace"],"url":"/docs/content-hub/content-hub#the-writing-workspace","text":"Articles and briefs open in the same workspace. It is one full-screen view with two halves: - an AI writing assistant chat - a rich text editor The two halves fill the screen. Thus, you do not go between a separate \"generate\" step and an \"edit\" step. All the work is in one location. This design has two results: - You talk to the assistant before Genezio writes anything. You do not complete a form and hope for a good result. First, you chat with the writing assistant about the angle, the structure, and the emphasis. Genezio starts the draft only when you agree on the content of the piece. - A brief becomes an article in the same location. When you work on a brief, you use Generate Article from Brief in the same screen. You do not export, import again, or use a second entry point. See Briefs.","keywords":""},{"kind":"section","title":"Content Hub","heading":"Optimized by design","section":"Documentation","crumbs":["Content Hub","Content Hub","Optimized by design"],"url":"/docs/content-hub/content-hub#optimized-by-design","text":"The Content Hub applies the Content Analyzer to the content that it generates. Thus, the same checks that you can run after the draft shape the output. The objective is content that is optimized when Genezio writes it, not repaired after.","keywords":""},{"kind":"section","title":"Content Hub","heading":"Make a new article","section":"Documentation","crumbs":["Content Hub","Content Hub","Make a new article"],"url":"/docs/content-hub/content-hub#make-a-new-article","text":"Open Content Hub. Click New Content. Select Article as the content type. If you select Brief, you get a structured outline. See Briefs. The New Content button replaced the old New Article button, because this entry point makes a brief or an article. The procedure to make an article has three parts.","keywords":""},{"kind":"section","title":"Content Hub","heading":"1. Topic & Scenarios","section":"Documentation","crumbs":["Content Hub","Content Hub","1. Topic & Scenarios"],"url":"/docs/content-hub/content-hub#1-topic-scenarios","text":"First, select the topic and one or more scenarios for the article. These are the same topics and scenarios that Genezio uses when it runs conversations. You can select many scenarios for one piece of content. Thus, one asset can address many user intents at the same time. This is useful when one page (for example, a comparison or an FAQ) answers more than one scenario. This step makes sure that the article agrees with the real user questions that answer engines already get. You can also select the article type (for example, blog post) and apply a template. See Content templates.","keywords":""},{"kind":"section","title":"Content Hub","heading":"2. Style & Length","section":"Documentation","crumbs":["Content Hub","Content Hub","2. Style & Length"],"url":"/docs/content-hub/content-hub#2-style-length","text":"Next, specify how Genezio writes the article. You can specify: - Tone of voice (for example: professional, friendly, analytical) - Content length (short, medium, long, or in-depth) - Language These settings adapt the article to your target audience.","keywords":""},{"kind":"section","title":"Content Hub","heading":"3. Additional Details","section":"Documentation","crumbs":["Content Hub","Content Hub","3. Additional Details"],"url":"/docs/content-hub/content-hub#3-additional-details","text":"Last, you can give more context to the AI. For example: - Target audience - Focus query fanouts: the specific follow-up searches that the article must address - Website URLs that Genezio uses as reference sources - Additional instructions about the writing style or the focus When you configure these parameters, the piece opens in the writing workspace. There, you chat with the AI writing assistant about what you want before Genezio writes a draft. Thus, the generation starts from a direction that you agreed on, not from a guess.","keywords":""},{"kind":"section","title":"Content Hub","heading":"Content templates","section":"Documentation","crumbs":["Content Hub","Content Hub","Content templates"],"url":"/docs/content-hub/content-hub#content-templates","text":"A template is a document structure that you can use again. New content gets this structure automatically. You can: - upload your own template documents to define the shape of a content type - select a saved template when you make new content - apply templates to articles and briefs Templates are useful for teams that make content with a consistent shape. Examples are case studies, comparison pages, FAQ pages, landing pages, and agency brief formats. You define the shape one time and use it again for each new piece. Each new article or brief that you generate from a template starts from that structure, not from an empty page.","keywords":""},{"kind":"section","title":"Content Hub","heading":"Edit articles","section":"Documentation","crumbs":["Content Hub","Content Hub","Edit articles"],"url":"/docs/content-hub/content-hub#edit-articles","text":"After Genezio generates an article, the article fills the editor half of the writing workspace. The assistant chat is next to it. In this editor, you can change the article directly in the browser. You can: - edit text - change the structure of sections - change the formatting - add or remove content The editor operates like a traditional writing environment and gives you full control of the article. A rich-text toolbar gives headings, lists, links, emphasis, and other formatting. Thus, you do not have to guess the markup when you edit. See Edit articles.","keywords":""},{"kind":"section","title":"Content Hub","heading":"Conversational editing","section":"Documentation","crumbs":["Content Hub","Content Hub","Conversational editing"],"url":"/docs/content-hub/content-hub#conversational-editing","text":"You can also edit the article in a conversation. Use the assistant panel next to the editor. It is the same chat that you used to shape the article before the draft. You can ask Genezio to: - change the tone or the style - expand some sections - add new sections - make explanations simpler - improve SEO - verify factual information At this step, the edits are fully conversational. You can chat with the article and ask the system to improve it.","keywords":""},{"kind":"section","title":"Content Hub","heading":"Article data and SEO optimization","section":"Documentation","crumbs":["Content Hub","Content Hub","Article data and SEO optimization"],"url":"/docs/content-hub/content-hub#article-data-and-seo-optimization","text":"The editor also shows structured metadata about the article: - SEO metadata - content score - keyword density - internal and external links - structural analysis These signals help you make sure that the article has a good structure and is optimized for discoverability.","keywords":""},{"kind":"section","title":"Content Hub","heading":"Export articles as PDF","section":"Documentation","crumbs":["Content Hub","Content Hub","Export articles as PDF"],"url":"/docs/content-hub/content-hub#export-articles-as-pdf","text":"You can export articles from the editor. The export gives you the article as a PDF. You can then share it, review it, or publish it in a different location.","keywords":""},{"kind":"section","title":"Content Hub","heading":"Publish articles and track AI citations","section":"Documentation","crumbs":["Content Hub","Content Hub","Publish articles and track AI citations"],"url":"/docs/content-hub/content-hub#publish-articles-and-track-ai-citations","text":"When you publish an article on your website, you can record its URL in Genezio. Then Genezio can: - monitor the article as part of the content footprint of your brand - monitor if the article appears in AI citations - include the article in the My Citations section When Genezio monitors published articles, you can measure if the new content improves AI Recommendations and AI Visibility over time.","keywords":""},{"kind":"section","title":"Content Hub","heading":"Content Hub and Actionable Insights","section":"Documentation","crumbs":["Content Hub","Content Hub","Content Hub and Actionable Insights"],"url":"/docs/content-hub/content-hub#content-hub-and-actionable-insights","text":"The Content Hub has a close connection with Actionable Insights. Many insights recommend specific articles to address: - topics that your content does not cover - citation gaps - queries where competitors are dominant With the Content Hub, you can quickly change these insights into concrete content actions. For a playbook for each type of signal, read From data to content strategy.","keywords":""},{"kind":"section","title":"Content Hub","heading":"Next steps","section":"Documentation","crumbs":["Content Hub","Content Hub","Next steps"],"url":"/docs/content-hub/content-hub#next-steps","text":"To understand how insights recommend content opportunities, read Actionable Insights. To understand how citations monitor published content, read Citations.","keywords":""},{"kind":"section","title":"Content Hub","heading":"In the API","section":"Documentation","crumbs":["Content Hub","Content Hub","In the API"],"url":"/docs/content-hub/content-hub#in-the-api","text":"The public API reads articles in Articles and templates in Templates.","keywords":""},{"kind":"page","title":"From data to content strategy","heading":"","section":"Documentation","crumbs":["Documentation","Content Hub","From data to content strategy"],"url":"/docs/content-hub/from-data-to-content-strategy","text":"A practical GEO content strategy guide: change the AI visibility, citation and perception signals of Genezio into a content plan with four playbooks. This guide tells SEO specialists and content marketers how to change the AI visibility signals of Genezio into a GEO content strategy. It gives one content playbook for each type of signal, so that your content plan has a real effect on AI answers.","keywords":""},{"kind":"section","title":"From data to content strategy","heading":"The four signals of Genezio","section":"Documentation","crumbs":["Content Hub","From data to content strategy","The four signals of Genezio"],"url":"/docs/content-hub/from-data-to-content-strategy#the-four-signals-of-genezio","text":"Genezio gives you four types of signal. When you understand what each signal tells you, you know which content to make. AI Recommendations % and AI Visibility %: Which topics have good scores? Which topics have low visibility, although your SEO rankings are strong? These gaps are your priority list. Citations: Which URLs do answer engines use when they recommend competitors and not you? You must get a feature on these domains, or outrank them. Perceptions: Which specific claims do answer engines make about your brand? Make the positive claims stronger. Correct the negative or old claims with better content. Actionable Insights: Recommendations that Genezio already analyzed: growth opportunities, critical gaps, citation sources to target, and positioning problems to repair.","keywords":""},{"kind":"section","title":"From data to content strategy","heading":"Content playbooks for each signal type","section":"Documentation","crumbs":["Content Hub","From data to content strategy","Content playbooks for each signal type"],"url":"/docs/content-hub/from-data-to-content-strategy#content-playbooks-for-each-signal-type","text":"","keywords":""},{"kind":"section","title":"From data to content strategy","heading":"Playbook 1: Low AI Visibility on a key topic","section":"Documentation","crumbs":["Content Hub","From data to content strategy","Playbook 1: Low AI Visibility on a key topic"],"url":"/docs/content-hub/from-data-to-content-strategy#playbook-1-low-ai-visibility-on-a-key-topic","text":"Signal: Your AI Visibility % is less than 20% on a topic where your SEO already ranks. HIGH PRIORITY. What occurs: Answer engines know that your page exists, but they do not trust it sufficiently to cite you in their answers. You can rank 1 on Google and still be absent from the answer of ChatGPT. The cause is that the third-party sources that answer engines use do not refer to you. To learn how this occurs, read How AI citations work. What to make: - In-depth pillar pages - FAQ content for the topic - Guest posts on authoritative sites in the category - Data studies or original research, which answer engines frequently cite Tip for SEO specialists: Look at the Citations section of Genezio for this topic. The sites that answer engines cite in place of you are your list of targets for PR and link building. One mention from a domain that answer engines trust can increase your AI Visibility % in some weeks.","keywords":""},{"kind":"section","title":"From data to content strategy","heading":"Playbook 2: Answer engines cite a competitor, but not you","section":"Documentation","crumbs":["Content Hub","From data to content strategy","Playbook 2: Answer engines cite a competitor, but not you"],"url":"/docs/content-hub/from-data-to-content-strategy#playbook-2-answer-engines-cite-a-competitor-but-not-you","text":"Signal: The Citations section shows competitor URLs from sources where you are not present. MEDIUM-HIGH PRIORITY. What occurs: When an answer engine answers a question in your category, it refers to specific third-party sources: review sites, industry blogs, and comparison platforms. These sources feature your competitor. They do not feature you. What to make: - Optimized profiles on G2, Capterra, and Trustpilot - Contributed articles on the domains that give citations - Press coverage in industry publications that answer engines trust - Updated case studies to offer to journalists Tip for content marketers: Do not offer only your brand. Offer a story. The publications that answer engines cite frequently usually publish original data, expert roundups, and how-to guides. Make content assets that journalists want to refer to.","keywords":""},{"kind":"section","title":"From data to content strategy","heading":"Playbook 3: Negative or old perceptions","section":"Documentation","crumbs":["Content Hub","From data to content strategy","Playbook 3: Negative or old perceptions"],"url":"/docs/content-hub/from-data-to-content-strategy#playbook-3-negative-or-old-perceptions","text":"Signal: The perceptions analysis shows that answer engines make incorrect or unfavorable claims about your brand. HIGH PRIORITY. What occurs: Answer engines learn from the content that is on the web. If the web says that your product is \"hard to set up\" or \"expensive\", this narrative stays in AI answers. It stays even if it is not true now. What to make: - \"New in [Year]\" posts that directly address the old claim - Comparison pages that address the objection directly - Customer stories that oppose the narrative - Updated answers to reviews on G2 and Capterra, to change the sentiment Tip for SEO specialists: Add FAQPage schema that directly answers \"Is [Brand] expensive?\" or \"Is [Brand] hard to use?\". The schema gives structured data that answer engines can extract and cite more easily. This data frequently has priority over older content with less structure.","keywords":""},{"kind":"section","title":"From data to content strategy","heading":"Playbook 4: Genezio found a growth opportunity","section":"Documentation","crumbs":["Content Hub","From data to content strategy","Playbook 4: Genezio found a growth opportunity"],"url":"/docs/content-hub/from-data-to-content-strategy#playbook-4-genezio-found-a-growth-opportunity","text":"Signal: Actionable Insights show a topic where competitors rank, but the category does not have sufficient content. EXPANSION OPPORTUNITY. What occurs: Genezio found a set of scenarios where answer engines consistently recommend brands, but one of these conditions is true: - The content quality is low. - There are few citations. - Your brand can realistically compete with a targeted effort. What to make: - Dedicated landing pages for the topic without sufficient content - Comparison content (\"Best X for Y use case\") - Integration or partnership content, to add more associations - Targeted outreach for citations in that specific niche Tip for content marketers: The growth opportunities that Genezio finds are frequently long-tail topics. In these topics, one strong piece of content can quickly give you authority. Give these topics priority over the most crowded terms.","keywords":""},{"kind":"section","title":"From data to content strategy","heading":"Which content to make for each agent type","section":"Documentation","crumbs":["Content Hub","From data to content strategy","Which content to make for each agent type"],"url":"/docs/content-hub/from-data-to-content-strategy#which-content-to-make-for-each-agent-type","text":"Each of the five agent types of Genezio shows a different content need: Prompter: you are not in the answers about category discovery: Category pillar pages, \"best of\" content, in-depth guides, original data Make answer engines relate your brand to the category. Recommender: answer engines do not recommend you for specific use cases: Landing pages for use cases, content for each persona, ROI calculators Be the answer to specific buyer intent queries. Introspector: answer engines describe your brand incorrectly: Updated About and Product pages, FAQ schema, press coverage, analyst coverage Correct and strengthen how answer engines describe you. Comparer: answer engines frame you unfavorably against a competitor: \"[Brand] vs [Competitor]\" comparison pages, migration guides, case studies of customers who changed brands Control the competitive narrative in AI answers. Fact Checker: answer engines state incorrect or unverifiable facts about you: Updated product, pricing, and certification pages, corrected third-party listings, authoritative reference content for high-risk claims Make sure that the correct facts are public, crawlable, and confirmed across answer engines The 80/20 rule for GEO content: 80% of the increase in your AI visibility comes from citations by the 3–5 authoritative sources that answer engines in your category trust most. Find these sources in the Citations section of Genezio. Make a feature in these sources your top content priority. For SEO specialists: A GEO content strategy does not replace your SEO strategy. It extends it. Pages that rank well and also appear in AI answers get much more total brand exposure than pages that do only one of the two. Use the data of Genezio to find your high-ranking pages that have few AI citations. Optimize these pages first.","keywords":""},{"kind":"section","title":"From data to content strategy","heading":"Next steps","section":"Documentation","crumbs":["Content Hub","From data to content strategy","Next steps"],"url":"/docs/content-hub/from-data-to-content-strategy#next-steps","text":"- To write the content, use the Content Hub. - To check a draft before you publish it, use the Content Analyzer. - To plan your first days in Genezio, read Your first week.","keywords":""},{"kind":"page","title":"Generate articles","heading":"","section":"Documentation","crumbs":["Documentation","Content Hub","Generate articles"],"url":"/docs/content-hub/generating-articles","text":"The Content Hub writes a full article from a brief, with the topic, query fanouts, audience, tone, and sources that you selected. Learn what to select. The Content Hub writes a full article from a brief. The topic, the query fanouts, and the sources that you selected are already attached.","keywords":""},{"kind":"section","title":"Generate articles","heading":"Start from a brief, not from a prompt","section":"Documentation","crumbs":["Content Hub","Generate articles","Start from a brief, not from a prompt"],"url":"/docs/content-hub/generating-articles#start-from-a-brief-not-from-a-prompt","text":"If you generate an article from an empty instruction, you get a general article. If you generate it from a brief, Genezio writes it for specific questions, with specific sources, for a specific audience. You do the important thinking in the brief. The article is the easy part.","keywords":""},{"kind":"section","title":"Generate articles","heading":"What you select","section":"Documentation","crumbs":["Content Hub","Generate articles","What you select"],"url":"/docs/content-hub/generating-articles#what-you-select","text":"- The topic or topics. A piece can cover more than one topic when a reader expects them together. See Cover several topics in one piece. - The audience. The audience changes the level, the assumptions, and the vocabulary. See Select the target audience. - The tone. See Select the tone of voice. - The sources. The writer uses this material, and the files that you attach. Genezio keeps each source with the piece. Thus, you can see later which material the piece came from.","keywords":""},{"kind":"section","title":"Generate articles","heading":"Work in your language","section":"Documentation","crumbs":["Content Hub","Generate articles","Work in your language"],"url":"/docs/content-hub/generating-articles#work-in-your-language","text":"The article and the conversation about it use the language of your brand. See Work in your brand language.","keywords":""},{"kind":"section","title":"Generate articles","heading":"After Genezio generates the article","section":"Documentation","crumbs":["Content Hub","Generate articles","After Genezio generates the article"],"url":"/docs/content-hub/generating-articles#after-genezio-generates-the-article","text":"Read the article as an editor, not as an approver. Do not ask \"is this good?\" Ask \"can an answer engine lift a passage from this to answer the question that we targeted?\" If no paragraph is a full answer alone, the article needs more work. This is true even when the article is easy to read. Then, do these steps: 1. Edit the article. 2. Examine its schema markup. 3. Publish the article.","keywords":""},{"kind":"page","title":"Briefs","heading":"","section":"Documentation","crumbs":["Documentation","Content Hub","Briefs"],"url":"/docs/content-hub/briefs","text":"Learn how to make a content brief in the Genezio Content Hub, refine it with the AI writing assistant, and change it into a full article for AI search. A brief is a short, structured outline of a piece of content. It tells what the content must cover, who it is for, and how to write it. A brief is not a finished article. In the Content Hub, a brief is the fastest path from an idea to an outline. You can give a brief to a writer, an editor, or an external agency. You can also use it as a structured starting point, and later expand it into a full article.","keywords":""},{"kind":"section","title":"Briefs","heading":"Why briefs exist","section":"Documentation","crumbs":["Content Hub","Briefs","Why briefs exist"],"url":"/docs/content-hub/briefs#why-briefs-exist","text":"A complete article is not always the correct first step. Possibly you want to: - give the work to an in-house writer or an external agency - agree on the direction before you invest in a full draft - record the intent quickly, and then continue later With a brief, you change an idea into a structured specification in one pass. You do not have to write full content to start the discussion.","keywords":""},{"kind":"section","title":"Briefs","heading":"What a content brief contains","section":"Documentation","crumbs":["Content Hub","Briefs","What a content brief contains"],"url":"/docs/content-hub/briefs#what-a-content-brief-contains","text":"A brief is a structured document. It contains: - Target audience: who the content is for. - Tone of voice: how the content must read. - Keywords: the terms that the content must rank for and make stronger. - Competitors: the brands that the content must position against or show differences from. - Scenarios: the user scenarios that the content must address. You can select many scenarios, thus one brief can address many intents at the same time. - Additional instructions: more guidance for the writer or for the next step in the workflow. Together, these fields make a specification that a person or Genezio can act on.","keywords":""},{"kind":"section","title":"Briefs","heading":"Make a brief","section":"Documentation","crumbs":["Content Hub","Briefs","Make a brief"],"url":"/docs/content-hub/briefs#make-a-brief","text":"Open Content Hub. Click New Content. Select Brief as the content type. Complete the fields of the brief: audience, tone, keywords, competitors, scenarios, and additional instructions. Generate the brief. The New Content entry point had the label New Article in the past. It makes a brief or a full article. The result depends on what you select.","keywords":""},{"kind":"section","title":"Briefs","heading":"Refine a brief with the AI writing assistant","section":"Documentation","crumbs":["Content Hub","Briefs","Refine a brief with the AI writing assistant"],"url":"/docs/content-hub/briefs#refine-a-brief-with-the-ai-writing-assistant","text":"A brief opens in the same writing workspace as an article. The AI writing assistant chat is on one side, and the rich text editor is on the other side. To refine the brief, you talk to the assistant. You can ask Genezio to: - make specific sections shorter or longer - change the tone - add or remove competitors and scenarios - make the audience targeting more precise - change the keywords This chat operates in the same way as conversational editing for articles. The brief is a working document. You can talk to it until it is ready to give to the writer.","keywords":""},{"kind":"section","title":"Briefs","heading":"Briefs and articles together","section":"Documentation","crumbs":["Content Hub","Briefs","Briefs and articles together"],"url":"/docs/content-hub/briefs#briefs-and-articles-together","text":"Briefs are in your Content Hub library together with articles. You can browse, filter, and open them from the same location, and they open in the same workspace. When you are ready to change the outline into a draft, click Generate Article from Brief. This step occurs in the same workspace. You do not export, import again, or use a second entry point. The structure of the brief (audience, tone, keywords, scenarios, instructions) becomes the starting point for the article. Thus, you do not have to enter the context again.","keywords":""},{"kind":"section","title":"Briefs","heading":"Templates and briefs","section":"Documentation","crumbs":["Content Hub","Briefs","Templates and briefs"],"url":"/docs/content-hub/briefs#templates-and-briefs","text":"Briefs use Content Hub templates. If you uploaded or selected a template, new briefs get its structure automatically. Templates are useful for teams that make briefs with a consistent shape. An example is a standard brief format that each external agency gets. For more information, see Content templates.","keywords":""},{"kind":"section","title":"Briefs","heading":"When to use a brief or an article","section":"Documentation","crumbs":["Content Hub","Briefs","When to use a brief or an article"],"url":"/docs/content-hub/briefs#when-to-use-a-brief-or-an-article","text":"You give the work to a writer or an agency: You want Genezio to write the finished draft. You want to record the intent quickly: You are ready to commit to a full piece. You do not agree on the direction yet: The direction is already clear. You want a specification that you can use again: You want an asset that you can publish The two start from the same New Content entry point. When you are ready, you can change a brief into an article.","keywords":""},{"kind":"section","title":"Briefs","heading":"Related pages","section":"Documentation","crumbs":["Content Hub","Briefs","Related pages"],"url":"/docs/content-hub/briefs#related-pages","text":"- Content Hub - Edit articles - From data to content strategy","keywords":""},{"kind":"section","title":"Briefs","heading":"In the API","section":"Documentation","crumbs":["Content Hub","Briefs","In the API"],"url":"/docs/content-hub/briefs#in-the-api","text":"The public API reads briefs in Briefs.","keywords":""},{"kind":"page","title":"Content Analyzer","heading":"","section":"Documentation","crumbs":["Documentation","Content Hub","Content Analyzer"],"url":"/docs/content-hub/content-analyzer","text":"The Genezio Content Analyzer checks if answer engines can find, trust and quote a page or a draft, and changes each finding into an action plan for the article. The Content Analyzer examines one piece of content and tells you how well answer engines can find, trust, and quote it. It also tells you exactly what to change to make the content easier to cite. The Content Analyzer is the input side of the loop. The rest of the platform measures what answer engines say about your brand. The Content Analyzer measures the pages on your side that answer engines read. The Content Analyzer replaces the old content scoring screen. The old screen gave only a score. In the Content Analyzer, each check has a plain-language takeaway and concrete recommendations. You can change these recommendations into an action plan.","keywords":""},{"kind":"section","title":"Content Analyzer","heading":"Why answer engines cite some pages and not others","section":"Documentation","crumbs":["Content Hub","Content Analyzer","Why answer engines cite some pages and not others"],"url":"/docs/content-hub/content-analyzer#why-answer-engines-cite-some-pages-and-not-others","text":"Answer engines do not cite pages because the pages rank well. They cite pages that have these properties: - The page answers a question clearly. - The page is easy to parse. - The page comes from a domain that the model already trusts. It is difficult to see these properties by eye. The Content Analyzer makes them measurable. You can use it before you publish, or on a live page that has bad results in AI answers. To learn more, read How LLMs select sources.","keywords":""},{"kind":"section","title":"Content Analyzer","heading":"What you can analyze","section":"Documentation","crumbs":["Content Hub","Content Analyzer","What you can analyze"],"url":"/docs/content-hub/content-analyzer#what-you-can-analyze","text":"You can run the analyzer on three types of input: - A URL: check a page that is already published. Genezio validates the URL first. Thus, you know immediately if Genezio cannot read the URL. You do not have to wait for the analysis. - An uploaded file: analyze a document that you work on. - Pasted draft text: paste a draft directly and check it before you publish it. The last option is the most important. With it, you can repair a piece while a change is still cheap.","keywords":""},{"kind":"section","title":"Content Analyzer","heading":"What the Content Analyzer report checks","section":"Documentation","crumbs":["Content Hub","Content Analyzer","What the Content Analyzer report checks"],"url":"/docs/content-hub/content-analyzer#what-the-content-analyzer-report-checks","text":"Each check gives a plain-language takeaway. The takeaway tells you what the result means and what to do. You do not get a number that you must interpret.","keywords":""},{"kind":"section","title":"Content Analyzer","heading":"Brand Favorability","section":"Documentation","crumbs":["Content Hub","Content Analyzer","Brand Favorability"],"url":"/docs/content-hub/content-analyzer#brand-favorability","text":"Brand Favorability shows the tone that answer engines take toward your brand in this content. A page can be technically correct and still frame your brand weakly.","keywords":""},{"kind":"section","title":"Content Analyzer","heading":"Question Coverage","section":"Documentation","crumbs":["Content Hub","Content Analyzer","Question Coverage"],"url":"/docs/content-hub/content-analyzer#question-coverage","text":"Question Coverage shows which real user questions the content answers. It shows the difference between a page that is about a topic and a page that solves the question of the user.","keywords":""},{"kind":"section","title":"Content Analyzer","heading":"Citation Standing","section":"Documentation","crumbs":["Content Hub","Content Analyzer","Citation Standing"],"url":"/docs/content-hub/content-analyzer#citation-standing","text":"Citation Standing compares the domain where you plan to publish with the domains that answer engines already cite in your space. The domain is optional. The check tells you if you publish in a location that models trust, or in a location that models never showed. If Genezio cannot reach the domain, the report says so clearly. The check does not fail silently.","keywords":""},{"kind":"section","title":"Content Analyzer","heading":"Brand Fit","section":"Documentation","crumbs":["Content Hub","Content Analyzer","Brand Fit"],"url":"/docs/content-hub/content-analyzer#brand-fit","text":"Brand Fit gives a Strong, Good, or Weak verdict on how well the content fits your brand. The card shows the reasons. Thus, you can decide if you agree.","keywords":""},{"kind":"section","title":"Content Analyzer","heading":"Structure and technical checks","section":"Documentation","crumbs":["Content Hub","Content Analyzer","Structure and technical checks"],"url":"/docs/content-hub/content-analyzer#structure-and-technical-checks","text":"Each of these checks also gives a plain-language takeaway: - GEO structure: Is the organization of the content the one that answer engines prefer? - Readability: Is the content sufficiently clear to quote? - Findability: Can users and answer engines find the content? - Schema and rendering: Can machines read the markup and the rendered output? - Content-type fit: Is this the correct format for the intent (guide, comparison, FAQ, and others)?","keywords":""},{"kind":"section","title":"Content Analyzer","heading":"The action plan","section":"Documentation","crumbs":["Content Hub","Content Analyzer","The action plan"],"url":"/docs/content-hub/content-analyzer#the-action-plan","text":"The action plan makes the report useful, not only informative. Each insight has concrete recommendations. For example: - \"Add a comparison table\" - \"Open with an answer-first definition\" - \"Add source citations\" Select the recommendations that you want to act on. Genezio collects them into an action plan. When the list is correct, click Continue in the Content Hub. The items go directly into the draft. Thus, the repairs go into the real article, not into a forgotten document.","keywords":""},{"kind":"section","title":"Content Analyzer","heading":"When content does not match a topic","section":"Documentation","crumbs":["Content Hub","Content Analyzer","When content does not match a topic"],"url":"/docs/content-hub/content-analyzer#when-content-does-not-match-a-topic","text":"If the content does not match a topic that you monitor, the report shows a \"no match\" state. This state is not a failure. It is a finding. Possibly the content is about a subject outside your monitored space. Or possibly you do not measure a topic that you must measure. See Topics.","keywords":""},{"kind":"section","title":"Content Analyzer","heading":"Export the report as a PDF","section":"Documentation","crumbs":["Content Hub","Content Analyzer","Export the report as a PDF"],"url":"/docs/content-hub/content-analyzer#export-the-report-as-a-pdf","text":"You can export the full report as a PDF. This is the easy way to give the findings to a writer, an agency, or a stakeholder who does not use Genezio.","keywords":""},{"kind":"section","title":"Content Analyzer","heading":"A typical Content Analyzer workflow","section":"Documentation","crumbs":["Content Hub","Content Analyzer","A typical Content Analyzer workflow"],"url":"/docs/content-hub/content-analyzer#a-typical-content-analyzer-workflow","text":"Open the Content Analyzer and give a URL, a file, or a pasted draft. Add the domain where you plan to publish, to get the citation standing. Read the report from the top to the bottom. Usually, Question Coverage and Brand Fit explain the most. Select the useful recommendations for the action plan. Click Continue to send the recommendations to the Content Hub. Then draft the improved piece. Publish the page. Then monitor if the page starts to appear in your citations.","keywords":""},{"kind":"section","title":"Content Analyzer","heading":"Related pages","section":"Documentation","crumbs":["Content Hub","Content Analyzer","Related pages"],"url":"/docs/content-hub/content-analyzer#related-pages","text":"- Content Hub - From data to content strategy - Citations - Topics","keywords":""},{"kind":"section","title":"Content Analyzer","heading":"In the API","section":"Documentation","crumbs":["Content Hub","Content Analyzer","In the API"],"url":"/docs/content-hub/content-analyzer#in-the-api","text":"The public API reads the results of the Content Analyzer in Content analyses.","keywords":""},{"kind":"page","title":"Analyze landing pages","heading":"","section":"Documentation","crumbs":["Documentation","Content Hub","Analyze landing pages"],"url":"/docs/content-hub/analyzing-landing-pages","text":"The Content Analyzer measures a landing page against all the topics of your brand. Learn about topic reach, quotable structure, and copy against page furniture. A landing page is not an article. If you examine a landing page as an article, you get the wrong result. A writer makes an article to answer a question. A writer makes a landing page to sell. Thus, the Content Analyzer reads the two types of page in different ways.","keywords":""},{"kind":"section","title":"Analyze landing pages","heading":"What changes","section":"Documentation","crumbs":["Content Hub","Analyze landing pages","What changes"],"url":"/docs/content-hub/analyzing-landing-pages#what-changes","text":"The Content Analyzer measures an article against one scenario. The scenario is one question, and the analyzer measures how well the article answers it. The Content Analyzer measures a landing page against all the topics of your brand at the same time. It does not use one scenario. This agrees with how a landing page gets citations. The page does not answer one question well. Answer engines find the page across the full range of questions that buyers ask. Thus, the report starts with Topic reach: can AI find this page for the questions that it gets? The report does not start with the fit to one prompt.","keywords":""},{"kind":"section","title":"Analyze landing pages","heading":"What you get","section":"Documentation","crumbs":["Content Hub","Analyze landing pages","What you get"],"url":"/docs/content-hub/analyzing-landing-pages#what-you-get","text":"The report has these parts: - Topic reach. How well a retriever finds this page across the questions that people ask AI about your topics. This part also shows the vocabulary of that demand. - Quotable structure. Whether the structure of the page lets an answer engine lift a passage. Marketing pages frequently fail here. They have strong claims, but no passage that an engine can quote. - Copy vs. page furniture. How much of the page is real prose, compared with navigation, banners, and boilerplate. A page with 12% prose gives a model little material, even when that 12% is good. - What it could support. A headline figure. It shows how much of the current AI text about your topics this page could support.","keywords":""},{"kind":"section","title":"Analyze landing pages","heading":"How to read the result","section":"Documentation","crumbs":["Content Hub","Analyze landing pages","How to read the result"],"url":"/docs/content-hub/analyzing-landing-pages#how-to-read-the-result","text":"Do not ask \"what score did the page get?\" Ask \"which of the three parts is weak?\" Each weak part needs a different type of work: - Weak topic reach is a content problem. The page is about a subject that is different from the questions of the buyers. - Weak quotable structure is a format problem. The page has the substance, but it has no passage that an engine can lift. - Weak copy vs. furniture is a design problem. Most of the page is chrome.","keywords":""},{"kind":"section","title":"Analyze landing pages","heading":"Pages that the analyzer does not read","section":"Documentation","crumbs":["Content Hub","Analyze landing pages","Pages that the analyzer does not read"],"url":"/docs/content-hub/analyzing-landing-pages#pages-that-the-analyzer-does-not-read","text":"This analysis cannot examine some types of page. For these pages, the analyzer stops. It does not give a number that looks correct but has no meaning. If you get this result, the page usually has almost no prose.","keywords":""},{"kind":"page","title":"Analyze many URLs at once","heading":"","section":"Documentation","crumbs":["Documentation","Content Hub","Analyze many URLs at once"],"url":"/docs/content-hub/analyzing-many-urls-at-once","text":"Paste a list of URLs into the Content Analyzer to get one analysis and one report for each page. Audit a site section, compare competitors, or check migrations. You seldom want to analyze only one page. For example, you want to know which of your forty product pages are weak. Or, you want to compare the full blog of a competitor with your blog. Paste a list of URLs. The analyzer makes a separate analysis for each URL.","keywords":""},{"kind":"section","title":"Analyze many URLs at once","heading":"How to run the analysis","section":"Documentation","crumbs":["Content Hub","Analyze many URLs at once","How to run the analysis"],"url":"/docs/content-hub/analyzing-many-urls-at-once#how-to-run-the-analysis","text":"1. Open the Content Analyzer. 2. Paste the addresses as full URLs (https://example.com/your-article), one URL on each line. 3. Start the run. Each page gets its own analysis and its own report. Thus, you can open the full result of one page, and you do not have to read an average. Each run accepts a fixed maximum number of URLs. If your list has more URLs than this maximum, the analyzer tells you.","keywords":""},{"kind":"section","title":"Analyze many URLs at once","heading":"What it costs","section":"Documentation","crumbs":["Content Hub","Analyze many URLs at once","What it costs"],"url":"/docs/content-hub/analyzing-many-urls-at-once#what-it-costs","text":"Each page counts separately against your daily analysis limit. Ten URLs are ten analyses, not one analysis. The analyzer shows you the count before you start. Thus, you can decide to run the full list today or to divide it. If some addresses are not correct, the analyzer tells you which addresses they are. The correct addresses run.","keywords":""},{"kind":"section","title":"Analyze many URLs at once","heading":"When to use it","section":"Documentation","crumbs":["Content Hub","Analyze many URLs at once","When to use it"],"url":"/docs/content-hub/analyzing-many-urls-at-once#when-to-use-it","text":"- Audit a section of your site. Analyze all the pages in one category, and sort them from the weakest. Then, \"our content needs work\" becomes a specific list of pages, with a specific reason for each page. - Compare your pages with a competitor. Run the pages of the competitor next to your pages on the same topics. Then, see where the pages of the competitor are more quotable. - Examine a migration. After a redesign, run the same list again and compare the results. If the redesign increased the page furniture and decreased the prose, the results show it immediately.","keywords":""},{"kind":"page","title":"Cover several topics in one piece","heading":"","section":"Documentation","crumbs":["Documentation","Content Hub","Cover several topics in one piece"],"url":"/docs/content-hub/covering-several-topics-in-one-piece","text":"Learn when to select more than one topic for a Content Hub brief or article, when not to combine topics, and how the result shows under each selected topic. A brief or an article can cover more than one topic.","keywords":""},{"kind":"section","title":"Cover several topics in one piece","heading":"Why you combine topics","section":"Documentation","crumbs":["Content Hub","Cover several topics in one piece","Why you combine topics"],"url":"/docs/content-hub/covering-several-topics-in-one-piece#why-you-combine-topics","text":"Topics in Genezio are narrow on purpose. The narrow scope lets Genezio measure visibility. Real content is frequently wider. One buyer's guide can correctly answer questions from three different topics. If you divide it into three thin pages, the pages help nobody. When you select more than one topic, Genezio writes one piece that gets citations for all of the topics. Genezio does not write one piece for each topic.","keywords":""},{"kind":"section","title":"Cover several topics in one piece","heading":"When to combine topics","section":"Documentation","crumbs":["Content Hub","Cover several topics in one piece","When to combine topics"],"url":"/docs/content-hub/covering-several-topics-in-one-piece#when-to-combine-topics","text":"Do combine topics that a reader expects to find in one location. For example, a comparison page usually covers pricing, features, and alternatives. Do not combine topics only to save time. A piece that covers unrelated questions does not answer any of them well. Answer engines lift passages. If a page is about four subjects in a vague way, it has no passage that an engine can lift. Use this test: if a person searches for topic A and finds this page, is the person satisfied? If the answer is no, write two pieces.","keywords":""},{"kind":"section","title":"Cover several topics in one piece","heading":"What changes in the result","section":"Documentation","crumbs":["Content Hub","Cover several topics in one piece","What changes in the result"],"url":"/docs/content-hub/covering-several-topics-in-one-piece#what-changes-in-the-result","text":"The brief covers the questions from all the topics that you selected. Genezio writes the article to answer these questions in one structure. Later, when you look at the topics that the piece supports, the piece shows under each of these topics.","keywords":""},{"kind":"page","title":"Schema markup","heading":"","section":"Documentation","crumbs":["Documentation","Content Hub","Schema markup"],"url":"/docs/content-hub/schema-markup","text":"Each Content Hub article includes schema.org markup. Learn why structured data matters for answer engines, and how to read, copy, edit, and paste it as JSON-LD. Each article that you generate includes its schema.org markup. This structured description tells search engines and answer engines what the page is, who wrote it, and what it is about.","keywords":""},{"kind":"section","title":"Schema markup","heading":"Why it is important","section":"Documentation","crumbs":["Content Hub","Schema markup","Why it is important"],"url":"/docs/content-hub/schema-markup#why-it-is-important","text":"Prose tells a model what you say. Structured markup tells the model what the page is. Without markup, an engine must find the type of the page from the text alone. For example, the page can be a product, a recipe, a how-to, or a review. The engine is frequently wrong. The markup also controls whether a page can get a rich result. A product page without markup cannot show a price or a rating. This is true even when the page is very good.","keywords":""},{"kind":"section","title":"Schema markup","heading":"Use the markup","section":"Documentation","crumbs":["Content Hub","Schema markup","Use the markup"],"url":"/docs/content-hub/schema-markup#use-the-markup","text":"Open an article. You can read, copy, or edit its markup. Paste the markup into your CMS as a JSON-LD block in the of the page. If your CMS has a structured-data field, paste the markup there.","keywords":""},{"kind":"section","title":"Schema markup","heading":"Edit the markup","section":"Documentation","crumbs":["Content Hub","Schema markup","Edit the markup"],"url":"/docs/content-hub/schema-markup#edit-the-markup","text":"The generated markup is a start. Sometimes you know facts that Genezio does not know. Edit the markup directly on the article. The editor tells you when the markup does not have a property that a rich result needs. For example, for a product, Google also needs offers, review, or aggregateRating before the page can get a rich result. Thus, you do not have to find the problem in a search console some weeks later.","keywords":""},{"kind":"section","title":"Schema markup","heading":"Use the markup from a different tool","section":"Documentation","crumbs":["Content Hub","Schema markup","Use the markup from a different tool"],"url":"/docs/content-hub/schema-markup#use-the-markup-from-a-different-tool","text":"An MCP client can read and edit the schema markup of an article. Thus, you can make a correction in the same conversation where you found the problem. See Genezio MCP.","keywords":""},{"kind":"page","title":"Work in your brand language","heading":"","section":"Documentation","crumbs":["Documentation","Content Hub","Work in your brand language"],"url":"/docs/content-hub/working-in-your-brand-language","text":"The Content Hub writes, chats, and reports in the language of your brand, with no translation at the end. Learn why this helps AI citations and how to set it. The Content Hub works in the language of your brand. It does not work in English and then translate at the end.","keywords":""},{"kind":"section","title":"Work in your brand language","heading":"What this means","section":"Documentation","crumbs":["Content Hub","Work in your brand language","What this means"],"url":"/docs/content-hub/working-in-your-brand-language#what-this-means","text":"- The conversation. When you chat with the Content Hub about a brief or an article, it answers in the language of your brand. You do not change to a different language to brief a writer, and then to a different language again to read the result. - The content analysis report. The analysis of a page uses the language of your brand. Thus, you can send a report to a local team, and you do not have to write it again.","keywords":""},{"kind":"section","title":"Work in your brand language","heading":"Why it is important","section":"Documentation","crumbs":["Content Hub","Work in your brand language","Why it is important"],"url":"/docs/content-hub/working-in-your-brand-language#why-it-is-important","text":"This is not only for convenience. Content in English that you translate loses the item that gets citations: the vocabulary that people use when they ask. \"Best CRM for small teams\" and its equivalent in Romanian are not the same phrase in translation. They contain different words, and answer engines match on these words. When you work in the language of the brand from the start, the piece agrees with the demand that Genezio measured. It does not agree with only an English version of that demand.","keywords":""},{"kind":"section","title":"Work in your brand language","heading":"Set the language","section":"Documentation","crumbs":["Content Hub","Work in your brand language","Set the language"],"url":"/docs/content-hub/working-in-your-brand-language#set-the-language","text":"The language comes from the brand. Set the language on the brand. Then, all the subsequent items use it: the conversation, the writing, and the analysis. If a brand operates in several markets, it is usually better to make several brands, one brand for each market. Then, each brand has its own language, topics, and competitors.","keywords":""},{"kind":"page","title":"Use query fanouts for content","heading":"","section":"Documentation","crumbs":["Documentation","Content Hub","Use query fanouts for content"],"url":"/docs/content-hub/using-query-fanouts-for-content","text":"Write content for the queries that answer engines search after a question. Learn what to take from query fanouts and how to carry them into a Content Hub brief. A query fanout is a query that an answer engine searched for after a person asked it a question. When you write for the fanouts, you write for what the retrieval looks for. You do not write for the sentence that a person typed.","keywords":""},{"kind":"section","title":"Use query fanouts for content","heading":"Why fanouts change the brief","section":"Documentation","crumbs":["Content Hub","Use query fanouts for content","Why fanouts change the brief"],"url":"/docs/content-hub/using-query-fanouts-for-content#why-fanouts-change-the-brief","text":"A question from a person and the queries behind it are different texts. For example, a person asks \"which CRM should I use for a small agency\". The engine then searches for these items: - pricing for small teams - integrations with design tools - limits on user seats A page for the question of the person covers the subject. A page for the fanouts answers the five items that the engine really tried to find. The engine matches on these items.","keywords":""},{"kind":"section","title":"Use query fanouts for content","heading":"What to take from fanouts","section":"Documentation","crumbs":["Content Hub","Use query fanouts for content","What to take from fanouts"],"url":"/docs/content-hub/using-query-fanouts-for-content#what-to-take-from-fanouts","text":"- The vocabulary. Fanouts use the words that an engine expects in a good source. If your page says \"seats\" and the fanouts say \"users\", engines retrieve your page less easily, without a good reason. - The sub-questions. One question from a buyer frequently becomes several questions. Each of these questions is a section that the piece must answer fully, not only mention. - The fanouts that you almost match. Your current pages almost answer some fanouts. These fanouts are easier wins than the fanouts that you do not match at all. Frequently, an edit is sufficient, and you do not need a new piece.","keywords":""},{"kind":"section","title":"Use query fanouts for content","heading":"Fanouts in the Content Hub","section":"Documentation","crumbs":["Content Hub","Use query fanouts for content","Fanouts in the Content Hub"],"url":"/docs/content-hub/using-query-fanouts-for-content#fanouts-in-the-content-hub","text":"You can move fanouts directly into a brief. Then, Genezio makes the outline from the fanouts, not from an assumption about the topic. The writer gets a specific list of questions to answer, not only a subject to cover. See Briefs and From data to content strategy.","keywords":""},{"kind":"section","title":"Use query fanouts for content","heading":"A caution","section":"Documentation","crumbs":["Content Hub","Use query fanouts for content","A caution"],"url":"/docs/content-hub/using-query-fanouts-for-content#a-caution","text":"Do not write one page for each fanout. Fanouts are parts of one intent. A page about one part is thin by design. Put the fanouts of one real question in a group. Then, write one large piece that answers all of them.","keywords":""},{"kind":"page","title":"Select the tone of voice","heading":"","section":"Documentation","crumbs":["Documentation","Content Hub","Select the tone of voice"],"url":"/docs/content-hub/selecting-tone-of-voice","text":"Tone controls how a piece sounds, but a bad tone can stop a good piece from being useful to answer engines. Learn which tones work and a simple test to apply. The tone controls how the piece sounds. It does not control whether engines cite the piece. But an incorrect tone can make a good piece less useful.","keywords":""},{"kind":"section","title":"Select the tone of voice","heading":"What to select","section":"Documentation","crumbs":["Content Hub","Select the tone of voice","What to select"],"url":"/docs/content-hub/selecting-tone-of-voice#what-to-select","text":"Select the tone that agrees with the situation of the reader. A person who compares tools wants claims that the person can examine. A person who learns a concept wants patience. The practical constraint for AI visibility is this: the tone must not prevent a quotable passage. An engine cannot easily lift text that prepares for three sentences before it gives a point. An engine that looks for an answer wants a paragraph that contains the answer.","keywords":""},{"kind":"section","title":"Select the tone of voice","heading":"Tones that decrease your visibility","section":"Documentation","crumbs":["Content Hub","Select the tone of voice","Tones that decrease your visibility"],"url":"/docs/content-hub/selecting-tone-of-voice#tones-that-decrease-your-visibility","text":"- Strong marketing voice. Superlatives and positioning language give an engine no facts to quote. Independent sources then rank higher than you for the same question. - Too many qualifications. A passage that qualifies each claim does not answer anything with certainty. Engines cite passages that answer with certainty. - Clever framing. Analogies and narrative introductions make the answer come later. This is acceptable for a newsletter. It is costly for a page that engines must retrieve.","keywords":""},{"kind":"section","title":"Select the tone of voice","heading":"A simple test","section":"Documentation","crumbs":["Content Hub","Select the tone of voice","A simple test"],"url":"/docs/content-hub/selecting-tone-of-voice#a-simple-test","text":"Read the piece and ask: can I copy one paragraph that answers the question alone, without the paragraph before it? If the answer is yes, the tone is correct. If the answer is no, the tone is only decoration, independent of how it sounds.","keywords":""},{"kind":"section","title":"Select the tone of voice","heading":"Consistency","section":"Documentation","crumbs":["Content Hub","Select the tone of voice","Consistency"],"url":"/docs/content-hub/selecting-tone-of-voice#consistency","text":"Keep the tone consistent across a brand. Answer engines cite your pages together. When a reader goes from one page to a different page, the reader sees if the pages sound like different companies.","keywords":""},{"kind":"page","title":"Select the target audience","heading":"","section":"Documentation","crumbs":["Documentation","Content Hub","Select the target audience"],"url":"/docs/content-hub/selecting-target-audience","text":"The audience of a piece changes what it says, what it assumes, and its vocabulary. Learn how a specific target audience makes your articles more quotable. The audience of a piece changes what the piece must say, what it can assume, and which words it uses. The Content Hub asks for the audience because a general audience gives general content. General content does not answer the question of a person exactly.","keywords":""},{"kind":"section","title":"Select the target audience","heading":"Why it is important for AI visibility","section":"Documentation","crumbs":["Content Hub","Select the target audience","Why it is important for AI visibility"],"url":"/docs/content-hub/selecting-target-audience#why-it-is-important-for-ai-visibility","text":"Answer engines lift passages that answer a question directly. A passage for \"anyone interested in CRM\" seldom answers a question exactly enough for an engine to lift it. A passage for a sales lead at an agency with twenty people answers a question that a person really asked. Engines quote this type of passage.","keywords":""},{"kind":"section","title":"Select the target audience","heading":"Select the audience","section":"Documentation","crumbs":["Content Hub","Select the target audience","Select the audience"],"url":"/docs/content-hub/selecting-target-audience#select-the-audience","text":"- Select the person who asks the question, not the person who signs the contract. These are frequently different persons. The question of the buyer and the question of the user have different vocabulary and different sources. - Use the personas that you already have. If the brand has personas, the personas give the context: the role, the constraints, and what the person already knows. Personas also keep the pieces of a team consistent. - Tell exactly what the audience already knows. The largest difference between a useful piece and a padded piece is how much basic information it does not repeat.","keywords":""},{"kind":"section","title":"Select the target audience","heading":"A frequent mistake","section":"Documentation","crumbs":["Content Hub","Select the target audience","A frequent mistake"],"url":"/docs/content-hub/selecting-target-audience#a-frequent-mistake","text":"Some teams select the widest audience to \"reach more people\". The result is a piece that explains what all readers already know, and that does not answer a specific question. Narrow pieces are more quotable. More quotable pieces get more citations.","keywords":""},{"kind":"page","title":"Edit articles","heading":"","section":"Documentation","crumbs":["Documentation","Content Hub","Edit articles"],"url":"/docs/content-hub/editing-articles","text":"Learn how to edit an AI-written article in the Genezio Content Hub, with the rich-text toolbar, conversational editing with the AI assistant, and templates. This guide tells how to edit an AI-written article in the editor of the Genezio Content Hub. After Genezio generates an article, the article fills the editor half of the writing workspace. There, you refine the draft, change the structure, and make the text better before you publish. The writing workspace is a full-screen view. The AI writing assistant chat is on one side, and a rich text editor is on the other side. Articles and briefs both open here. Thus, you have one location to write, independent of how you started. See Content Hub.","keywords":""},{"kind":"section","title":"Edit articles","heading":"Edit the text in the editor","section":"Documentation","crumbs":["Content Hub","Edit articles","Edit the text in the editor"],"url":"/docs/content-hub/editing-articles#edit-the-text-in-the-editor","text":"The editor operates like a traditional writing environment. You can: - edit text directly - change the structure of sections - add or remove content - change the formatting Genezio saves your changes while you work. The article stays in your Content Hub library together with your other articles and briefs.","keywords":""},{"kind":"section","title":"Edit articles","heading":"Format the article with the rich-text toolbar","section":"Documentation","crumbs":["Content Hub","Edit articles","Format the article with the rich-text toolbar"],"url":"/docs/content-hub/editing-articles#format-the-article-with-the-rich-text-toolbar","text":"The editor has a rich-text toolbar. Thus, you do not have to guess how to format the text. The toolbar supports: - headings (H1, H2, H3,...) - bold, italic, and other inline emphasis - bulleted and numbered lists - links - block quotes and code blocks You can format content visually, and you do not have to write markup. This is especially useful when you change the structure of an article before you give it to a different person or publish it.","keywords":""},{"kind":"section","title":"Edit articles","heading":"Conversational editing with the AI assistant","section":"Documentation","crumbs":["Content Hub","Edit articles","Conversational editing with the AI assistant"],"url":"/docs/content-hub/editing-articles#conversational-editing-with-the-ai-assistant","text":"You can also edit the article in a conversation. Use the assistant next to the editor. It is the same chat that you use to shape a piece before Genezio writes the draft. You can ask Genezio to: - change the tone or the style - expand sections or make them simpler - add new sections - improve SEO - verify factual claims","keywords":""},{"kind":"section","title":"Edit articles","heading":"Edit an article that uses a template","section":"Documentation","crumbs":["Content Hub","Edit articles","Edit an article that uses a template"],"url":"/docs/content-hub/editing-articles#edit-an-article-that-uses-a-template","text":"If Genezio generated the article from a template, the structure of the template is already in the article when the editor opens. You can edit in this structure. The headings and sections that the template defines stay in the article, unless you change them. For more information, see Content templates.","keywords":""},{"kind":"section","title":"Edit articles","heading":"Related pages","section":"Documentation","crumbs":["Content Hub","Edit articles","Related pages"],"url":"/docs/content-hub/editing-articles#related-pages","text":"- Content Hub - Briefs - Content Analyzer","keywords":""},{"kind":"section","title":"Edit articles","heading":"In the API","section":"Documentation","crumbs":["Content Hub","Edit articles","In the API"],"url":"/docs/content-hub/editing-articles#in-the-api","text":"The public API reads articles in Articles.","keywords":""},{"kind":"page","title":"Chat with your article","heading":"","section":"Documentation","crumbs":["Documentation","Content Hub","Chat with your article"],"url":"/docs/content-hub/chatting-with-your-article","text":"Each Content Hub article has a conversation attached. Ask for changes in words, attach files as source material, and see when manual edits are better. Each article in the Content Hub has a conversation attached. You can ask for changes in words. You do not have to make all the changes manually.","keywords":""},{"kind":"section","title":"Chat with your article","heading":"When to use the conversation","section":"Documentation","crumbs":["Content Hub","Chat with your article","When to use the conversation"],"url":"/docs/content-hub/chatting-with-your-article#when-to-use-the-conversation","text":"- Changes to all of the article. An instruction such as \"Make this for practitioners, not executives\" changes each paragraph. It is faster to write this instruction one time than to write each paragraph again. - Missing content. For example, ask \"Add a section on what this costs for a team of five\". The writer already has the brief, the sources, and the fanouts. - A check of the article against its purpose. Before you publish, ask a question such as \"Which of our target questions does this not answer yet?\"","keywords":""},{"kind":"section","title":"Chat with your article","heading":"When not to use the conversation","section":"Documentation","crumbs":["Content Hub","Chat with your article","When not to use the conversation"],"url":"/docs/content-hub/chatting-with-your-article#when-not-to-use-the-conversation","text":"Do not use the conversation for small, exact edits. It is faster to correct a word, a number, or a link manually. See Edit articles.","keywords":""},{"kind":"section","title":"Chat with your article","heading":"Attachments","section":"Documentation","crumbs":["Content Hub","Chat with your article","Attachments"],"url":"/docs/content-hub/chatting-with-your-article#attachments","text":"You can attach files to the conversation. Then, a style guide, a product sheet, or a research document becomes material for the writer. Genezio keeps the attached files with the article. Thus, you can see later which material the article came from.","keywords":""},{"kind":"section","title":"Chat with your article","heading":"Your language","section":"Documentation","crumbs":["Content Hub","Chat with your article","Your language"],"url":"/docs/content-hub/chatting-with-your-article#your-language","text":"The conversation uses the language of your brand. You write the brief and read the article in the same language. See Work in your brand language.","keywords":""},{"kind":"section","title":"Chat with your article","heading":"A good habit","section":"Documentation","crumbs":["Content Hub","Chat with your article","A good habit"],"url":"/docs/content-hub/chatting-with-your-article#a-good-habit","text":"After each large change, read one passage again and apply the citation test. Ask: does this passage still answer a question alone? Broad instructions sometimes make an article easier to read, but the article then answers less.","keywords":""},{"kind":"page","title":"Publish content","heading":"","section":"Documentation","crumbs":["Documentation","Content Hub","Publish content"],"url":"/docs/content-hub/publishing-content","text":"Publish Content Hub articles on your own site with their schema markup, in a format that answer engines can read, and then measure when the citations start. Genezio writes and stores content. You publish the content on your own site. The method of publication has an effect on whether an engine can cite the piece.","keywords":""},{"kind":"section","title":"Publish content","heading":"Move the content to your site","section":"Documentation","crumbs":["Content Hub","Publish content","Move the content to your site"],"url":"/docs/content-hub/publishing-content#move-the-content-to-your-site","text":"Copy the article into your CMS, or export it. The text is the easy part. Teams frequently forget the two items in the sections that follow.","keywords":""},{"kind":"section","title":"Publish content","heading":"Include the schema markup","section":"Documentation","crumbs":["Content Hub","Publish content","Include the schema markup"],"url":"/docs/content-hub/publishing-content#include-the-schema-markup","text":"Each article has its schema.org markup. Paste the markup into the page as a JSON-LD block, or into the structured-data field of your CMS. Without the markup, an engine must find the type of the page from the prose, and it is frequently wrong. With the markup, the page can also get rich results that it cannot get without it.","keywords":""},{"kind":"section","title":"Publish content","heading":"Publish it where engines can read it","section":"Documentation","crumbs":["Content Hub","Publish content","Publish it where engines can read it"],"url":"/docs/content-hub/publishing-content#publish-it-where-engines-can-read-it","text":"If answer engines cannot retrieve a page, they cannot cite it. This is true even when the page is very good. Obey these rules: - Do not put the content behind a login or a form. Engines cannot see gated content. - Do not publish the content only in a PDF, an image, or a video. If the substance is not in text on the page, the engine has nothing to lift. A video needs a transcript. - Use a URL that crawlers can read. If robots.txt excludes the page, or if the page has a noindex mark, the retrieval that supplies answers excludes it.","keywords":""},{"kind":"section","title":"Publish content","heading":"Measure the result","section":"Documentation","crumbs":["Content Hub","Publish content","Measure the result"],"url":"/docs/content-hub/publishing-content#measure-the-result","text":"Publication is the start of the measurement. It is not the end of the work. 1. Record the date. 2. Continue to run the topics that the piece targets. 3. Monitor whether the page starts to show in Most cited sources. Citations usually start some weeks after publication. Engines must find the page first, and then decide that the page is useful. If the piece has no citation after one run, it did not fail. If the piece has no citation after several runs, read the sources that engines cite for that question. Then, compare them with your piece honestly.","keywords":""},{"kind":"page","title":"Geo Assistant","heading":"","section":"Documentation","crumbs":["Documentation","Geo Assistant","Geo Assistant"],"url":"/docs/geo-assistant/geo-assistant","text":"Geo is the AI assistant of Genezio. Learn how to chat with your AI visibility data, investigate changes, get reports and recommendations, and use Geo via MCP. Geo is the AI assistant of Genezio: the conversational layer of the platform that lets you chat with your AI visibility data. Geo is in the dashboard, and it has live access to all the data that Genezio collected for your brand: - Each metric - Each conversation - Each citation - Each competitor - Each perception - Each SWOT entry The name comes from GEO, Generative Engine Optimization. GEO is the equivalent of SEO for answer engines. Genezio is a GEO platform, and Geo is the assistant on top of it.","keywords":""},{"kind":"section","title":"Geo Assistant","heading":"Why Geo exists","section":"Documentation","crumbs":["Geo Assistant","Geo Assistant","Why Geo exists"],"url":"/docs/geo-assistant/geo-assistant#why-geo-exists","text":"A typical brand on Genezio has a large quantity of data: - Hundreds of topics - Thousands of conversations with answer engines each month - Citations from dozens of sources - Competitor mentions across many answer engines - Sentiment trends for each region, persona, and language To answer daily business questions, you must examine this data manually, and that is slow. Geo changes these questions into a chat. You ask in plain language. Geo runs the correct queries on your data. Then it gives the answer and the data that it used. Typical questions: - \"Am I winning or losing share of voice this month, and against whom?\" - \"Which LLM is hurting me most, and on which topics?\" - \"Why did my recommendation rate drop on ChatGPT last week?\" - \"Which citation sources are driving negative sentiment about us?\" - \"What's a competitor doing in Perplexity that we're not?\" - \"Give me three content ideas that would close the gap on topic X.\"","keywords":""},{"kind":"section","title":"Geo Assistant","heading":"What you can do with Geo","section":"Documentation","crumbs":["Geo Assistant","Geo Assistant","What you can do with Geo"],"url":"/docs/geo-assistant/geo-assistant#what-you-can-do-with-geo","text":"","keywords":""},{"kind":"section","title":"Geo Assistant","heading":"Investigate a change in a metric","section":"Documentation","crumbs":["Geo Assistant","Geo Assistant","Investigate a change in a metric"],"url":"/docs/geo-assistant/geo-assistant#investigate-a-change-in-a-metric","text":"A metric on a chart can change. For example, the recommendation rate drops, a new competitor appears, or citations from a domain become negative. Do not click through filtered views to find the cause. Send the chart, the drawer, or the object to Geo and ask \"what's behind this?\". Geo gets the applicable data, makes a summary of it, and shows the conversations that are the source of the data. For the full list of the surfaces that have this action, refer to Ask Geo.","keywords":""},{"kind":"section","title":"Geo Assistant","heading":"Write reports and briefings","section":"Documentation","crumbs":["Geo Assistant","Geo Assistant","Write reports and briefings"],"url":"/docs/geo-assistant/geo-assistant#write-reports-and-briefings","text":"Geo writes briefings when you ask for them: - \"Give me a one-paragraph executive summary of our GEO performance for the last 30 days.\" - \"List the top five competitor mentions in Introspector topics this quarter.\" - \"Summarize sentiment trends by region for the last month.\" These briefings help you with stakeholder updates, board decks, and weekly stand-ups. You do not have to collect the numbers manually.","keywords":""},{"kind":"section","title":"Geo Assistant","heading":"Get recommendations","section":"Documentation","crumbs":["Geo Assistant","Geo Assistant","Get recommendations"],"url":"/docs/geo-assistant/geo-assistant#get-recommendations","text":"Geo can recommend actions. It does not only describe the data. Geo can recommend: - The content to create - The citation sources to contact - The topics to add or to make better - How to counter the strength of a competitor Geo gets these recommendations from the real data of your brand, not from general best practice.","keywords":""},{"kind":"section","title":"Geo Assistant","heading":"Do workflows across features","section":"Documentation","crumbs":["Geo Assistant","Geo Assistant","Do workflows across features"],"url":"/docs/geo-assistant/geo-assistant#do-workflows-across-features","text":"Geo connects the dashboards, the competitive analysis, and the content production. This is a usual workflow: Investigate a competitor in the Competitors view. Send that competitor to Geo. Ask \"what SWOT perceptions show their strengths?\" Ask \"draft a brief that counters their top strength\". Send the brief to the Content Hub. In one conversation, you go from an observation to a deliverable. For the changes that Geo can make for you, read Actions Geo can take.","keywords":""},{"kind":"section","title":"Geo Assistant","heading":"How Geo works","section":"Documentation","crumbs":["Geo Assistant","Geo Assistant","How Geo works"],"url":"/docs/geo-assistant/geo-assistant#how-geo-works","text":"The data layer of the platform has a read endpoint for each chart and each drawer. Geo gets these endpoints as a set of tools. For each question, Geo selects the tools to call. It runs them with the permissions of your account. Then it makes one answer from the results in the chat. You see the answer and the tool calls that Geo made. Thus, you can audit each answer. For how Geo shows the trace of the tool calls, refer to Sessions and History. In a conversation, Geo replies in one language. After a thread starts, each answer stays in the same language. Thus, a briefing does not change its language in the middle of a thread.","keywords":""},{"kind":"section","title":"Geo Assistant","heading":"What Geo knows about your brand","section":"Documentation","crumbs":["Geo Assistant","Geo Assistant","What Geo knows about your brand"],"url":"/docs/geo-assistant/geo-assistant#what-geo-knows-about-your-brand","text":"The context of Geo also includes the AI Perception Summary of your brand. This summary is the general narrative, made by AI, of how answer engines see your brand. Thus, you can ask Geo about the perception of your brand at the highest level, not only one metric at a time: - \"How is our brand perceived overall by AI right now?\" - \"What's the headline of our AI perception, and what's dragging it down?\" For the narrative that Geo uses, refer to AI Perception Summary.","keywords":""},{"kind":"section","title":"Geo Assistant","heading":"More tools","section":"Documentation","crumbs":["Geo Assistant","Geo Assistant","More tools"],"url":"/docs/geo-assistant/geo-assistant#more-tools","text":"The set of tools of Geo becomes larger with time. In addition to the data tools, Geo can use: - Recycle Bin operations: restore deleted topics and scenarios from a conversation.","keywords":""},{"kind":"section","title":"Geo Assistant","heading":"Where to find Geo in the dashboard","section":"Documentation","crumbs":["Geo Assistant","Geo Assistant","Where to find Geo in the dashboard"],"url":"/docs/geo-assistant/geo-assistant#where-to-find-geo-in-the-dashboard","text":"You can use Geo in all of the dashboard: - In the assistant panel. - With the Ask Geo action on almost each important surface. These surfaces include the topic drawer, the scenario drawer, the conversations page, the competitors-by-LLM view, the overview, SWOT, perceptions, citations, and the competitor details drawer. - With an Ask Geo card next to the data views. The card has starter questions for frequent subjects. - With one-click Ask Geo questions on the Shopping cards. These questions use your real numbers. For example, a metric such as \"0% product visibility\" comes with a \"where do we start?\" question.","keywords":""},{"kind":"section","title":"Geo Assistant","heading":"Use Geo outside the dashboard (MCP)","section":"Documentation","crumbs":["Geo Assistant","Geo Assistant","Use Geo outside the dashboard (MCP)"],"url":"/docs/geo-assistant/geo-assistant#use-geo-outside-the-dashboard-mcp","text":"Geo is also available through MCP (Model Context Protocol) with OAuth. Thus, you can connect Geo to the MCP clients that you use, for example Claude Desktop or Claude Code. Then you can query the data of your Genezio brand from outside the dashboard. The same tools (read tools and action tools) are available, and the same permissions apply. MCP access helps when you want to use Geo in your own AI tools. Refer to Connect an MCP client for the server URL and the steps for each client.","keywords":""},{"kind":"section","title":"Geo Assistant","heading":"Related pages","section":"Documentation","crumbs":["Geo Assistant","Geo Assistant","Related pages"],"url":"/docs/geo-assistant/geo-assistant#related-pages","text":"- Ask Geo - Sessions and History - Actions Geo Can Take - Insights: SWOT Analysis - Content Hub: Briefs","keywords":""},{"kind":"page","title":"Ask Geo","heading":"","section":"Documentation","crumbs":["Documentation","Geo Assistant","Ask Geo"],"url":"/docs/geo-assistant/ask-geo","text":"Ask Geo opens the Geo Assistant with a topic, scenario, competitor, citation or perception from the dashboard attached, so you can ask about it in one click. Most data surfaces of the Genezio dashboard have an Ask Geo button with a sparkle icon. The button is in the corner of a drawer or next to a chart. You find it on a topic, a scenario, a competitor, a citation, a SWOT entry, or a perception. When you click the button, the Geo Assistant opens. The object that you look at is already attached, with its context. Then you can ask more questions about that object. You do not have to explain again what you look at.","keywords":""},{"kind":"section","title":"Ask Geo","heading":"Why Ask Geo exists","section":"Documentation","crumbs":["Geo Assistant","Ask Geo","Why Ask Geo exists"],"url":"/docs/geo-assistant/ask-geo#why-ask-geo-exists","text":"An investigation in the dashboard usually starts when you see something on a chart or in a drawer. Before Ask Geo, these were the steps to ask a question: 1. Find something interesting on a chart. 2. Open the Geo Assistant. 3. Describe again, in words, what you look at. 4. Hope that Geo finds the correct data. Ask Geo changes these steps into one click. The assistant gets the exact object that you looked at: that competitor, that scenario, or that citation source. Then you can ask immediately: - \"why is this happening?\" - \"what should I do about it?\" - \"compare this to last month\" - \"show me the conversations behind this number\" Thus, the time from an observation on a chart to a decision becomes shorter.","keywords":""},{"kind":"section","title":"Ask Geo","heading":"Where Ask Geo is available","section":"Documentation","crumbs":["Geo Assistant","Ask Geo","Where Ask Geo is available"],"url":"/docs/geo-assistant/ask-geo#where-ask-geo-is-available","text":"The action is on almost each important surface of the dashboard: Topic drawer: A topic, to investigate its performance, citations, or competitors. Scenario drawer: A scenario, to examine its performance across the answer engines. Conversations page: A conversation, for an explanation or a comparison. Competitors-by-LLM view: The behavior of a competitor on one answer engine. Competitor details drawer: The full profile of a competitor, for questions about SWOT, share of voice, or a counter-strategy. Overview: The high-level snapshot of the performance, for a summary or an explanation of a trend. SWOT: A SWOT entry (yours or of a competitor), to examine the perceptions behind it. Perceptions: An extracted perception, for an analysis of its sentiment, source, or context. Citations: A citation source, to see the conversations that cite it and what it says. Look for the Ask Geo action on each important object on the screen.","keywords":""},{"kind":"section","title":"Ask Geo","heading":"What occurs when you click Ask Geo","section":"Documentation","crumbs":["Geo Assistant","Ask Geo","What occurs when you click Ask Geo"],"url":"/docs/geo-assistant/ask-geo#what-occurs-when-you-click-ask-geo","text":"Click Ask Geo on the object. Geo opens, or moves to the front, with the object attached as context. Read what Geo received. For example: \"I'm looking at the CRM for startups topic for the last 30 days.\" Type your question. Geo uses that object for its answer. The attached context stays with the conversation. You can ask many questions about the same object, and you do not have to attach it again.","keywords":""},{"kind":"section","title":"Ask Geo","heading":"Typical Ask Geo workflows","section":"Documentation","crumbs":["Geo Assistant","Ask Geo","Typical Ask Geo workflows"],"url":"/docs/geo-assistant/ask-geo#typical-ask-geo-workflows","text":"Find the cause of a drop. On the Overview, your recommendation rate drops on ChatGPT. Click Ask Geo on the chart and ask \"what's behind this drop?\". Geo gets the conversations behind the data and explains the drop. Counter a competitor. In the competitor details drawer, click Ask Geo. Ask \"what are their top three strengths in SWOT, and how do I counter them?\". After Geo answers, you can ask it to \"draft a brief that addresses the top strength.\" Audit a citation source. In the Citations view, send a domain to Geo. Ask \"what does this source say about us versus competitors?\". Geo makes a summary of the perceptions from the conversations that cite the domain. Explain a scenario. In the scenario drawer, send a scenario. Ask \"which LLM performs worst on this and why?\". Geo compares the conversations across the answer engines.","keywords":""},{"kind":"section","title":"Ask Geo","heading":"Related pages","section":"Documentation","crumbs":["Geo Assistant","Ask Geo","Related pages"],"url":"/docs/geo-assistant/ask-geo#related-pages","text":"- Geo Assistant - Sessions and History - Actions Geo Can Take - Insights: SWOT Analysis - Core Concepts: Competitors","keywords":""},{"kind":"page","title":"Sessions and history","heading":"","section":"Documentation","crumbs":["Documentation","Geo Assistant","Sessions and history"],"url":"/docs/geo-assistant/sessions-and-history","text":"Learn how to keep many Geo Assistant sessions, find them again in the history sidebar, and audit each AI answer with the trace of its tool calls. The Geo Assistant has many chat sessions, a history sidebar, and a trace of the tool calls for each answer. With them, you can go from one conversation to a different one, go back to a conversation later, and examine how Geo made each answer. Thus, Geo is not a chat for one question only. It is a workspace that stays.","keywords":""},{"kind":"section","title":"Sessions and history","heading":"Keep many Geo sessions open","section":"Documentation","crumbs":["Geo Assistant","Sessions and history","Keep many Geo sessions open"],"url":"/docs/geo-assistant/sessions-and-history#keep-many-geo-sessions-open","text":"You can keep many Geo sessions open at the same time. Each session is independent and has its own: - Conversation history - Attached context (from Ask Geo actions) - Thread of questions Frequent examples: - A \"Competitor X investigation\" session. You go back to it when new data comes in. - A \"Q2 content planning\" session. It becomes larger over weeks, when you ask Geo for ideas for briefs. - A \"Fix our Perplexity gap\" session. It is about one answer engine only. - Short sessions for one question, for example \"summarize last week's performance\". You can go from one session to a different one at any time, and continue each thread on any day.","keywords":""},{"kind":"section","title":"Sessions and history","heading":"The history sidebar","section":"Documentation","crumbs":["Geo Assistant","Sessions and history","The history sidebar"],"url":"/docs/geo-assistant/sessions-and-history#the-history-sidebar","text":"The history sidebar shows each session of your account. In the sidebar, you can: - Open a previous session and continue the conversation. - Rename a session, to find it easily later. - Delete the sessions that you do not need. Each session keeps its full message history, its attached context, and its trace of tool calls. When you go back to a session after some weeks, you see the same view that you left.","keywords":""},{"kind":"section","title":"Sessions and history","heading":"Traces of the tool calls","section":"Documentation","crumbs":["Geo Assistant","Sessions and history","Traces of the tool calls"],"url":"/docs/geo-assistant/sessions-and-history#traces-of-the-tool-calls","text":"A tool call trace is the record of the calls that Geo made to the data of the platform for one answer. Geo uses the same read endpoints that supply the charts and the drawers of the dashboard, and it shows you each tool call. You can see: - The tool that Geo called (for example, a lookup of a metric or a fetch of a conversation). - The parameters of the call (the topic, the date range, the filter). - The result of the call. Thus, you can audit the answers of Geo. You do not have to trust a response that you cannot examine, because you can see the exact dashboard query that made the answer. For example, Geo can say that your share of voice dropped 12% on a topic. Then you can see the call and confirm the number. The trace also shows the actions that Geo takes, together with the read calls.","keywords":""},{"kind":"section","title":"Sessions and history","heading":"Why sessions and traces are important","section":"Documentation","crumbs":["Geo Assistant","Sessions and history","Why sessions and traces are important"],"url":"/docs/geo-assistant/sessions-and-history#why-sessions-and-traces-are-important","text":"Sessions and traces of tool calls change how teams use the assistant: - Investigations that stay: long threads stay across days and team members. You do not lose them when you close the tab. - Trust: you can see the trace of the tool calls, thus you can examine the work of Geo. The assistant is a colleague whose logic you can examine, not a black box. - Parallel work: you can keep a competitor investigation, a content plan, and a session for stakeholder reports open at the same time. Thus, Geo is a workspace, not a search bar.","keywords":""},{"kind":"section","title":"Sessions and history","heading":"Related pages","section":"Documentation","crumbs":["Geo Assistant","Sessions and history","Related pages"],"url":"/docs/geo-assistant/sessions-and-history#related-pages","text":"- Geo Assistant - Ask Geo - Actions Geo Can Take","keywords":""},{"kind":"page","title":"Actions of the Geo Assistant","heading":"","section":"Documentation","crumbs":["Documentation","Geo Assistant","Actions of the Geo Assistant"],"url":"/docs/geo-assistant/actions-geo-can-take","text":"Learn which changes the Geo Assistant can make in Genezio for you: manage brands, competitors, personas, topics and scenarios, and start briefs and articles. The Geo Assistant does not only read data and make summaries. It can also do actions in the Genezio platform: it can create objects, make them better, and move them for you, with the permissions of your account. Thus, Geo connects the investigation, the decision, and the deliverable.","keywords":""},{"kind":"section","title":"Actions of the Geo Assistant","heading":"Read tools and action tools","section":"Documentation","crumbs":["Geo Assistant","Actions of the Geo Assistant","Read tools and action tools"],"url":"/docs/geo-assistant/actions-geo-can-take#read-tools-and-action-tools","text":"Most answers of Geo come from read tools. These tools are the same data endpoints that supply the charts and the drawers of the dashboard. Geo selects the tools to call, runs them, and makes one answer from the results. Some tools of Geo are action tools. These tools do not only read your data. They change it. When Geo does an action, it makes a real change to your account. You see the action in the trace of the tool calls, the same as the read calls. Refer to Sessions and History.","keywords":""},{"kind":"section","title":"Actions of the Geo Assistant","heading":"What Geo can do","section":"Documentation","crumbs":["Geo Assistant","Actions of the Geo Assistant","What Geo can do"],"url":"/docs/geo-assistant/actions-geo-can-take#what-geo-can-do","text":"These are the action tools of Geo.","keywords":""},{"kind":"section","title":"Actions of the Geo Assistant","heading":"Manage brands and competitors","section":"Documentation","crumbs":["Geo Assistant","Actions of the Geo Assistant","Manage brands and competitors"],"url":"/docs/geo-assistant/actions-geo-can-take#manage-brands-and-competitors","text":"- Move a brand entry: promote, demote, merge, or attach competitor entries, from what Geo finds in the data. - Change the class of a detected brand: change a competitor that Genezio tracked by mistake into a sub-brand of yours. Geo can also do the opposite change.","keywords":""},{"kind":"section","title":"Actions of the Geo Assistant","heading":"Manage personas","section":"Documentation","crumbs":["Geo Assistant","Actions of the Geo Assistant","Manage personas"],"url":"/docs/geo-assistant/actions-geo-can-take#manage-personas","text":"- Create a persona: when Geo finds a missing audience segment, it can make a new persona with a name, a role, a country, a city, and a language. - Update a persona: change the attributes when the scenarios show that the persona does not fit.","keywords":""},{"kind":"section","title":"Actions of the Geo Assistant","heading":"Manage topics and scenarios","section":"Documentation","crumbs":["Geo Assistant","Actions of the Geo Assistant","Manage topics and scenarios"],"url":"/docs/geo-assistant/actions-geo-can-take#manage-topics-and-scenarios","text":"- Make a topic better: change its description, its persona, or its selected competitors. - Create a scenario: write a new scenario for a topic when Geo finds an intent that no scenario covers. - Update a scenario: make the text more precise, change the agent type, or change the competitors of a Comparer scenario. - Restore deleted topics or scenarios: recover an object from the Recycle Bin when a user deleted it by mistake. To compare these actions with the manual procedure, read Create topics and scenarios.","keywords":""},{"kind":"section","title":"Actions of the Geo Assistant","heading":"Make content","section":"Documentation","crumbs":["Geo Assistant","Actions of the Geo Assistant","Make content"],"url":"/docs/geo-assistant/actions-geo-can-take#make-content","text":"- Make a brief: send the findings of Geo to the Content Hub as a structured brief. The brief has the audience, the tone, the keywords, the competitors, the scenarios, and the instructions. - Start an article: start a full article in the Content Hub. The article uses the context of the current conversation. For the content that Geo makes, refer to Content Hub: Briefs.","keywords":""},{"kind":"section","title":"Actions of the Geo Assistant","heading":"How Geo starts an action","section":"Documentation","crumbs":["Geo Assistant","Actions of the Geo Assistant","How Geo starts an action"],"url":"/docs/geo-assistant/actions-geo-can-take#how-geo-starts-an-action","text":"You do not have to ask for an action by its name. Geo decides from the conversation when an action is applicable. For example: - You ask: \"This competitor X is actually a product line we own — fix that.\" Geo proposes to attach the brand to yours and does the move. - You ask: \"We're missing a persona for European mid-market buyers.\" Geo writes a new persona and creates it. - You ask: \"Draft a brief that counters their top SWOT strength.\" Geo writes the brief and sends it to the Content Hub. The action runs with the permissions of your account. Geo cannot do an action that you cannot do yourself in the UI.","keywords":""},{"kind":"section","title":"Actions of the Geo Assistant","heading":"Audit the actions of Geo","section":"Documentation","crumbs":["Geo Assistant","Actions of the Geo Assistant","Audit the actions of Geo"],"url":"/docs/geo-assistant/actions-geo-can-take#audit-the-actions-of-geo","text":"Each action of Geo shows in the trace of the tool calls of the session, with the read calls. You can see: - The action that Geo started. - The parameters of the action (the object that it changed and the new values). - The result. Thus, you can audit the actions of Geo the same as each change that a user makes in the UI.","keywords":""},{"kind":"section","title":"Actions of the Geo Assistant","heading":"Related pages","section":"Documentation","crumbs":["Geo Assistant","Actions of the Geo Assistant","Related pages"],"url":"/docs/geo-assistant/actions-geo-can-take#related-pages","text":"- Geo Assistant - Sessions and History - Ask Geo - Content Hub: Briefs - Core Concepts: Personas - Core Concepts: Competitors","keywords":""},{"kind":"page","title":"Business Scorecard","heading":"","section":"Documentation","crumbs":["Documentation","Business Scorecard","Business Scorecard"],"url":"/docs/scorecard","text":"The Business Scorecard holds the goals of your brand in your own words and measures each goal every day, so one board shows where each goal stands. Visibility numbers tell you what occurs. The Business Scorecard tells you if that is what you wanted. A scorecard holds your goals. These are the few items that are important to you in this quarter. The scorecard measures each goal every day. You do not have to read six dashboards and decide if the quarter is good. You open one board and see where each goal stands.","keywords":""},{"kind":"section","title":"Business Scorecard","heading":"Why the Business Scorecard exists","section":"Documentation","crumbs":["Business Scorecard","Business Scorecard","Why the Business Scorecard exists"],"url":"/docs/scorecard#why-the-business-scorecard-exists","text":"Two brands can have the same visibility score but be in very different positions. One brand wins the topics that cause a sale. The other brand wins topics that no buyer buys from. A general metric cannot show this difference. A goal can, because you said what is important. You write a goal in your own words. For example: - Be recommended for \"best CRM for small teams\" on ChatGPT and Perplexity. - Appear in more answers than our two closest competitors on pricing topics. - Be cited by the review sites our buyers read. The scorecard then measures exactly that, every day, and shows you the trend.","keywords":""},{"kind":"section","title":"Business Scorecard","heading":"What a goal gives you","section":"Documentation","crumbs":["Business Scorecard","Business Scorecard","What a goal gives you"],"url":"/docs/scorecard#what-a-goal-gives-you","text":"- A number that you selected. It is not an average of the platform. It is the item that you said you wanted. - A history. Each goal records where it stood yesterday. Thus, you see the movement, not only one value. - Its calculation. Open a card to see the source of the number, down to the conversations behind it. You do not have to trust a black box. - A category and the data that it reads. Each goal shows its type of goal and the parts of your brand data that it uses. Thus, each person who opens the board understands the goal without a question to you.","keywords":""},{"kind":"section","title":"Business Scorecard","heading":"Where to go next","section":"Documentation","crumbs":["Business Scorecard","Business Scorecard","Where to go next"],"url":"/docs/scorecard#where-to-go-next","text":"- Creating a Goal: write a goal in a conversation. - Reading a Goal: the card, the trend, and the calculation. - Managing Your Board: tags, order, and the archive. You enable the Business Scorecard for each brand. If you do not see it, ask your account manager to enable it.","keywords":""},{"kind":"page","title":"Creating a Goal","heading":"","section":"Documentation","crumbs":["Documentation","Business Scorecard","Creating a Goal"],"url":"/docs/scorecard/creating-a-goal","text":"Create a Business Scorecard goal when you describe it in a conversation. The agent reads your brand, decides what to measure, and shows the goal for approval. To create a goal, you describe it in a conversation. You do not configure metrics, select dimensions, or write a formula.","keywords":""},{"kind":"section","title":"Creating a Goal","heading":"Start with a sentence","section":"Documentation","crumbs":["Business Scorecard","Creating a Goal","Start with a sentence"],"url":"/docs/scorecard/creating-a-goal#start-with-a-sentence","text":"Open the Business Scorecard for your brand and start a new goal. Tell what you want in plain language: I want to be the brand ChatGPT recommends when someone asks for a project management tool for agencies. The agent reads your brand: its topics, competitors, personas, and the answer engines that you track. Then it decides what to measure. If an item is really ambiguous, the agent asks a question. The question is about the business, not about settings.","keywords":""},{"kind":"section","title":"Creating a Goal","heading":"Make the goal narrower","section":"Documentation","crumbs":["Business Scorecard","Creating a Goal","Make the goal narrower"],"url":"/docs/scorecard/creating-a-goal#make-the-goal-narrower","text":"The conversation continues until the goal is correct. You can add a limit at any time: Only for the UK market. Compare us against Asana and Monday, not the whole market. Just the topics where someone is close to buying. After each change, the agent builds the goal again and shows you what it will track. When the card agrees with what you meant, accept it.","keywords":""},{"kind":"section","title":"Creating a Goal","heading":"You do not read code","section":"Documentation","crumbs":["Business Scorecard","Creating a Goal","You do not read code"],"url":"/docs/scorecard/creating-a-goal#you-do-not-read-code","text":"A small program that the agent writes measures the goal. You do not see this program, and you do not have to. The agent tells you in words what the goal measures. Then you accept the goal or continue to change it. If a goal looks wrong, tell the agent in the same conversation. The agent then changes the goal.","keywords":""},{"kind":"section","title":"Creating a Goal","heading":"Genezio keeps your draft","section":"Documentation","crumbs":["Business Scorecard","Creating a Goal","Genezio keeps your draft"],"url":"/docs/scorecard/creating-a-goal#genezio-keeps-your-draft","text":"If you close the tab before the end, your draft is there when you come back. Thus, a reload does not delete a goal that took a long conversation.","keywords":""},{"kind":"section","title":"Creating a Goal","heading":"Change a goal later","section":"Documentation","crumbs":["Business Scorecard","Creating a Goal","Change a goal later"],"url":"/docs/scorecard/creating-a-goal#change-a-goal-later","text":"The conversation does not stop when the agent creates the goal. Open the goal and continue the conversation. You can make the market wider, replace a competitor, or change the answer engines. The goal continues from the history that it already has.","keywords":""},{"kind":"section","title":"Creating a Goal","heading":"Create goals from another tool","section":"Documentation","crumbs":["Business Scorecard","Creating a Goal","Create goals from another tool"],"url":"/docs/scorecard/creating-a-goal#create-goals-from-another-tool","text":"If your team works in an MCP client such as Claude, you can create and change goals from that client. Refer to Scorecards over MCP.","keywords":""},{"kind":"page","title":"Reading a Goal","heading":"","section":"Documentation","crumbs":["Documentation","Business Scorecard","Reading a Goal"],"url":"/docs/scorecard/reading-a-goal","text":"Read a goal card of the Business Scorecard: its current state, its daily trend, its chart shape, and the conversations that are the source of each number. Each goal is a card on your board. The card starts with where the goal stands today, not with its configuration.","keywords":""},{"kind":"section","title":"Reading a Goal","heading":"The card","section":"Documentation","crumbs":["Business Scorecard","Reading a Goal","The card"],"url":"/docs/scorecard/reading-a-goal#the-card","text":"A card shows the current state of the goal and its direction. Each card has the shape that is correct for its goal: - A trend over days - A comparison across answer engines - A share between brands - A grid Genezio selects the shape for the goal. Thus, a share-of-voice goal does not show as a line chart that hides the split. When a goal is about a competitor or an answer engine, the card shows its logo. Thus, you can read the board quickly. When a row names one of your products, the row shows the photo of the product.","keywords":""},{"kind":"section","title":"Reading a Goal","heading":"Find the source of a number","section":"Documentation","crumbs":["Business Scorecard","Reading a Goal","Find the source of a number"],"url":"/docs/scorecard/reading-a-goal#find-the-source-of-a-number","text":"Open a card to follow the number back to the conversations that are its source. This is important at two times: - When a goal moves sharply, and you must know if a real event occurred or if the measurement changed. - When you show the number to a person who will ask about its source. You do not have to trust a number on a scorecard without proof.","keywords":""},{"kind":"section","title":"Reading a Goal","heading":"Each day, automatically","section":"Documentation","crumbs":["Business Scorecard","Reading a Goal","Each day, automatically"],"url":"/docs/scorecard/reading-a-goal#each-day-automatically","text":"Genezio measures each goal one time each day and keeps the value of yesterday. Thus, you get a history and you do not have to run anything. A goal that you create today starts its trend immediately.","keywords":""},{"kind":"section","title":"Reading a Goal","heading":"A goal for many languages or markets","section":"Documentation","crumbs":["Business Scorecard","Reading a Goal","A goal for many languages or markets"],"url":"/docs/scorecard/reading-a-goal#a-goal-for-many-languages-or-markets","text":"You can set a goal to one language. Thus, a brand that works in many markets can keep one goal for each market, not one unclear average.","keywords":""},{"kind":"page","title":"Managing Your Board","heading":"","section":"Documentation","crumbs":["Documentation","Business Scorecard","Managing Your Board"],"url":"/docs/scorecard/managing-your-board","text":"Keep the Business Scorecard board clear: put the goal cards in order, add tags, rename goals, move old goals to the archive, and read the audit log. People read a board much more frequently than they edit it. Thus, keep on the board only the goals that are important now.","keywords":""},{"kind":"section","title":"Managing Your Board","heading":"Order","section":"Documentation","crumbs":["Business Scorecard","Managing Your Board","Order"],"url":"/docs/scorecard/managing-your-board#order","text":"Drag the cards into the order in which you want to read them. Genezio keeps your order. Thus, the goal that you examine first each morning stays at the top.","keywords":""},{"kind":"section","title":"Managing Your Board","heading":"Tags","section":"Documentation","crumbs":["Business Scorecard","Managing Your Board","Tags"],"url":"/docs/scorecard/managing-your-board#tags","text":"Add a tag to a goal to put it in a group. For example, you can group goals by market, by campaign, by the team that owns them, or by quarter. You can change or remove tags at any time. When a board has more than a small number of goals, tags are the fastest way to keep it easy to read.","keywords":""},{"kind":"section","title":"Managing Your Board","heading":"Rename a goal","section":"Documentation","crumbs":["Business Scorecard","Managing Your Board","Rename a goal"],"url":"/docs/scorecard/managing-your-board#rename-a-goal","text":"You can change the title of a goal on the page of the goal. You can rename a goal at any time, because the measurement does not use the title.","keywords":""},{"kind":"section","title":"Managing Your Board","heading":"Archive a goal","section":"Documentation","crumbs":["Business Scorecard","Managing Your Board","Archive a goal"],"url":"/docs/scorecard/managing-your-board#archive-a-goal","text":"You can remove a goal that you do not track any more from the board. Genezio does not delete the goal. It moves the goal to the archive, with its full history. Later, you can restore the goal to the board. This is important for seasonal work. When a campaign ends, you can archive its goal. Next year, you can restore the goal, and the trend of the last year is still attached.","keywords":""},{"kind":"section","title":"Managing Your Board","heading":"Who changed what","section":"Documentation","crumbs":["Business Scorecard","Managing Your Board","Who changed what"],"url":"/docs/scorecard/managing-your-board#who-changed-what","text":"The audit log records each action on a scorecard: goals that a person created, changed, archived, or restored. If a goal moves and no person remembers a change, the log gives the answer.","keywords":""},{"kind":"page","title":"Dashboards and metrics","heading":"","section":"Documentation","crumbs":["Documentation","Dashboards and metrics","Dashboards and metrics"],"url":"/docs/dashboards","text":"A guide to each Genezio dashboard card: what it measures, how to divide it by topic, scenario or competitor, and how to read trends before you act. The pages in this section tell you how to read each card of the dashboard. They also tell you how to divide each card before you act on it. Start with the two headline numbers: - Visibility score explained - Share of Voice Then read the breakdowns. Most of your work comes from them: - Topic performance - Scenario performance - Competitor comparison Citation frequency shows the sources that have an effect on your category. Trend tracking tells about a problem that surprises most teams. Answer engines give different results from run to run. Thus, the direction of a metric is the signal, and a single reading is not.","keywords":""},{"kind":"page","title":"Visibility score explained","heading":"","section":"Documentation","crumbs":["Documentation","Dashboards and metrics","Visibility score explained"],"url":"/docs/dashboards/visibility-score-explained","text":"Read the visibility card of your dashboard: the formula, the change against the previous period, normal noise between runs, and the breakdowns to check first. The visibility number on your dashboard is the percentage of relevant conversations in which your brand appeared. Formula: Conversations where your brand appears ÷ Total eligible conversations × 100.","keywords":""},{"kind":"section","title":"Visibility score explained","heading":"How to read the card","section":"Documentation","crumbs":["Dashboards and metrics","Visibility score explained","How to read the card"],"url":"/docs/dashboards/visibility-score-explained#how-to-read-the-card","text":"The number is for the current period. The change next to it is the difference from the previous period. The change is usually the more useful part, because a score has meaning only when you see how it moves. The card shows an average across each topic and each answer engine that you monitor. Use this number on a board slide. Do not use it to decide your next actions, because it hides the location of the problem.","keywords":""},{"kind":"section","title":"Visibility score explained","heading":"When the number changes and nothing changed","section":"Documentation","crumbs":["Dashboards and metrics","Visibility score explained","When the number changes and nothing changed"],"url":"/docs/dashboards/visibility-score-explained#when-the-number-changes-and-nothing-changed","text":"Answer engines are not deterministic. When you ask the same question two times, the wording is different. Sometimes the brands are also different. A change of some points between runs is the normal noise of the medium. Use a single reading as an estimate. A direction that continues over several runs is a finding.","keywords":""},{"kind":"section","title":"Visibility score explained","heading":"Before you act on the score","section":"Documentation","crumbs":["Dashboards and metrics","Visibility score explained","Before you act on the score"],"url":"/docs/dashboards/visibility-score-explained#before-you-act-on-the-score","text":"Divide the score before you spend money or time: - By topic. You are rarely weak on all topics. For more information, see Topic performance. - By answer engine. Possibly a brand is strong on one answer engine and absent on a different one. Then the brand has a source problem that is specific to that answer engine. - Against competitors. A score of 40% has no meaning until you know the score of the leader. The leader can be at 45% or at 90%.","keywords":""},{"kind":"page","title":"Share of Voice","heading":"","section":"Documentation","crumbs":["Documentation","Dashboards and metrics","Share of Voice"],"url":"/docs/dashboards/share-of-voice","text":"Read the Share of Voice card: how it differs from visibility, why it is zero-sum, how to watch the category leader, and how to trace a sudden drop in share. Share of Voice is the part of the conversation in your category that is about your brand and not about other brands. Formula: Mentions of your brand ÷ Mentions of all brands (your brand and the competitors) × 100.","keywords":""},{"kind":"section","title":"Share of Voice","heading":"How Share of Voice is different from visibility","section":"Documentation","crumbs":["Dashboards and metrics","Share of Voice","How Share of Voice is different from visibility"],"url":"/docs/dashboards/share-of-voice#how-share-of-voice-is-different-from-visibility","text":"Visibility tells you whether you appear. Share of Voice tells you how much of the space you get when you appear. The two numbers change independently, and the difference between them gives information. Possibly you appear in most answers but you have a small share. Then the answers mention you only briefly, and they discuss another brand in detail.","keywords":""},{"kind":"section","title":"Share of Voice","heading":"How to read Share of Voice","section":"Documentation","crumbs":["Dashboards and metrics","Share of Voice","How to read Share of Voice"],"url":"/docs/dashboards/share-of-voice#how-to-read-share-of-voice","text":"Share of Voice is zero-sum. Your share can decrease while your visibility increases, if a competitor grew faster. This is not a failure of your work. The category becomes more crowded. Say this clearly when you report the number. Monitor the share of the leader, not only your share. A category where the leader has 60% is different from a category where each of the top three brands has 20%. In the first category, you attack a position. In the second category, the field is still open. A new brand takes share from a different brand. When your share decreases, find out if a new brand appeared in the answers.","keywords":""},{"kind":"section","title":"Share of Voice","heading":"Where the data comes from","section":"Documentation","crumbs":["Dashboards and metrics","Share of Voice","Where the data comes from"],"url":"/docs/dashboards/share-of-voice#where-the-data-comes-from","text":"Genezio counts the mentions in the same conversations that it uses for each other number. Thus, Share of Voice always agrees with what you can read in Conversations. When a number surprises you, open the answers that it comes from. For the full definition of the metric and its views in the dashboard, see Share of Voice.","keywords":""},{"kind":"page","title":"Topic performance","heading":"","section":"Documentation","crumbs":["Documentation","Dashboards and metrics","Topic performance"],"url":"/docs/dashboards/topic-performance","text":"Topic performance divides your AI visibility by topic. Find the topics with zero visibility, the topics with no recommendation and the topics that change. One visibility number for a brand hides the only data that you can act on. You are strong on some topics and absent from other topics. Topic performance shows this breakdown.","keywords":""},{"kind":"section","title":"Topic performance","heading":"What to look for","section":"Documentation","crumbs":["Dashboards and metrics","Topic performance","What to look for"],"url":"/docs/dashboards/topic-performance#what-to-look-for","text":"Topics where your score is zero. These rows give the most actions. A zero means that answer engines answer that question without your brand. The conversations show which brands they use in place of yours. Topics where you are visible but not recommended. The answers mention you, but they select a different brand. This is a positioning problem, and the answers usually tell you the reason. Topics that change. A topic that decreases over several runs needs more attention than a topic that was always weak. Something changed recently, and it is easiest to find the cause of a recent change.","keywords":""},{"kind":"section","title":"Topic performance","heading":"How to sort by importance","section":"Documentation","crumbs":["Dashboards and metrics","Topic performance","How to sort by importance"],"url":"/docs/dashboards/topic-performance#how-to-sort-by-importance","text":"A topic with high demand and low visibility has more value than a topic with low demand that you lead. Sort by the gap, not by the score. Your worst topic is not automatically your best opportunity. For some questions, you must not win, because the honest answer is a competitor.","keywords":""},{"kind":"section","title":"Topic performance","heading":"How to examine a topic in more detail","section":"Documentation","crumbs":["Dashboards and metrics","Topic performance","How to examine a topic in more detail"],"url":"/docs/dashboards/topic-performance#how-to-examine-a-topic-in-more-detail","text":"Open a topic to see its scenarios and the conversations of each scenario. The conversation gives the explanation. The score tells you where to look, and the answer tells you why. For more information, see Topics and Scenario performance.","keywords":""},{"kind":"page","title":"Scenario performance","heading":"","section":"Documentation","crumbs":["Documentation","Dashboards and metrics","Scenario performance"],"url":"/docs/dashboards/scenario-performance","text":"Scenario performance shows which phrasings of a topic your brand wins in AI answers and which it loses, so you can find weak framings and write better briefs. A topic is the subject. A scenario is one realistic question that a person asks about that subject. Scenario performance shows the phrasings that you win and the phrasings that you lose.","keywords":""},{"kind":"section","title":"Scenario performance","heading":"Why the difference is important","section":"Documentation","crumbs":["Dashboards and metrics","Scenario performance","Why the difference is important"],"url":"/docs/dashboards/scenario-performance#why-the-difference-is-important","text":"Two scenarios in one topic can give opposite results. For example, \"Best CRM for a small team\" and \"CRM for a team of five that outgrew spreadsheets\" are in the same topic. They frequently get different answers, because the answer engines use different sources for them. A topic score is an average, and it hides this difference. The scenario view shows it.","keywords":""},{"kind":"section","title":"Scenario performance","heading":"What scenario performance tells you","section":"Documentation","crumbs":["Dashboards and metrics","Scenario performance","What scenario performance tells you"],"url":"/docs/dashboards/scenario-performance#what-scenario-performance-tells-you","text":"The framing that you lose. Possibly you win the general question and lose each specific question. Then your content is positioning material with no concrete detail in it. If the opposite occurs, you have the detail but no overview. The scenarios where a competitor is strong. When a competitor wins one scenario each time, one source usually supports it. Open the conversation and examine the cited sources. Whether a scenario is realistic. A scenario that no person types gives answers that no person gets. Scenario performance is where you find a badly written scenario. Frequently, the fix is to write the scenario again.","keywords":""},{"kind":"section","title":"Scenario performance","heading":"How to act on scenario performance","section":"Documentation","crumbs":["Dashboards and metrics","Scenario performance","How to act on scenario performance"],"url":"/docs/dashboards/scenario-performance#how-to-act-on-scenario-performance","text":"Write content briefs at the level of the scenario. A brief for a scenario has a question to answer. A brief for a topic has only a subject to discuss, and that is much weaker. For more information, see Creating scenarios.","keywords":""},{"kind":"page","title":"Citation frequency","heading":"","section":"Documentation","crumbs":["Documentation","Dashboards and metrics","Citation frequency"],"url":"/docs/dashboards/citation-frequency","text":"Citation frequency shows which sources answer engines use in your category and how frequently. Learn how to read the rank, your own domain and the gaps. Citation frequency shows which sources answer engines use in your category, and how frequently they use them.","keywords":""},{"kind":"section","title":"Citation frequency","heading":"What the number means","section":"Documentation","crumbs":["Dashboards and metrics","Citation frequency","What the number means"],"url":"/docs/dashboards/citation-frequency#what-the-number-means","text":"A citation count is the number of times that an answer used a source as support in your conversations. A source with forty citations has an effect on your category. A source with two citations is noise. The rank is more important than the absolute number. The absolute number changes with the quantity of conversations that you ran. The order does not change with it.","keywords":""},{"kind":"section","title":"Citation frequency","heading":"How to read the list","section":"Documentation","crumbs":["Dashboards and metrics","Citation frequency","How to read the list"],"url":"/docs/dashboards/citation-frequency#how-to-read-the-list","text":"The top of the list is your real competition for attention. These are sources, not brands. When a comparison site is at the top, that site decides a large part of your position in the category. The position of your own domain gives a diagnosis. A high position is good, but it is fragile. Answer engines give more weight to independent sources, because a description of a brand by the brand itself is weak evidence. A profile of only your own pages is easy to replace. One competitor that gets coverage from third parties can push your pages down. A missing source is a finding. When you expect a publication and you do not see it, an answer engine does not trust it for this category. Or the answer engine cannot parse it.","keywords":""},{"kind":"section","title":"Citation frequency","heading":"From a domain to an action","section":"Documentation","crumbs":["Dashboards and metrics","Citation frequency","From a domain to an action"],"url":"/docs/dashboards/citation-frequency#from-a-domain-to-an-action","text":"A domain is frequently too general. When Genezio can identify the publisher of the cited material, it shows the channel or the author. For example, it shows the specific YouTube channel, the subreddit, or the named writer. For more information, see Channels and authors. This detail changes the action. \"Do something about YouTube\" is not a plan. \"This channel cited us eleven times, so we send it an email\" is a plan.","keywords":""},{"kind":"page","title":"Competitor comparison","heading":"","section":"Documentation","crumbs":["Documentation","Dashboards and metrics","Competitor comparison"],"url":"/docs/dashboards/competitor-comparison","text":"Compare each competitor with your brand on the same topics, scenarios and answer engines, and learn to read visibility and recommendation side by side. Competitor comparison shows each metric of your brand for the brands that you compete with. Genezio measures each metric in the same way for each brand.","keywords":""},{"kind":"section","title":"Competitor comparison","heading":"How to read the comparison","section":"Documentation","crumbs":["Dashboards and metrics","Competitor comparison","How to read the comparison"],"url":"/docs/dashboards/competitor-comparison#how-to-read-the-comparison","text":"Put the three numbers side by side. Then the profile of each competitor becomes clear: - High visibility, high recommendation. This is the brand to beat. Answer engines know this brand and they also recommend it. - High visibility, low recommendation. Answer engines know this brand, but they do not prefer it. Frequently, this is the incumbent that all answers name and no answer selects. - Low visibility, high recommendation. This brand is dangerous. Answer engines show it rarely, but the answers recommend it when they show it. If its sources become better, it becomes a real threat. - Low visibility, low recommendation. This brand does not compete in AI answers at this time, independent of what it does in other channels. Your own position in this grid tells you which problem you have.","keywords":""},{"kind":"section","title":"Competitor comparison","heading":"How Genezio makes the comparison fair","section":"Documentation","crumbs":["Dashboards and metrics","Competitor comparison","How Genezio makes the comparison fair"],"url":"/docs/dashboards/competitor-comparison#how-genezio-makes-the-comparison-fair","text":"The comparison is valid because Genezio runs each brand through the same topics, the same scenarios, and the same answer engines. The number of a competitor on this page is not its global position. It is its position on your set of questions. This is intentional. The comparison answers the question that is important: \"Which brand wins the questions that my buyers ask?\" It does not answer \"Which brand is bigger?\"","keywords":""},{"kind":"section","title":"Competitor comparison","heading":"When an unknown competitor appears","section":"Documentation","crumbs":["Dashboards and metrics","Competitor comparison","When an unknown competitor appears"],"url":"/docs/dashboards/competitor-comparison#when-an-unknown-competitor-appears","text":"Examine the brand seriously before you ignore it. Answer engines show brands according to how well sources support them, not according to market share. You possibly never lost a deal to this brand, but an answer engine can still recommend it. Such a problem starts in AI answers and comes into your sales pipeline later. For more information, see Competitor insights.","keywords":""},{"kind":"page","title":"Trend tracking","heading":"","section":"Documentation","crumbs":["Documentation","Dashboards and metrics","Trend tracking"],"url":"/docs/dashboards/trend-tracking","text":"Answer engines give different results from run to run. Learn which trends to trust, how to find the cause of a change, and how to report it to leadership. A single reading can give you an incorrect conclusion. Answer engines give different results from run to run. Thus, the value of repeated measurement is the direction, not the single point.","keywords":""},{"kind":"section","title":"Trend tracking","heading":"What to trust","section":"Documentation","crumbs":["Dashboards and metrics","Trend tracking","What to trust"],"url":"/docs/dashboards/trend-tracking#what-to-trust","text":"A direction that continues over several runs is a finding. A single sudden change is usually noise. The exception is when a specific event occurred. For example, you published content, a competitor launched a product, or an answer engine got an update. Before you examine a change, find out if your competitors changed in the same direction at the same time. If all brands changed, the answer engine changed, not your brand.","keywords":""},{"kind":"section","title":"Trend tracking","heading":"How to find the cause of a change","section":"Documentation","crumbs":["Dashboards and metrics","Trend tracking","How to find the cause of a change"],"url":"/docs/dashboards/trend-tracking#how-to-find-the-cause-of-a-change","text":"A trend line tells you when a change occurred. It does not tell you why. To find the cause, do these steps: 1. Find the date when the change started. 2. Find out if the change applies to all topics or only to some topics. 3. Open the conversations from before and after the change. 4. Compare the cited sources. A changed answer almost always means a changed source. Step 4 explains most changes that you cannot explain at first.","keywords":""},{"kind":"section","title":"Trend tracking","heading":"How to report a change","section":"Documentation","crumbs":["Dashboards and metrics","Trend tracking","How to report a change"],"url":"/docs/dashboards/trend-tracking#how-to-report-a-change","text":"Report the direction and the cause, not the number. \"Visibility is 62%\" causes a question that no person can answer. \"Visibility increased by nine points on pricing topics after we published the comparison page, and the answer engines now cite it\" is a result. Sometimes the number must have a specific meaning for your business, not for the platform. For this, use the Business Scorecard.","keywords":""},{"kind":"page","title":"Improving AI visibility","heading":"","section":"Documentation","crumbs":["Documentation","Improving AI visibility","Improving AI visibility"],"url":"/docs/improving-ai-visibility","text":"A strategy guide to AI visibility in the order of effect: source selection, quotable content, updates, authority, entities, topic coverage, and monitoring. This section gives a strategy to improve your AI visibility. The pages are approximately in the order of their effect. How LLMs choose sources explains the mechanism: retrieval, then trust, then quotation. Each stage fails in a different way. Usually, the work with the lowest cost is one of these two tasks: - Structure the content that you already have, so that an answer engine can quote a passage. - Update the pages that answer engines already retrieve. The work with a long-term effect is to build authority through independent sources. Backlinks and citations explains why this work is not the same as link building. The foundations are these two tasks: - Entity reinforcement, so that answer engines know what you are. - Semantic topic coverage, so that your content covers a topic in substance, not only in name. Then, monitor your AI visibility. Do not read noise as a result.","keywords":""},{"kind":"page","title":"How LLMs choose sources","heading":"","section":"Documentation","crumbs":["Documentation","Improving AI visibility","How LLMs choose sources"],"url":"/docs/improving-ai-visibility/how-llms-choose-sources","text":"Answer engines retrieve, trust, and then quote sources. See why a page fails at each stage and which four actions improve the place of your brand in AI answers. When an answer engine answers a question about your category, it does retrieval before generation. It finds material, it judges the material, and it writes the answer from the material that it trusts. This order explains most of the actions that you can take.","keywords":""},{"kind":"section","title":"How LLMs choose sources","heading":"What answer engines retrieve","section":"Documentation","crumbs":["Improving AI visibility","How LLMs choose sources","What answer engines retrieve"],"url":"/docs/improving-ai-visibility/how-llms-choose-sources#what-answer-engines-retrieve","text":"Retrieval matches substance, not brand. An answer engine finds a page because the page contains the words of the question and answers the question directly. For more, read Query fanouts. Thus, a smaller brand with one useful page can rank higher than a large brand with no such page.","keywords":""},{"kind":"section","title":"How LLMs choose sources","heading":"What answer engines trust","section":"Documentation","crumbs":["Improving AI visibility","How LLMs choose sources","What answer engines trust"],"url":"/docs/improving-ai-visibility/how-llms-choose-sources#what-answer-engines-trust","text":"From the retrieved material, answer engines prefer sources that look like evidence: - Independent sources are better than self-published sources. When a brand describes itself, the evidence is weak. When a third party describes the brand, the evidence is stronger. This is the largest factor, and most teams do not use it sufficiently. - Specific pages are better than general pages. A page that fully answers one question is better than a page that touches twenty questions. - Consistent sources are better than contradictory sources. When sources do not agree about you, answer engines give an uncertain answer. Or, they select the claim with more support.","keywords":""},{"kind":"section","title":"How LLMs choose sources","heading":"What answer engines quote","section":"Documentation","crumbs":["Improving AI visibility","How LLMs choose sources","What answer engines quote"],"url":"/docs/improving-ai-visibility/how-llms-choose-sources#what-answer-engines-quote","text":"Answer engines quote passages. An answer engine can retrieve and trust a page, but the page can still give nothing to the answer. The cause is that no paragraph in the page is a full answer by itself. This failure is the most frequent failure, and it is the easiest to correct. Read Structuring content for LLMs.","keywords":""},{"kind":"section","title":"How LLMs choose sources","heading":"What this means for you","section":"Documentation","crumbs":["Improving AI visibility","How LLMs choose sources","What this means for you"],"url":"/docs/improving-ai-visibility/how-llms-choose-sources#what-this-means-for-you","text":"Do these actions in the order of their effect: 1. Make your content quotable for the questions where you already rank. 2. Get independent mentions in the places where buyers already look. 3. Cover the questions where you are absent. Select them by their commercial value. 4. Keep your facts consistent in all the places where they occur. You cannot submit your content to an answer engine. There is no ranking factor to manipulate, and there is no person to appeal to. The work is to be the best available source for a question.","keywords":""},{"kind":"page","title":"Entity reinforcement","heading":"","section":"Documentation","crumbs":["Documentation","Improving AI visibility","Entity reinforcement"],"url":"/docs/improving-ai-visibility/entity-reinforcement","text":"Answer engines must know what your brand is before they can recommend it. Learn how to make your brand and category unambiguous with clear words and schema. An answer engine must know what you are before it can decide if you answer a question. Entity reinforcement makes this information unambiguous.","keywords":""},{"kind":"section","title":"Entity reinforcement","heading":"The problem it solves","section":"Documentation","crumbs":["Improving AI visibility","Entity reinforcement","The problem it solves"],"url":"/docs/improving-ai-visibility/entity-reinforcement#the-problem-it-solves","text":"For a machine, a brand is frequently ambiguous: - The name is also a usual word, or the name of a different company. - Different places describe the category with different words, for example \"project management\", \"work OS\", or \"collaboration platform\". - The product changed, and older sources describe the old product. When an answer engine identifies you inconsistently, it sometimes answers about a different item. Sometimes, it does not include you in answers where you belong.","keywords":""},{"kind":"section","title":"Entity reinforcement","heading":"What reinforcement looks like","section":"Documentation","crumbs":["Improving AI visibility","Entity reinforcement","What reinforcement looks like"],"url":"/docs/improving-ai-visibility/entity-reinforcement#what-reinforcement-looks-like","text":"- Say what you are clearly, in the words that buyers use. Positioning text that does not use the category name makes you harder to identify. \"The operating system for modern teams\" gives no information to a machine. - Say it consistently in all places. This includes your site, your profiles, and your documentation. Contradictions make answer engines give uncertain answers. - Use structured markup. Schema tells answer engines what a page is, thus they do not have to infer it. Read Schema markup. - Remove ambiguity on purpose. If your name is the same as a different name, put the category next to your name. Do this in text that answer engines will probably retrieve.","keywords":""},{"kind":"section","title":"Entity reinforcement","heading":"How to know if it works","section":"Documentation","crumbs":["Improving AI visibility","Entity reinforcement","How to know if it works"],"url":"/docs/improving-ai-visibility/entity-reinforcement#how-to-know-if-it-works","text":"Read your conversations. Answer engines can describe your category inconsistently, or they can confuse you with a different brand. This is an entity problem. Content about specific questions cannot correct it. This work is a foundation. Do it before you work on individual topics, because all other work assumes that the answer engine knows which company you are.","keywords":""},{"kind":"page","title":"Structuring content for LLMs","heading":"","section":"Documentation","crumbs":["Documentation","Improving AI visibility","Structuring content for LLMs"],"url":"/docs/improving-ai-visibility/structuring-content-for-llms","text":"Answer engines quote passages, not pages. Use a one-paragraph test and four rules to make each passage self-contained, specific, and quotable in AI answers. Answer engines quote passages. Frequently, an answer engine never cites a good page. The most frequent cause is that no paragraph in the page answers a question by itself.","keywords":""},{"kind":"section","title":"Structuring content for LLMs","heading":"The test","section":"Documentation","crumbs":["Improving AI visibility","Structuring content for LLMs","The test"],"url":"/docs/improving-ai-visibility/structuring-content-for-llms#the-test","text":"1. Copy a paragraph from the page. 2. Read the paragraph alone. 3. Find if it answers a question without the paragraph before it. If it does, the paragraph is quotable. If it needs an introduction, a pronoun reference, or a heading to have meaning, an answer engine cannot use it. Most marketing pages fail this test in all their paragraphs.","keywords":""},{"kind":"section","title":"Structuring content for LLMs","heading":"What makes a passage quotable","section":"Documentation","crumbs":["Improving AI visibility","Structuring content for LLMs","What makes a passage quotable"],"url":"/docs/improving-ai-visibility/structuring-content-for-llms#what-makes-a-passage-quotable","text":"- It answers in the first sentence. Put the conclusion first and the reasoning after it. If you build slowly to a point, you lose the quote. - It is self-contained. Write the name of the subject, not \"it\" or \"this\". When an answer engine puts a passage into an answer, the passage loses its context. - It is specific. Give numbers, conditions, and limits. \"Pricing starts at $X per user for teams under ten\" is quotable. \"Flexible pricing for teams of all sizes\" is not quotable. - It uses the words of the question. Read Query fanouts.","keywords":""},{"kind":"section","title":"Structuring content for LLMs","heading":"Structure that helps","section":"Documentation","crumbs":["Improving AI visibility","Structuring content for LLMs","Structure that helps"],"url":"/docs/improving-ai-visibility/structuring-content-for-llms#structure-that-helps","text":"- Headings that are questions. - Short paragraphs, with one claim in each paragraph. - Tables for comparisons, because a table row is a self-contained fact. - Explicit statements of what an item is not. Answer engines frequently quote these statements, because they resolve a difference that the user considered.","keywords":""},{"kind":"section","title":"Structuring content for LLMs","heading":"Structure that causes problems","section":"Documentation","crumbs":["Improving AI visibility","Structuring content for LLMs","Structure that causes problems"],"url":"/docs/improving-ai-visibility/structuring-content-for-llms#structure-that-causes-problems","text":"- Long blocks of text, with the answer hidden inside. - Information that is only in an image, a PDF, or a video without a transcript. - Key facts in three different sections, thus no section is complete. - A conclusion that only refers vaguely to the previous text.","keywords":""},{"kind":"section","title":"Structuring content for LLMs","heading":"Where to start","section":"Documentation","crumbs":["Improving AI visibility","Structuring content for LLMs","Where to start"],"url":"/docs/improving-ai-visibility/structuring-content-for-llms#where-to-start","text":"Do not write the full site again. Select the pages that you already have for the questions where you already appear. Then, make their key paragraphs self-contained. This is the visibility work with the lowest cost, because retrieval already works. Only the quotation fails.","keywords":""},{"kind":"page","title":"Semantic topic coverage","heading":"","section":"Documentation","crumbs":["Documentation","Improving AI visibility","Semantic topic coverage"],"url":"/docs/improving-ai-visibility/semantic-topic-coverage","text":"To cover a topic, answer the full cluster of buyer questions. Use query fanouts to find the cluster, write deep pieces, and check coverage across scenarios. To cover a topic, answer the cluster of questions that buyers ask about it. One page with the topic in the title is not coverage.","keywords":""},{"kind":"section","title":"Semantic topic coverage","heading":"Why one page is rarely sufficient","section":"Documentation","crumbs":["Improving AI visibility","Semantic topic coverage","Why one page is rarely sufficient"],"url":"/docs/improving-ai-visibility/semantic-topic-coverage#why-one-page-is-rarely-sufficient","text":"An answer engine expands a buyer question into many retrieval queries. Usually, one page answers one or two of these queries well, and the other queries only briefly. Thus, the answer engine retrieves the page for only a part of the cluster. Coverage means that your site has an answer for each question in the cluster. Each answer must be real, not only a mention.","keywords":""},{"kind":"section","title":"Semantic topic coverage","heading":"Find the cluster","section":"Documentation","crumbs":["Improving AI visibility","Semantic topic coverage","Find the cluster"],"url":"/docs/improving-ai-visibility/semantic-topic-coverage#find-the-cluster","text":"Use the query fanouts for the topic. Query fanouts are the expansion of the question by the answer engine itself. Thus, they give a better map of the cluster than an internal guess. Group the query fanouts by intent. For example, put the query fanouts about price together, and put the query fanouts about integrations together. Each group is a candidate for one piece of content.","keywords":""},{"kind":"section","title":"Semantic topic coverage","heading":"Depth before breadth","section":"Documentation","crumbs":["Improving AI visibility","Semantic topic coverage","Depth before breadth"],"url":"/docs/improving-ai-visibility/semantic-topic-coverage#depth-before-breadth","text":"A frequent failure is one page for each query fanout. Thin pages give answer engines nothing to quote. They also compete with each other, and they signal low quality. One substantial piece that answers a coherent group is better than five pages that each answer only half of one query fanout.","keywords":""},{"kind":"section","title":"Semantic topic coverage","heading":"Know when a topic is covered","section":"Documentation","crumbs":["Improving AI visibility","Semantic topic coverage","Know when a topic is covered"],"url":"/docs/improving-ai-visibility/semantic-topic-coverage#know-when-a-topic-is-covered","text":"A topic is not covered when you publish. It is covered when the conversations show you in many scenarios of the topic, not only in one scenario. Read Scenario performance. A topic can look covered but show uneven results across its scenarios. This is the clearest sign that the coverage is only in name, not real.","keywords":""},{"kind":"page","title":"Building authority","heading":"","section":"Documentation","crumbs":["Documentation","Improving AI visibility","Building authority"],"url":"/docs/improving-ai-visibility/building-authority","text":"Answer engines prefer independent sources to your own site. Learn where AI authority comes from, how to select where to spend, and which tactics have no effect. Answer engines prefer independent sources, because a description of a brand by the brand itself is not evidence. In this context, authority means that other parties describe you, in places that answer engines retrieve.","keywords":""},{"kind":"section","title":"Building authority","heading":"Why your own site has a limit","section":"Documentation","crumbs":["Improving AI visibility","Building authority","Why your own site has a limit"],"url":"/docs/improving-ai-visibility/building-authority#why-your-own-site-has-a-limit","text":"You can write a perfect page and still lose to an average comparison page from a third party. The comparison is evidence, and your page is a claim. A citation profile that contains only your own domain is also fragile. When one competitor gets independent coverage, that competitor can replace you. Read Most cited sources.","keywords":""},{"kind":"section","title":"Building authority","heading":"Where authority comes from","section":"Documentation","crumbs":["Improving AI visibility","Building authority","Where authority comes from"],"url":"/docs/improving-ai-visibility/building-authority#where-authority-comes-from","text":"- Review and comparison sites, where buyers in your category already look. Usually, these sites have the largest effect, because answer engines cite them frequently. Also, you can partly control your profile on them. - Community discussion: the subreddit or forum where people discuss your category. Answer engines cite these places. You cannot buy a place in them. You can only earn it. - Independent writers and creators that write about your field. When Genezio can identify them, it gives the name of the specific channel or author. Read Channels and authors. Thus, your outreach starts from a list, not from a guess. - Documentation and technical text that other parties refer to. This work is slow but durable. For some categories, it is the only route.","keywords":""},{"kind":"section","title":"Building authority","heading":"How to select where to spend","section":"Documentation","crumbs":["Improving AI visibility","Building authority","How to select where to spend"],"url":"/docs/improving-ai-visibility/building-authority#how-to-select-where-to-spend","text":"Start from the sources that answer engines already cite in your category. Do not start from a general list of good publications. A site with forty citations for your questions has more value than a famous site with zero citations.","keywords":""},{"kind":"section","title":"Building authority","heading":"What does not work","section":"Documentation","crumbs":["Improving AI visibility","Building authority","What does not work"],"url":"/docs/improving-ai-visibility/building-authority#what-does-not-work","text":"- Paid placements that answer engines do not retrieve. - Press releases that no publication uses. - Link building for search rankings. Answer engines read the text. They do not count the links.","keywords":""},{"kind":"page","title":"Backlinks and citations","heading":"","section":"Documentation","crumbs":["Documentation","Improving AI visibility","Backlinks and citations"],"url":"/docs/improving-ai-visibility/backlinks-and-citations","text":"A backlink is an HTML link. A citation is an answer engine that uses a source to support a claim. Learn why mentions without links count for AI visibility. Backlinks and citations are not the same item. If you confuse them, you waste money.","keywords":""},{"kind":"section","title":"Backlinks and citations","heading":"The difference","section":"Documentation","crumbs":["Improving AI visibility","Backlinks and citations","The difference"],"url":"/docs/improving-ai-visibility/backlinks-and-citations#the-difference","text":"A backlink is an HTML link to your page. Backlinks are important for search engines, which use links as votes. A citation occurs when an answer engine uses a source to support a claim. A citation does not need a link. A page that describes your product accurately can be a citation without a link to you. Also, an answer engine can ignore a page that has a link to you. Answer engines read the text. For them, a link is not a vote. It is only markup.","keywords":""},{"kind":"section","title":"Backlinks and citations","heading":"What this changes","section":"Documentation","crumbs":["Improving AI visibility","Backlinks and citations","What this changes"],"url":"/docs/improving-ai-visibility/backlinks-and-citations#what-this-changes","text":"A mention without a link still has an effect. When a cited source describes you accurately, the mention has an effect, with or without a link. Outreach that demands a link sometimes loses the mention. A link from a page that answer engines never retrieve has no effect. Directory links and low-quality placements for search rankings are not visible to answer engines. An incorrect description is worse than no description. A cited source can give incorrect data about your pricing or your category. Answer engines then copy that error into their answers. The correction of that error is real visibility work.","keywords":""},{"kind":"section","title":"Backlinks and citations","heading":"Where the two overlap","section":"Documentation","crumbs":["Improving AI visibility","Backlinks and citations","Where the two overlap"],"url":"/docs/improving-ai-visibility/backlinks-and-citations#where-the-two-overlap","text":"A good page on a site that buyers read usually gets both a backlink and a citation. The overlap is large, thus good work serves both purposes. The mistake is work in the shape of link building, which serves only one purpose, and frequently neither.","keywords":""},{"kind":"section","title":"Backlinks and citations","heading":"What to do","section":"Documentation","crumbs":["Improving AI visibility","Backlinks and citations","What to do"],"url":"/docs/improving-ai-visibility/backlinks-and-citations#what-to-do","text":"1. Find the sources that answer engines already cite in your category. 2. Examine what these sources say about you. 3. Correct the incorrect information, and add the missing information. 4. Get a place in the sources where you are absent. A link with the mention is of secondary importance.","keywords":""},{"kind":"page","title":"Updating content","heading":"","section":"Documentation","crumbs":["Documentation","Improving AI visibility","Updating content"],"url":"/docs/improving-ai-visibility/updating-content","text":"Existing pages are often a better investment than new pages. Learn which pages to update first, why a new date has no effect, and what to expect next. Existing pages are usually a better investment than new pages, because retrieval already works for them. Only the quotation, the accuracy, or the completeness fails, and you can correct these with edits.","keywords":""},{"kind":"section","title":"Updating content","heading":"What to update first","section":"Documentation","crumbs":["Improving AI visibility","Updating content","What to update first"],"url":"/docs/improving-ai-visibility/updating-content#what-to-update-first","text":"- Pages for questions where you appear but answer engines do not recommend you. The answer engine already finds you and selects a different brand. Usually, the answers tell you why. - Pages that answer engines retrieve but never quote. Apply the test from Structuring content for LLMs. If no paragraph is self-contained, correct that problem. - Pages that newer sources contradict. Your page and a cited third-party source can disagree about your own product. Then, answer engines give an uncertain answer, or they use the third-party source. - All content with old facts. For example: old pricing, removed features, or replaced integrations. Answer engines copy old facts into answers. The cost of these errors is high, but it is not easy to see.","keywords":""},{"kind":"section","title":"Updating content","heading":"What an update is not","section":"Documentation","crumbs":["Improving AI visibility","Updating content","What an update is not"],"url":"/docs/improving-ai-visibility/updating-content#what-an-update-is-not","text":"An update is not a new date on a page that you publish again. Answer engines read the text. A changed timestamp has no effect.","keywords":""},{"kind":"section","title":"Updating content","heading":"After an update","section":"Documentation","crumbs":["Improving AI visibility","Updating content","After an update"],"url":"/docs/improving-ai-visibility/updating-content#after-an-update","text":"Record the date, and continue to run the topic. Expect a delay. Answer engines must retrieve the page again and decide that it is better than the previous source. One run is not a verdict. Read Monitoring AI visibility.","keywords":""},{"kind":"page","title":"Monitoring AI visibility","heading":"","section":"Documentation","crumbs":["Documentation","Improving AI visibility","Monitoring AI visibility"],"url":"/docs/improving-ai-visibility/monitoring-ai-visibility","text":"AI visibility has a long feedback loop with much noise. Learn a weekly schedule, how to read a change before you act, and how to report direction and cause. Visibility work has a long feedback loop with much noise. Good monitoring is mostly about patience: do not make conclusions too early.","keywords":""},{"kind":"section","title":"Monitoring AI visibility","heading":"A practical schedule","section":"Documentation","crumbs":["Improving AI visibility","Monitoring AI visibility","A practical schedule"],"url":"/docs/improving-ai-visibility/monitoring-ai-visibility#a-practical-schedule","text":"- Run conversations weekly on a stable set of topics. This frequency is sufficient to find a real change, and it is not so frequent that noise hides the signal. Read How often should I run conversations?. - Keep the set of questions fixed between runs. If you change scenarios while you watch a trend, you measure two different things. - Record what you changed, and when. Almost each \"why did this change\" question has its answer in a note that nobody wrote.","keywords":""},{"kind":"section","title":"Monitoring AI visibility","heading":"How to read a change","section":"Documentation","crumbs":["Improving AI visibility","Monitoring AI visibility","How to read a change"],"url":"/docs/improving-ai-visibility/monitoring-ai-visibility#how-to-read-a-change","text":"Before you investigate a change, do these checks in sequence: 1. Find if your competitors changed in the same direction at the same time. If the full category changed, the answer engine changed, not you. This one check prevents most of the wasted investigation. 2. Find if the change occurs in all topics or only in some topics. A change in some topics is usually yours. A change in all topics is usually not yours. 3. Open the conversations from before and after the change, and compare the cited sources. A changed answer almost always means a changed source.","keywords":""},{"kind":"section","title":"Monitoring AI visibility","heading":"What to report","section":"Documentation","crumbs":["Improving AI visibility","Monitoring AI visibility","What to report"],"url":"/docs/improving-ai-visibility/monitoring-ai-visibility#what-to-report","text":"Report the direction and the cause, not the number. This sentence is a result: \"Visibility increased by nine points on pricing topics after we published the comparison page, and answer engines now cite it.\" In contrast, \"Visibility is 62%\" causes a question that nobody can answer.","keywords":""},{"kind":"section","title":"Monitoring AI visibility","heading":"Measure what is important to you","section":"Documentation","crumbs":["Improving AI visibility","Monitoring AI visibility","Measure what is important to you"],"url":"/docs/improving-ai-visibility/monitoring-ai-visibility#measure-what-is-important-to-you","text":"Platform metrics describe the medium. You possibly must track a specific item for your business. For example: a recommendation for a specific phrase, in one market, against two named competitors. Use the Business Scorecard for this.","keywords":""},{"kind":"page","title":"Shopping overview","heading":"","section":"Documentation","crumbs":["Documentation","Shopping","Shopping overview"],"url":"/docs/shopping/shopping-overview","text":"The Shopping Overview dashboard shows the AI visibility, recommendation and share of voice of your products, with Top 10 leaderboards and trend charts. This page explains the Shopping Overview, the AI shopping dashboard for the performance of your products in answer engines. It is the first page that you open in the shopping section. You use it for weekly checks and for monthly reports. The brand dashboards answer the question \"how does my brand perform in AI answers?\". The Shopping Overview answers a different question: \"which of my products get a good result, and which products do the answer engines omit?\". Refer to Product Visibility for what product tracking is and why it is important. Shopping is an enterprise feature. The Genezio team enables it for a brand.","keywords":""},{"kind":"section","title":"Shopping overview","heading":"What the Shopping Overview dashboard shows","section":"Documentation","crumbs":["Shopping","Shopping overview","What the Shopping Overview dashboard shows"],"url":"/docs/shopping/shopping-overview#what-the-shopping-overview-dashboard-shows","text":"","keywords":""},{"kind":"section","title":"Shopping overview","heading":"The three product metrics","section":"Documentation","crumbs":["Shopping","Shopping overview","The three product metrics"],"url":"/docs/shopping/shopping-overview#the-three-product-metrics","text":"The main numbers are at the top of the page: - Product Visibility: how frequently your products appear in eligible conversations. - Product Recommendation: in the conversations where a product appears, how frequently the answer engine recommends it. - Share of Voice: how much of the product conversation is yours, compared with the competing products. These metrics use the same formulas as the brand KPIs, but for products. Refer to Insights: Share of Voice for how to read share.","keywords":""},{"kind":"section","title":"Shopping overview","heading":"Top 10 leaderboards","section":"Documentation","crumbs":["Shopping","Shopping overview","Top 10 leaderboards"],"url":"/docs/shopping/shopping-overview#top-10-leaderboards","text":"For each of the three metrics, the dashboard shows a Top 10 product leaderboard. It shows your ten strongest products for that metric.","keywords":""},{"kind":"section","title":"Shopping overview","heading":"Trend charts","section":"Documentation","crumbs":["Shopping","Shopping overview","Trend charts"],"url":"/docs/shopping/shopping-overview#trend-charts","text":"Each metric has a trend chart. The chart shows how the metric changed across the selected period. Thus, you see the direction, not only one snapshot.","keywords":""},{"kind":"section","title":"Shopping overview","heading":"Change from the previous period","section":"Documentation","crumbs":["Shopping","Shopping overview","Change from the previous period"],"url":"/docs/shopping/shopping-overview#change-from-the-previous-period","text":"Each metric shows a percentage change from the previous period. Thus, you see immediately if the metric becomes better, stays the same, or becomes worse.","keywords":""},{"kind":"section","title":"Shopping overview","heading":"Total retailers card","section":"Documentation","crumbs":["Shopping","Shopping overview","Total retailers card"],"url":"/docs/shopping/shopping-overview#total-retailers-card","text":"A total-retailers card shows how many retailers are related to your products in the conversations that Genezio analyzed. Refer to Merchants and Retailers for the details.","keywords":""},{"kind":"section","title":"Shopping overview","heading":"How to read the shopping metrics","section":"Documentation","crumbs":["Shopping","Shopping overview","How to read the shopping metrics"],"url":"/docs/shopping/shopping-overview#how-to-read-the-shopping-metrics","text":"Start with the three metrics, in sequence. Visibility tells you the reach. Recommendation tells you if that reach changes into recommendations. Share of Voice tells you your position against the competing products. A drop in Visibility and a drop in Recommendation are different problems. They need different solutions. Read the leaderboards as three different questions. One product is rarely at the top of all three lists. The differences give you the insight: - Top by Visibility: the products that the answer engines mention most. These products carry your presence. - Top by Recommendation: the products that the answer engines recommend actively. These are your strongest products, also when the answer engines mention them less frequently. - Top by Share of Voice: the products that own their competitive space. The most useful comparison is what is missing. Maybe the product that you promote now is not on a leaderboard. Then the answer engines do not carry your campaign. That is a problem of content and sources, not a problem of media. A large change from the previous period is a reason to investigate. It is not a conclusion. A large change usually has one of three causes: - Your content or your coverage changed. - A competitor got or lost position. - The set of detected products changed, because the answer engines started to name a new product. Look at the trend chart to see if the change is a spike or a stable direction. Then open Products to see the products that caused it. Remember the source of the products. Genezio detects products automatically from the conversations with answer engines only. There is no catalog upload and no feed. Thus, the dashboard shows the products that the answer engines talk about, not your full catalog.","keywords":""},{"kind":"section","title":"Shopping overview","heading":"Ask Geo a question from the shopping cards","section":"Documentation","crumbs":["Shopping","Shopping overview","Ask Geo a question from the shopping cards"],"url":"/docs/shopping/shopping-overview#ask-geo-a-question-from-the-shopping-cards","text":"The shopping cards have one-click questions with context. These questions use your real numbers. You do not have to read a metric and then decide what to do. Click a question, and you get an answer from your data. For example: We have 0% visibility — where do we start? The question includes the context of its card. Thus, a number on a dashboard becomes a next action. Refer to Geo Assistant for all the things that Geo can do.","keywords":""},{"kind":"section","title":"Shopping overview","heading":"Export the shopping data","section":"Documentation","crumbs":["Shopping","Shopping overview","Export the shopping data"],"url":"/docs/shopping/shopping-overview#export-the-shopping-data","text":"- General PDF export: the standard PDF export includes the shopping pages and the product data, with the other reports of your brand. - Product export: a separate export for the product data only. Use it when you need the list of products and their metrics separately.","keywords":""},{"kind":"section","title":"Shopping overview","heading":"Related pages","section":"Documentation","crumbs":["Shopping","Shopping overview","Related pages"],"url":"/docs/shopping/shopping-overview#related-pages","text":"- Product Visibility - Products - Merchants and Retailers - Geo Assistant - Insights: Share of Voice","keywords":""},{"kind":"section","title":"Shopping overview","heading":"In the API","section":"Documentation","crumbs":["Shopping","Shopping overview","In the API"],"url":"/docs/shopping/shopping-overview#in-the-api","text":"The products report gives the product metrics over time.","keywords":""},{"kind":"page","title":"Product visibility","heading":"","section":"Documentation","crumbs":["Documentation","Shopping","Product visibility"],"url":"/docs/shopping/product-visibility","text":"Product visibility measures how often answer engines name and recommend each of your products. Learn how Genezio detects products and calculates the metrics. Product Visibility measures how answer engines talk about each of your products, not only about your brand. This page explains how Genezio detects the products and how it calculates the product metrics. At the brand level, Genezio answers the question \"how do AI assistants talk about my brand?\". Shopping and Product Visibility answer a more precise question: \"how do AI assistants talk about my products?\" This question becomes more important each month. Answer engines frequently do not stop at the name of a brand. They name a specific product: a model number, a SKU, or a variant. The shopper then buys that named product.","keywords":""},{"kind":"section","title":"Product visibility","heading":"Brand visibility compared with product visibility","section":"Documentation","crumbs":["Shopping","Product visibility","Brand visibility compared with product visibility"],"url":"/docs/shopping/product-visibility#brand-visibility-compared-with-product-visibility","text":"Brand Visibility tells you if your company shows in the answer. Product Visibility tells you which of your products shows in the answer. This is the difference: - Brand level: \"For running shoes, the AI mentions your brand in most answers.\" This is good, but it does not tell you which shoe. - Product level: \"When it mentions your brand, it names the model of last season, not the model that you promote now.\" Two brands can have the same brand visibility and very different commercial results. One brand has its main product in the answer. The other brand has a discontinued line in the answer. The product data shows this gap.","keywords":""},{"kind":"section","title":"Product visibility","heading":"Why product visibility is important for marketers","section":"Documentation","crumbs":["Shopping","Product visibility","Why product visibility is important for marketers"],"url":"/docs/shopping/product-visibility#why-product-visibility-is-important-for-marketers","text":"- The named product gets the sale. When an AI assistant recommends a specific product, the shopper often considers only the named products. It is not sufficient to be the brand in the answer. - You find which products the answer engines know. Your catalog and the products that the answer engines talk about are rarely the same list. Product Visibility shows you the real overlap. - You can measure launches and campaigns for each product. You can see if the product that you push now starts to appear in the answers. Or you can see if an older model still gets the attention. - Competitive comparisons become concrete. You do not get \"we lose share in this category\". You get \"we lose with this specific product against this specific competitor product\".","keywords":""},{"kind":"section","title":"Product visibility","heading":"How Genezio detects products in AI answers","section":"Documentation","crumbs":["Shopping","Product visibility","How Genezio detects products in AI answers"],"url":"/docs/shopping/product-visibility#how-genezio-detects-products-in-ai-answers","text":"Genezio detects products automatically, from the conversations with answer engines only. Genezio reads the conversations that it runs on answer engines. It finds the products that these engines mention. There is: - No catalog upload - No product feed - No site crawl - No manual product entry The result: the list of products is not your catalog. It is the set of products that the answer engines talk about. If a product never appears in an answer, it does not appear in Genezio. This absence is also a finding, and usually you can act on it.","keywords":""},{"kind":"section","title":"Product visibility","heading":"How Genezio measures product visibility","section":"Documentation","crumbs":["Shopping","Product visibility","How Genezio measures product visibility"],"url":"/docs/shopping/product-visibility#how-genezio-measures-product-visibility","text":"The product metrics use the same formulas as the brand KPIs. The only difference is the scope: one product, not the full brand. These are the brand definitions: For a product, the metrics are: - Product Visibility: how frequently this product appears in eligible conversations. This is the reach number. - Product Recommendation: in the conversations where this product appears, how frequently the answer engine recommends it. This is the conversion number. - Share of Voice: how much of the conversation is about this product, compared with the competing products in the same space. The formulas are the same as the brand KPIs. Thus, all that you know about how to read those numbers applies here too. Refer to Insights: Your KPIs Explained for the full explanation. Refer to Insights: Share of Voice for how to read share. Compare the two metrics together: - High Product Visibility and low Product Recommendation: the answer engines know the product, but they do not recommend it. - Low visibility and high recommendation: the product wins when an answer engine mentions it, but the answer engines do not mention it frequently.","keywords":"AI Visibility % = conversations where it appears / total eligible conversations AI Recommendations % = conversations where recommended / conversations where visible"},{"kind":"section","title":"Product visibility","heading":"Availability of product tracking","section":"Documentation","crumbs":["Shopping","Product visibility","Availability of product tracking"],"url":"/docs/shopping/product-visibility#availability-of-product-tracking","text":"Shopping and Product Visibility is an enterprise feature. The Genezio team enables it for a brand, as your contract specifies. The interface has no self-serve switch. To enable product tracking for your brand, ask the Genezio team. The team enables it. After that, the shopping pages appear in your navigation. Genezio then starts to detect products in the conversations that it runs.","keywords":""},{"kind":"section","title":"Product visibility","heading":"The shopping pages after activation","section":"Documentation","crumbs":["Shopping","Product visibility","The shopping pages after activation"],"url":"/docs/shopping/product-visibility#the-shopping-pages-after-activation","text":"- Shopping Overview: the dashboard. It has the three metrics, Top 10 product leaderboards, trend charts, the change from the previous period, and a total-retailers card. - Products: the full list of products that Genezio detected for your brand. - Product Details: a drawer with the detail view of one product. - Merchants and Retailers: the retailers that the answer engines relate to your products. The conversation transcripts also show the products. Genezio tags and highlights the products in the text. Tables and maps show in the transcript. Thus, you can see exactly how the answer presented a product.","keywords":""},{"kind":"section","title":"Product visibility","heading":"Related pages","section":"Documentation","crumbs":["Shopping","Product visibility","Related pages"],"url":"/docs/shopping/product-visibility#related-pages","text":"- Shopping Overview - Products - Merchants and Retailers - Core Concepts: Brand Visibility - Insights: Your KPIs Explained - Insights: Share of Voice","keywords":""},{"kind":"section","title":"Product visibility","heading":"In the API","section":"Documentation","crumbs":["Shopping","Product visibility","In the API"],"url":"/docs/shopping/product-visibility#in-the-api","text":"The products report gives the visibility, recommendation, and share of voice of the products over time.","keywords":""},{"kind":"page","title":"Products","heading":"","section":"Documentation","crumbs":["Documentation","Shopping","Products"],"url":"/docs/shopping/products","text":"Use the Products page to track each product that answer engines mention, sort by visibility and recommendation, and find the competitor products in its answers. This page explains how to use the Products page for AI product tracking: the full list of your products that the answer engines mention. It is not the catalog that you sell. It is the catalog that the answer engines talk about. Each row shows a product, its image, and its performance in the answers. You do not upload data to this page. Genezio detects products automatically, from the conversations with answer engines only. There is no catalog upload, no product feed, no crawl, and no manual entry. If a product is on this page, an answer engine named it. This page is not the same as Master Filters. The old name of Master Filters was \"Products\". Master Filters put topics into groups, so that you can divide your reports by product line. The Products page shows the products that the answer engines mention, which Genezio finds automatically. These are two different features. Shopping and Product Visibility is an enterprise feature. The Genezio team enables it for a brand, as your contract specifies. There is no self-serve switch. If you do not see the Shopping section, speak to your Genezio contact.","keywords":""},{"kind":"section","title":"Products","heading":"What the Products page shows","section":"Documentation","crumbs":["Shopping","Products","What the Products page shows"],"url":"/docs/shopping/products#what-the-products-page-shows","text":"The page shows each product that the answer engines named in your tracked conversations, with: - The product image. Thus, you can examine the list visually. You do not have to read SKU names. - A Product Visibility column: how frequently the product appears in the conversations for which it was eligible. - A Product Recommendation column: when the product appeared, how frequently the answer engine recommended it. - Co-mentioned products: the products that the answer engines name together with it. The two metrics use the same formulas as the brand KPIs, for one product: - Product Visibility = conversations in which the product appears ÷ eligible conversations - Product Recommendation = conversations in which the answer engine recommended it ÷ conversations in which it was visible Thus, you can compare the numbers of a product directly with the numbers of your brand, and with the numbers of other products.","keywords":""},{"kind":"section","title":"Products","heading":"Search and filter the products","section":"Documentation","crumbs":["Shopping","Products","Search and filter the products"],"url":"/docs/shopping/products#search-and-filter-the-products","text":"You can search and filter the list. This is important when the answer engines name dozens of your products. - Search by product name to go directly to the item that you need. The search helps when you prepare for a category review or a launch retrospective. - Filter the list to the part that you work on. Then sort in it. Start with a filter, not with the full list. A set of ten products tells a clearer story than the full list.","keywords":""},{"kind":"section","title":"Products","heading":"Sort the products to find priorities","section":"Documentation","crumbs":["Shopping","Products","Sort the products to find priorities"],"url":"/docs/shopping/products#sort-the-products-to-find-priorities","text":"You can sort the Visibility column and the Recommendation column. A sort is the fastest way to change the list into a to-do list, because the two metrics answer different questions. 1. Sort by Visibility, descending. These are the products that the answer engines know. They are your strongest assets in the answer engines. 2. Sort by Visibility, ascending. These are the products that the answer engines rarely show. Usually, the product is absent from the sources that the answer engines read. Or the sources describe it in words that buyers do not use. 3. Sort by Recommendation, ascending. Then look at the rows with high visibility. The third pattern is the important one: - High visibility and low recommendation: the answer engine names the product but does not select it. The product is in the consideration set and it loses. This is usually a problem of positioning, reviews, or comparison content. It is not a problem of awareness. - Low visibility and high recommendation: when the answer engine finds the product, it likes it. The product must have more presence in the sources that the answer engines read. - High visibility and high recommendation: these are your reference products. Examine their coverage and do the same for other products. - Low visibility and low recommendation: the product is almost invisible in shopping answers. Decide if it is worth an investment. Open a row to see the reasons for the numbers in Product Details.","keywords":""},{"kind":"section","title":"Products","heading":"Co-mentioned products and competitor products","section":"Documentation","crumbs":["Shopping","Products","Co-mentioned products and competitor products"],"url":"/docs/shopping/products#co-mentioned-products-and-competitor-products","text":"Answer engines rarely name one product alone. They make a shortlist. Co-mentioned products are the other products that an answer engine names in the same answer. The co-mentioned products view shows you that shortlist. It has two types.","keywords":""},{"kind":"section","title":"Products","heading":"Your products named together","section":"Documentation","crumbs":["Shopping","Products","Your products named together"],"url":"/docs/shopping/products#your-products-named-together","text":"When an answer engine names two of your products in the same answer, it puts them in one group. They can be alternatives, steps on a good, better, and best ladder, or products that go together. What to do: - Two products are frequently together and one is clearly a lower version of the other. Then check if the answer engine describes the difference as you want. - Products that you see as complements are frequently together. Then your content can use this for bundles and cross-sell. - The answer engines frequently confuse two products that you see as different. Then your product pages and the third-party coverage probably do not show the difference clearly.","keywords":""},{"kind":"section","title":"Products","heading":"Competitor products that take your position","section":"Documentation","crumbs":["Shopping","Products","Competitor products that take your position"],"url":"/docs/shopping/products#competitor-products-that-take-your-position","text":"The same view shows the competitor products that the answer engines name together with yours. These products take the position that you want. This is the most actionable information on the page. For each product, it tells you the products that the answer engines compare with yours. That set is frequently different from the competitors that your team expects. What to do: - Use the competitor products that occur frequently as your real comparison set. Make sure that you have comparison content for them. - Look for competitor products that occur against many of your products. That competitor wins the category, not only one comparison. - Compare the list with your tracked Competitors. If a competitor product frequently takes your position but you do not track its brand, add the brand.","keywords":""},{"kind":"section","title":"Products","heading":"Where to start your first product review","section":"Documentation","crumbs":["Shopping","Products","Where to start your first product review"],"url":"/docs/shopping/products#where-to-start-your-first-product-review","text":"For your first review: 1. Filter to the category that you own. 2. Sort by Visibility, descending. Make sure that the products that you expect at the top are at the top. 3. Sort by Recommendation, ascending, and look at the products with high visibility. These give the fastest results. 4. Open the co-mentions of your top three products to see which products take the position. 5. Open Product Details for the worst product to see exactly what the answer engines say about it. For the view of all the products together, refer to Shopping Overview.","keywords":""},{"kind":"section","title":"Products","heading":"Related pages","section":"Documentation","crumbs":["Shopping","Products","Related pages"],"url":"/docs/shopping/products#related-pages","text":"- Product Details - Product Visibility - Shopping Overview - Merchants and Retailers - Master Filters - Competitors","keywords":""},{"kind":"section","title":"Products","heading":"In the API","section":"Documentation","crumbs":["Shopping","Products","In the API"],"url":"/docs/shopping/products#in-the-api","text":"List the products gives the products of a brand, and the products of the same answers gives the co-mentioned products.","keywords":""},{"kind":"page","title":"Product details","heading":"","section":"Documentation","crumbs":["Documentation","Shopping","Product details"],"url":"/docs/shopping/product-details","text":"Learn what the Product Details drawer shows about one product in AI answers: its perception, SWOT, shopping queries, conversations and retailer offers. This page explains the Product Details drawer, which shows how AI answers describe and recommend one product. The drawer gives the full story of the product: - How the answer engines describe it. - The shopping questions in which it appears. - The exact conversations that named it. - Where the answer engines say that you can buy it. The Products page tells you which products need attention. The drawer tells you why. Usually, it also tells you what to do.","keywords":""},{"kind":"section","title":"Product details","heading":"Open the Product Details drawer","section":"Documentation","crumbs":["Shopping","Product details","Open the Product Details drawer"],"url":"/docs/shopping/product-details#open-the-product-details-drawer","text":"The drawer opens for each product, from two places: - The Products page: click a product row. - A leaderboard: click a product in its ranked position. The drawer opens over the current view. When you close it, you go back to the same position in the list. Thus, you can examine many products in one session.","keywords":""},{"kind":"section","title":"Product details","heading":"How the answer engines see the product","section":"Documentation","crumbs":["Shopping","Product details","How the answer engines see the product"],"url":"/docs/shopping/product-details#how-the-answer-engines-see-the-product","text":"The drawer starts with the perception of the product. This is how the answer engines describe it in their own words. Why this is important: your product page says one thing, and the answer engine can say a different thing. Buyers hear the perception. For example, an answer engine can describe your premium model as \"the budget option\". More money for ads does not solve this. The solution is in the sources that the answer engines read. The perception of a product works the same as the perception of a brand. Refer to Perceptions.","keywords":""},{"kind":"section","title":"Product details","heading":"The SWOT of the product","section":"Documentation","crumbs":["Shopping","Product details","The SWOT of the product"],"url":"/docs/shopping/product-details#the-swot-of-the-product","text":"The drawer includes a SWOT for the product: the strengths, weaknesses, opportunities, and threats. Genezio makes the SWOT from what the answer engines say about the product. Why this is important: the SWOT changes a large quantity of AI text into four groups that you can act on: - Weaknesses tell you which objection your content must answer. - Threats usually name the competitor products that win against you. - Strengths tell you which messages already work. You can use them more. Refer to SWOT Analysis for how Genezio makes a SWOT.","keywords":""},{"kind":"section","title":"Product details","heading":"The shopping queries of the product","section":"Documentation","crumbs":["Shopping","Product details","The shopping queries of the product"],"url":"/docs/shopping/product-details#the-shopping-queries-of-the-product","text":"The drawer shows the shopping queries in which the product appears. These are the real buying questions that showed the product. Why this is important: this is intent data. It tells you the buying moments that your product already owns. It also tells you, by their absence, the buying moments that it does not own. For example, you positioned a product for travel, but it never appears in travel queries. Then your positioning did not get to the answer engines.","keywords":""},{"kind":"section","title":"Product details","heading":"The query fanouts of the shopping queries","section":"Documentation","crumbs":["Shopping","Product details","The query fanouts of the shopping queries"],"url":"/docs/shopping/product-details#the-query-fanouts-of-the-shopping-queries","text":"For these queries, the drawer shows the query fanouts. A query fanout is a follow-up question that an answer engine makes when it researches a buying question. Why this is important: buyers do not ask one question and stop. Query fanouts show the chain: the comparison, the price check, the \"is it worth it\" question. At each step of the chain, an answer engine can include your product or remove it. Each step is a target for content. Refer to Query Fanouts.","keywords":""},{"kind":"section","title":"Product details","heading":"The conversations that named the product","section":"Documentation","crumbs":["Shopping","Product details","The conversations that named the product"],"url":"/docs/shopping/product-details#the-conversations-that-named-the-product","text":"The drawer shows each conversation in which the product occurred. You can open the full transcript. These transcripts show the products clearly: - The transcript tags and highlights the products in the text. Thus, you can quickly find your products and the products of your competitors. - Tables show in the transcript. Thus, you read the comparison grids as the buyer saw them. - Maps show in the transcript. Thus, you can read answers about locations. Why this is important: the metrics tell you that a product loses. The transcript shows you the sentence where it lost. You can see the exact moment when the answer engine recommended your product. Or you can see when it named your product and then selected a different one, with its reason in the text. Correct the cause that this reason shows. Refer to Conversations.","keywords":""},{"kind":"section","title":"Product details","heading":"Retailer offers, with prices and links","section":"Documentation","crumbs":["Shopping","Product details","Retailer offers, with prices and links"],"url":"/docs/shopping/product-details#retailer-offers-with-prices-and-links","text":"The drawer shows the retailer offers that the answer engines give for the product, with prices and links. Why this is important: shopping answers do not only name a product. They send the buyer to a store. The purchase occurs in that store. This section also shows price problems first. For example, the answer engine can give an old price. Or it can send the buyer to a grey-market listing that has a lower price than your own store.","keywords":""},{"kind":"section","title":"Product details","heading":"The retailers that sell the product","section":"Documentation","crumbs":["Shopping","Product details","The retailers that sell the product"],"url":"/docs/shopping/product-details#the-retailers-that-sell-the-product","text":"Next to the offers, the drawer shows the retailers that sell the product, as the answer engines say. Why this is important: this list is a distribution check from the system that your buyers ask: - A missing retailer means that the answer engine does not know that the retailer sells your product. - An unexpected retailer means that a company lists your product and you do not know about it. - If your own direct-to-consumer store is not on the list, the answer engine sends each buyer to a third party. For the view of the retailers across all products, refer to Merchants and Retailers.","keywords":""},{"kind":"section","title":"Product details","heading":"A practical sequence to diagnose a product","section":"Documentation","crumbs":["Shopping","Product details","A practical sequence to diagnose a product"],"url":"/docs/shopping/product-details#a-practical-sequence-to-diagnose-a-product","text":"For a product with high visibility and low recommendation: 1. Read the perception. Make sure that the answer engines describe the product as you intend. 2. Read the SWOT weaknesses and threats. Find the objection and the competitor. 3. Look at the shopping queries. Make sure that the product appears in the correct buying moments. 4. Open two or three conversations. Find the sentence where the answer engine did not select the product. 5. Look at the retailer offers. Find if a problem with price or availability is the real cause. After step five, you usually have a specific brief that you can write, not only a metric.","keywords":""},{"kind":"section","title":"Product details","heading":"Availability of Product Details","section":"Documentation","crumbs":["Shopping","Product details","Availability of Product Details"],"url":"/docs/shopping/product-details#availability-of-product-details","text":"Shopping and Product Visibility is an enterprise feature. The Genezio team enables it for a brand, as your contract specifies. There is no self-serve switch. Genezio detects the products automatically, from the conversations with answer engines only. There is no catalog upload, no feed, no crawl, and no manual entry.","keywords":""},{"kind":"section","title":"Product details","heading":"Related pages","section":"Documentation","crumbs":["Shopping","Product details","Related pages"],"url":"/docs/shopping/product-details#related-pages","text":"- Products - Product Visibility - Shopping Overview - Merchants and Retailers - Perceptions - SWOT Analysis - Query Fanouts - Conversations","keywords":""},{"kind":"section","title":"Product details","heading":"In the API","section":"Documentation","crumbs":["Shopping","Product details","In the API"],"url":"/docs/shopping/product-details#in-the-api","text":"Get a product gives one product. What the answers say about a product gives its perception, and the sources of a product gives its sources.","keywords":""},{"kind":"page","title":"Merchants and retailers","heading":"","section":"Documentation","crumbs":["Documentation","Shopping","Merchants and retailers"],"url":"/docs/shopping/merchants-and-retailers","text":"Find the retailers that ChatGPT and other answer engines cite for your product category, with their share of voice, prices and source conversations. This page explains how Genezio shows the retailers in AI answers: the merchants that answer engines cite when they recommend products in your category. Merchant and retailer intelligence also shows: - The prices of those retailers in the answers. - How frequently each retailer occurs. - The conversations behind each mention. This feature is part of Shopping and Product Visibility. For the full view of how Genezio tracks what the answer engines say about each of your products, start with Product Visibility.","keywords":""},{"kind":"section","title":"Merchants and retailers","heading":"Why the retailers in AI answers are important","section":"Documentation","crumbs":["Shopping","Merchants and retailers","Why the retailers in AI answers are important"],"url":"/docs/shopping/merchants-and-retailers#why-the-retailers-in-ai-answers-are-important","text":"When an answer engine recommends a product, it usually tells the shopper where to buy it. In the answers, those retailers are the shelf of your product. You do not control which retailers are on it. When you know which retailers have the largest part of that shelf, and at which prices, you know: - Where to put your work on retail and channel relationships. The retailers that the answer engines use most are the ones that are worth an investment. - Where your prices are incorrect. The shoppers see the prices that the answer engines show. If these prices are not the prices that you intend, speak to the channel. - Which retailers do not sell your products at all. A retailer can occur frequently for your category but never sell your products. That is a gap that you can see.","keywords":""},{"kind":"section","title":"Merchants and retailers","heading":"What the merchant drawer shows about a retailer","section":"Documentation","crumbs":["Shopping","Merchants and retailers","What the merchant drawer shows about a retailer"],"url":"/docs/shopping/merchants-and-retailers#what-the-merchant-drawer-shows-about-a-retailer","text":"The retailer intelligence is in the merchant drawer. Open a retailer to see: - Retailer share of voice: the percentage of the retailer citations in your category that this merchant has. This number tells you how large the shelf of the retailer is. - Prices: the prices of that retailer in the answers. - Occurrences: how many times the retailer occurred. - The conversations behind each occurrence: the real conversations in which an answer engine cited the retailer. Thus, you can read the source of each number. The last item is the most important. Share of voice tells you that a retailer is large. The conversations tell you why the answer engines send shoppers to that retailer. For the general metric, refer to Share of Voice. The Shopping Overview dashboard also has a total-retailers card. It shows the number of different retailers that the answer engines cite across your category. Refer to Shopping Overview.","keywords":""},{"kind":"section","title":"Merchants and retailers","heading":"Open the merchant drawer","section":"Documentation","crumbs":["Shopping","Merchants and retailers","Open the merchant drawer"],"url":"/docs/shopping/merchants-and-retailers#open-the-merchant-drawer","text":"Go to the shopping area of the Genezio dashboard and open Shopping Overview. Look at the total-retailers card to see the number of retailers in your category. Open the retailer view. Select a merchant to open the merchant drawer. Read the share of voice, the prices, and the occurrences at the top of the drawer. Open the linked conversations. See the exact answers that cited that retailer.","keywords":""},{"kind":"section","title":"Merchants and retailers","heading":"How to read the retailer data","section":"Documentation","crumbs":["Shopping","Merchants and retailers","How to read the retailer data"],"url":"/docs/shopping/merchants-and-retailers#how-to-read-the-retailer-data","text":"Start with the largest shelf. Then examine the smaller shelves: - Start with share of voice. Usually, the first few retailers have most of what the answer engines recommend in your category. Those are the important relationships. - Then look at the prices. Compare the prices in the answers with the prices that you expect from that retailer. A difference is a price problem or a feed problem. Tell the channel about it. - Use the occurrences to check consistency. A retailer that occurs frequently is a stable part of the category shelf. It is not a single event. - Read the conversations. They give you the context: the questions, the products, and the comparisons that put that retailer in the answer.","keywords":""},{"kind":"section","title":"Merchants and retailers","heading":"Connect the retailers to the offers of a product","section":"Documentation","crumbs":["Shopping","Merchants and retailers","Connect the retailers to the offers of a product"],"url":"/docs/shopping/merchants-and-retailers#connect-the-retailers-to-the-offers-of-a-product","text":"Retailer intelligence works in two directions. The merchant drawer starts from the retailer. It answers the question \"who does AI cite in my category?\". The detail drawer of a product starts from the product. It shows the retailer offers with prices and links, and the retailers that sell the product. Refer to Product Details. Together, the two drawers take you from \"this retailer has the largest part of my category\" to \"my products are on its shelf, or not, and at this price\". Use Products to go from one product to a different one.","keywords":""},{"kind":"section","title":"Merchants and retailers","heading":"Actions from the retailer data","section":"Documentation","crumbs":["Shopping","Merchants and retailers","Actions from the retailer data"],"url":"/docs/shopping/merchants-and-retailers#actions-from-the-retailer-data","text":"- Channel and retail relationships. Give priority to the retailers that the answer engines cite. Do not give priority only to the retailers with the largest offline presence. - Price checks. Compare the prices in the answers across retailers. Flag the retailers that show your product incorrectly. - Gaps of absence. Find the retailers that occur frequently in your category but do not sell your products. Make them distribution targets. - Evidence for internal discussions. Each number links to real conversations. Thus, your discussions about channels and prices use what the answer engines tell shoppers, not assumptions.","keywords":""},{"kind":"section","title":"Merchants and retailers","heading":"Availability of merchant intelligence","section":"Documentation","crumbs":["Shopping","Merchants and retailers","Availability of merchant intelligence"],"url":"/docs/shopping/merchants-and-retailers#availability-of-merchant-intelligence","text":"Shopping and Product Visibility, with merchant and retailer intelligence, is an enterprise feature. The Genezio team enables it for a brand, as your contract specifies. There is no self-serve switch. If you do not see the shopping area in your dashboard, speak to the Genezio team. Genezio detects the products and the retailers automatically, from the conversations with answer engines only. There is no upload, no feed, no crawl, and no manual entry. If a retailer is in the merchant drawer, an answer engine cited it in a real conversation.","keywords":""},{"kind":"section","title":"Merchants and retailers","heading":"Related pages","section":"Documentation","crumbs":["Shopping","Merchants and retailers","Related pages"],"url":"/docs/shopping/merchants-and-retailers#related-pages","text":"- Product Visibility: what shopping visibility tracks and why it is important - Shopping Overview: the dashboard, with the total-retailers card - Products: the products that Genezio detects in the conversations - Product Details: the retailer offers, prices, and links of one product - Citations: how Genezio records the sources of the answers - Conversations: the conversations behind each metric","keywords":""},{"kind":"section","title":"Merchants and retailers","heading":"In the API","section":"Documentation","crumbs":["Shopping","Merchants and retailers","In the API"],"url":"/docs/shopping/merchants-and-retailers#in-the-api","text":"List the retailers gives the retailers that the answer engines cite.","keywords":""},{"kind":"page","title":"Agentic commerce readiness","heading":"","section":"Documentation","crumbs":["Documentation","Agentic commerce","Agentic commerce readiness"],"url":"/docs/agentic-commerce/agentic-commerce-readiness","text":"Agentic commerce readiness shows if AI shopping agents can read your store and buy from it. Learn what UCP is and how Genezio measures your readiness. This page explains agentic commerce, the Universal Commerce Protocol (UCP), and how Genezio measures the agentic commerce readiness of your store. Agentic commerce is the change from people who browse your store to AI agents that browse it for them. Shopping assistants in tools such as ChatGPT start to browse, compare, and also buy for a user. For example, a shopper asks an assistant to \"find me a durable rain jacket under $150 and order it\". Then the agent, not the person, visits the stores, reads the product data, and makes the selection. Agentic commerce readiness tells you how ready your store is for these agents. The agents must understand your store and act on it. If an agent cannot read your catalog or get to your endpoints, it cannot recommend you. It also cannot buy from you. Thus, readiness for agents becomes as important as a high rank on Google.","keywords":""},{"kind":"section","title":"Agentic commerce readiness","heading":"Why agentic commerce readiness is important","section":"Documentation","crumbs":["Agentic commerce","Agentic commerce readiness","Why agentic commerce readiness is important"],"url":"/docs/agentic-commerce/agentic-commerce-readiness#why-agentic-commerce-readiness-is-important","text":"For most of the history of the web, the customer was a person who read a page. Now, the customer is frequently an agent. The agent needs structured information that a machine can read. This is important for an e-commerce brand for these reasons: - Agents select from what they can read. Maybe your products, prices, and availability are not available in a form that a machine can read. Then the agent has no reliable data to use. It selects competitors that are easier to read. - The selection occurs before the shopper sees it. When a recommendation gets to the user, the agent already made the list shorter. You want to be on that shortlist. - It extends AI recommendations to commerce. The general mission of Genezio is to make sure that AI systems recommend and represent your brand well. Agent readiness extends this mission to the commerce layer. The question is not only if AI talks about you. The question is also if AI can do transactions with you. Agent readiness is now necessary, not optional. More purchases go through assistants. Thus, the stores that agents can use get the sale.","keywords":""},{"kind":"section","title":"Agentic commerce readiness","heading":"What the Universal Commerce Protocol (UCP) is","section":"Documentation","crumbs":["Agentic commerce","Agentic commerce readiness","What the Universal Commerce Protocol (UCP) is"],"url":"/docs/agentic-commerce/agentic-commerce-readiness#what-the-universal-commerce-protocol-ucp-is","text":"The Universal Commerce Protocol (UCP) is the new standard that makes your store available to AI shopping agents. It is a contract between your store and the agents that visit it. It gives a known location for your catalog, your product details, and the endpoints that an agent needs to act. UCP publishes a manifest at a known location on your site (/.well-known/ucp). The manifest shows the agents: - Your product and catalog data, in a structured feed that they can read. - The live endpoints that an agent uses to check details and do actions. - The metadata that tells an agent what your store supports. You do not have to understand each field to get the benefit. For a marketer, the result is important: an agent can recommend and buy with confidence from a store that uses UCP.","keywords":""},{"kind":"section","title":"Agentic commerce readiness","heading":"How Genezio measures agentic commerce readiness","section":"Documentation","crumbs":["Agentic commerce","Agentic commerce readiness","How Genezio measures agentic commerce readiness"],"url":"/docs/agentic-commerce/agentic-commerce-readiness#how-genezio-measures-agentic-commerce-readiness","text":"Genezio helps you answer two questions about agentic commerce: - Is my store ready for agentic commerce? - How do I compare to my competitors? The main feature is the UCP Readiness Audit: - It crawls the UCP configuration of your store. - It gives a score for your readiness. - It does the same check on the competitors that you track in Genezio. Thus, you can compare the agent readiness side by side. - It records how your readiness changes over time. - It runs again automatically after you edit your website. Thus, the result stays current. The dashboard shows the audit on a UCP page. The page has a readiness card that gives a quick summary of your position. For the practical details, refer to the UCP Readiness Audit page. It tells you how to run the audit, what each check means, and how the score works.","keywords":""},{"kind":"section","title":"Agentic commerce readiness","heading":"Related pages","section":"Documentation","crumbs":["Agentic commerce","Agentic commerce readiness","Related pages"],"url":"/docs/agentic-commerce/agentic-commerce-readiness#related-pages","text":"- UCP Readiness Audit: run the audit and read your readiness score - Shopping: Product Visibility: how AI talks about each of your products now - Merchants and Retailers: the retailers that answer engines cite for your category - Competitors: how Genezio tracks the brands that you compare with - Knowledge Base: the source of truth that Genezio makes about your brand","keywords":""},{"kind":"section","title":"Agentic commerce readiness","heading":"In the API","section":"Documentation","crumbs":["Agentic commerce","Agentic commerce readiness","In the API"],"url":"/docs/agentic-commerce/agentic-commerce-readiness#in-the-api","text":"The newest e-commerce audit gives the latest UCP readiness result of your brand.","keywords":""},{"kind":"page","title":"UCP readiness audit","heading":"","section":"Documentation","crumbs":["Documentation","Agentic commerce","UCP readiness audit"],"url":"/docs/agentic-commerce/ucp-readiness-audit","text":"The UCP Readiness Audit checks if AI shopping agents can use your store. Learn what it checks, how the U1 to U7 score works, and how to run it. The UCP Readiness Audit checks if AI shopping agents can use your website. This page explains what the audit checks, how the readiness score works, and how to run the audit. The audit does these steps: - It crawls the Universal Commerce Protocol (UCP) configuration of your store. - It gives a score for how ready your store is. - It does the same check on your tracked competitors. Thus, you can compare the agent readiness side by side. If the concept is new for you, start with Agentic Commerce Readiness to learn why it is important.","keywords":""},{"kind":"section","title":"UCP readiness audit","heading":"Why a UCP readiness audit is important","section":"Documentation","crumbs":["Agentic commerce","UCP readiness audit","Why a UCP readiness audit is important"],"url":"/docs/agentic-commerce/ucp-readiness-audit#why-a-ucp-readiness-audit-is-important","text":"AI shopping agents can recommend and buy only from stores that they can read. The audit changes the question \"are we ready?\" from a guess into a concrete score that you can repeat. It shows you exactly which parts of your UCP configuration cause problems. The audit also gives a score to your competitors. Thus, you can see if agent readiness is an advantage that you can get or a gap that you must close.","keywords":""},{"kind":"section","title":"UCP readiness audit","heading":"What the UCP Readiness Audit checks","section":"Documentation","crumbs":["Agentic commerce","UCP readiness audit","What the UCP Readiness Audit checks"],"url":"/docs/agentic-commerce/ucp-readiness-audit#what-the-ucp-readiness-audit-checks","text":"The audit checks three items. The sequence is the same as the sequence in which an agent finds your store: 1. Manifest completeness. The audit crawls the /.well-known/ucp manifest of your site. Agents use this manifest as the entry point to find your store. The audit checks that the manifest exists. It also checks that the manifest has the information that an agent needs to continue. 2. Catalog feed quality. The audit follows the link from the manifest to your product or catalog feed. It checks if the feed is available. It also checks if the data is sufficiently complete for an agent to compare and select products. 3. Live-endpoint probes. The audit then probes the live endpoints in your configuration. It makes sure that they are available and that they reply. An agent needs endpoints that work to check details and do actions. Together, these checks follow the path of an agent: find the manifest, read the catalog, call the endpoints.","keywords":""},{"kind":"section","title":"UCP readiness audit","heading":"How the UCP readiness score works","section":"Documentation","crumbs":["Agentic commerce","UCP readiness audit","How the UCP readiness score works"],"url":"/docs/agentic-commerce/ucp-readiness-audit#how-the-ucp-readiness-score-works","text":"The audit compares your store with a set of rules, with the labels U1 to U7. These rules are for the three areas above: - Rules for the presence and completeness of the manifest. - Rules for the availability and quality of the catalog feed. - Rules for the availability of the live endpoints. Each rule adds to a general readiness score. The dashboard shows the result on a UCP page, as a readiness card. The card gives a summary of your position. It also highlights the rules that you fail. Thus, you know what to correct first.","keywords":""},{"kind":"section","title":"UCP readiness audit","heading":"How to run the UCP Readiness Audit","section":"Documentation","crumbs":["Agentic commerce","UCP readiness audit","How to run the UCP Readiness Audit"],"url":"/docs/agentic-commerce/ucp-readiness-audit#how-to-run-the-ucp-readiness-audit","text":"Open the UCP page in the Genezio dashboard. Make sure that the website of your brand is correct. The audit crawls the live site. Thus, the URL must point to your real store. You can check or change the URL in Settings - Brand Details. Run the audit. Genezio crawls the manifest, follows the catalog feed, and probes the live endpoints. Read the readiness card for your general score and the results of the rules U1 to U7. You do not have to run the audit manually each time. The audit runs again automatically after you edit your website. Thus, the score shows your latest changes with no more work.","keywords":""},{"kind":"section","title":"UCP readiness audit","heading":"Score history and drift","section":"Documentation","crumbs":["Agentic commerce","UCP readiness audit","Score history and drift"],"url":"/docs/agentic-commerce/ucp-readiness-audit#score-history-and-drift","text":"The audit keeps a score history. Thus, you can see how your readiness changes over time. The history helps you find regressions. For example, a change to the site can break your manifest by accident, or it can stop an endpoint. Then the readiness score decreases, and you see the decrease in the history. Thus, you find the problem before the agents stop their recommendations of your store.","keywords":""},{"kind":"section","title":"UCP readiness audit","heading":"Compare your UCP readiness with competitors","section":"Documentation","crumbs":["Agentic commerce","UCP readiness audit","Compare your UCP readiness with competitors"],"url":"/docs/agentic-commerce/ucp-readiness-audit#compare-your-ucp-readiness-with-competitors","text":"The audit does not give a score only to your brand. It does the same checks on the competitors that you track in Genezio. Thus, you can compare the agent readiness directly. The audit helps you answer these questions: - Are my competitors already ready for UCP, and I am not? - Can I get agent readiness as an advantage before they do? - Which rules do my competitors pass that I fail? For more about how Genezio finds and tracks the brands that you compare with, refer to Competitors.","keywords":""},{"kind":"section","title":"UCP readiness audit","heading":"Related pages","section":"Documentation","crumbs":["Agentic commerce","UCP readiness audit","Related pages"],"url":"/docs/agentic-commerce/ucp-readiness-audit#related-pages","text":"- Agentic Commerce Readiness: the concept, and why agent readiness is important - Competitors: how Genezio tracks the brands that you compare with - Knowledge Base: the source of truth that Genezio makes about your brand","keywords":""},{"kind":"section","title":"UCP readiness audit","heading":"In the API","section":"Documentation","crumbs":["Agentic commerce","UCP readiness audit","In the API"],"url":"/docs/agentic-commerce/ucp-readiness-audit#in-the-api","text":"Start an e-commerce audit runs the audit. The newest e-commerce audit, the history of the audit, and the audit of the competitors give the results.","keywords":""},{"kind":"page","title":"Tutorials","heading":"","section":"Documentation","crumbs":["Documentation","Tutorials","Tutorials"],"url":"/docs/tutorials","text":"End-to-end AI visibility tutorials for ecommerce brands and SaaS companies, and for competitor tracking, content that LLMs cite, and brand reputation. These tutorials give end-to-end sequences. Each tutorial is a quarter of work, not a checklist. Tutorials by type of company: - Improve AI visibility for an ecommerce brand - Improve AI visibility for a SaaS company Tutorials by task: - Track competitors in AI search - Generate content that LLMs cite - Monitor brand reputation in AI Each tutorial tells you which work usually changes the numbers and which work usually does not. In this field, most of the wasted effort goes into work that cannot change the result.","keywords":""},{"kind":"page","title":"Improve AI visibility for an ecommerce brand","heading":"","section":"Documentation","crumbs":["Documentation","Tutorials","Improve AI visibility for an ecommerce brand"],"url":"/docs/tutorials/improve-ai-visibility-for-an-ecommerce-brand","text":"A tutorial for retailers: one brand per market, topics from purchase questions, product visibility, third-party sources, and a Business Scorecard goal. This tutorial gives a sequence of work for a retailer. The retailer wants answer engines to recommend it when a user asks an assistant what to buy.","keywords":""},{"kind":"section","title":"Improve AI visibility for an ecommerce brand","heading":"1. Configure the brand as buyers think of it","section":"Documentation","crumbs":["Tutorials","Improve AI visibility for an ecommerce brand","1. Configure the brand as buyers think of it"],"url":"/docs/tutorials/improve-ai-visibility-for-an-ecommerce-brand#1-configure-the-brand-as-buyers-think-of-it","text":"Create one brand for each market. A shop that sells in two countries has two sets of questions, competitors, and sources. If you merge them, you get one unclear average. Read Create your first brand.","keywords":""},{"kind":"section","title":"Improve AI visibility for an ecommerce brand","heading":"2. Make topics from purchase questions","section":"Documentation","crumbs":["Tutorials","Improve AI visibility for an ecommerce brand","2. Make topics from purchase questions"],"url":"/docs/tutorials/improve-ai-visibility-for-an-ecommerce-brand#2-make-topics-from-purchase-questions","text":"Do not use product categories. Use the questions that people ask before they buy. \"Best running shoes for flat feet\" is a topic. \"Footwear\" is not a topic. Get these questions from what customers ask your support team. If you connected Search Console, also get them from Search Console.","keywords":""},{"kind":"section","title":"Improve AI visibility for an ecommerce brand","heading":"3. Run conversations, then read the answers","section":"Documentation","crumbs":["Tutorials","Improve AI visibility for an ecommerce brand","3. Run conversations, then read the answers"],"url":"/docs/tutorials/improve-ai-visibility-for-an-ecommerce-brand#3-run-conversations-then-read-the-answers","text":"Read the answers before you look at the scores. For retail, examine these three items: - Do answer engines name you? - Do they name specific products, or only the shop? - Which sources do they cite: review sites, YouTube, Reddit, or retailer guides?","keywords":""},{"kind":"section","title":"Improve AI visibility for an ecommerce brand","heading":"4. Work on the product layer","section":"Documentation","crumbs":["Tutorials","Improve AI visibility for an ecommerce brand","4. Work on the product layer"],"url":"/docs/tutorials/improve-ai-visibility-for-an-ecommerce-brand#4-work-on-the-product-layer","text":"Retail has a step that software does not have: individual products occur in answers. Product visibility shows which of your products occur. The readiness work for agentic commerce is important here. Structured and complete product data makes a product quotable.","keywords":""},{"kind":"section","title":"Improve AI visibility for an ecommerce brand","heading":"5. Correct the sources, not only the site","section":"Documentation","crumbs":["Tutorials","Improve AI visibility for an ecommerce brand","5. Correct the sources, not only the site"],"url":"/docs/tutorials/improve-ai-visibility-for-an-ecommerce-brand#5-correct-the-sources-not-only-the-site","text":"Ecommerce answers use third parties very much: reviews, comparison guides, and creator videos. Channels and authors gives the names of the specific YouTube channels and subreddits that have an effect on your category. That is your outreach list.","keywords":""},{"kind":"section","title":"Improve AI visibility for an ecommerce brand","heading":"6. Measure what is important to you","section":"Documentation","crumbs":["Tutorials","Improve AI visibility for an ecommerce brand","6. Measure what is important to you"],"url":"/docs/tutorials/improve-ai-visibility-for-an-ecommerce-brand#6-measure-what-is-important-to-you","text":"\"Be recommended for [your category] on ChatGPT in the UK\" is a Business Scorecard goal. Visibility across all topics is not such a goal.","keywords":""},{"kind":"section","title":"Improve AI visibility for an ecommerce brand","heading":"What usually has an effect","section":"Documentation","crumbs":["Tutorials","Improve AI visibility for an ecommerce brand","What usually has an effect"],"url":"/docs/tutorials/improve-ai-visibility-for-an-ecommerce-brand#what-usually-has-an-effect","text":"In this order: 1. Accurate and complete product data. 2. A presence on the review sites that answer engines already cite. 3. One useful buying guide that answers a cluster of questions. 4. Corrections to third-party pages that describe you incorrectly.","keywords":""},{"kind":"page","title":"Improve AI visibility for a SaaS company","heading":"","section":"Documentation","crumbs":["Documentation","Tutorials","Improve AI visibility for a SaaS company"],"url":"/docs/tutorials/improve-ai-visibility-for-a-saas-company","text":"A tutorial for SaaS teams: use lost comparisons as topics, correct how answer engines describe you, make pages quotable, and plan a 12-week first quarter. This tutorial gives a sequence of work for a software company. For software, the buying question is usually a comparison.","keywords":""},{"kind":"section","title":"Improve AI visibility for a SaaS company","heading":"1. Use the comparisons that you lose as topics","section":"Documentation","crumbs":["Tutorials","Improve AI visibility for a SaaS company","1. Use the comparisons that you lose as topics"],"url":"/docs/tutorials/improve-ai-visibility-for-a-saas-company#1-use-the-comparisons-that-you-lose-as-topics","text":"Your sales team already knows these comparisons. Examples are \"CRM for agencies\", \"alternative to [competitor]\", and \"tool for teams of five to fifty\". Start with ten topics. Read Define topics.","keywords":""},{"kind":"section","title":"Improve AI visibility for a SaaS company","heading":"2. Run conversations and read them before you look at scores","section":"Documentation","crumbs":["Tutorials","Improve AI visibility for a SaaS company","2. Run conversations and read them before you look at scores"],"url":"/docs/tutorials/improve-ai-visibility-for-a-saas-company#2-run-conversations-and-read-them-before-you-look-at-scores","text":"Read approximately twelve answers. For SaaS, find these items: - The competitors that answer engines treat as your real alternatives. Frequently, these are not the competitors that you track commercially. - How answer engines describe you when they name you, for example expensive, complex, or for enterprises. - If the answers cite comparison sites, documentation, or community threads.","keywords":""},{"kind":"section","title":"Improve AI visibility for a SaaS company","heading":"3. Correct the description before you add coverage","section":"Documentation","crumbs":["Tutorials","Improve AI visibility for a SaaS company","3. Correct the description before you add coverage"],"url":"/docs/tutorials/improve-ai-visibility-for-a-saas-company#3-correct-the-description-before-you-add-coverage","text":"When answer engines name a brand in most answers but recommend it in few answers, the brand has a positioning problem. More content cannot correct this problem. Use Monitor perceptions to track the specific claim, for example \"hard to set up\" or \"expensive\". Then, you can see if your work against the claim has an effect.","keywords":""},{"kind":"section","title":"Improve AI visibility for a SaaS company","heading":"4. Make your existing pages quotable","section":"Documentation","crumbs":["Tutorials","Improve AI visibility for a SaaS company","4. Make your existing pages quotable"],"url":"/docs/tutorials/improve-ai-visibility-for-a-saas-company#4-make-your-existing-pages-quotable","text":"SaaS sites usually have the information, but they hide it in marketing text. Apply the test in Structuring content for LLMs to your pricing, integration, and comparison pages. This is the work with the lowest cost.","keywords":""},{"kind":"section","title":"Improve AI visibility for a SaaS company","heading":"5. Get independent coverage","section":"Documentation","crumbs":["Tutorials","Improve AI visibility for a SaaS company","5. Get independent coverage"],"url":"/docs/tutorials/improve-ai-visibility-for-a-saas-company#5-get-independent-coverage","text":"Answer engines give more weight to third parties. For software, the third parties are review platforms, comparison sites, and the communities where people discuss your category. Read Building authority.","keywords":""},{"kind":"section","title":"Improve AI visibility for a SaaS company","heading":"6. Close gaps with real content","section":"Documentation","crumbs":["Tutorials","Improve AI visibility for a SaaS company","6. Close gaps with real content"],"url":"/docs/tutorials/improve-ai-visibility-for-a-saas-company#6-close-gaps-with-real-content","text":"Use Content opportunities. Write one substantial piece for each cluster, not one page for each question.","keywords":""},{"kind":"section","title":"Improve AI visibility for a SaaS company","heading":"A reasonable first quarter","section":"Documentation","crumbs":["Tutorials","Improve AI visibility for a SaaS company","A reasonable first quarter"],"url":"/docs/tutorials/improve-ai-visibility-for-a-saas-company#a-reasonable-first-quarter","text":"1–2: Configure Genezio, run conversations, and read the answers. 3–4: Make your top ten existing pages quotable. 5–8: Correct how answer engines describe you, where the description is incorrect. 9–12: Write one or two substantial pieces for the largest gaps. Then, measure.","keywords":""},{"kind":"page","title":"Track competitors in AI search","heading":"","section":"Documentation","crumbs":["Documentation","Tutorials","Track competitors in AI search"],"url":"/docs/tutorials/track-competitors-in-ai-search","text":"A tutorial to track competitors in AI answers: let answers name them, clean the list, read visibility with recommendation, and find the sources behind each win. This tutorial shows how to use Genezio to monitor a competitor correctly and calmly.","keywords":""},{"kind":"section","title":"Track competitors in AI search","heading":"1. Let the answers name the competitors","section":"Documentation","crumbs":["Tutorials","Track competitors in AI search","1. Let the answers name the competitors"],"url":"/docs/tutorials/track-competitors-in-ai-search#1-let-the-answers-name-the-competitors","text":"Genezio extracts competitors from the answers. Thus, start with the brands that answer engines actually put next to you. Read Extracting competitors. Expect a surprise. The brands that an answer engine treats as alternatives are not always the brands that win deals against you. Also, a brand that wins in AI answers today usually occurs in your sales pipeline later.","keywords":""},{"kind":"section","title":"Track competitors in AI search","heading":"2. Clean the list","section":"Documentation","crumbs":["Tutorials","Track competitors in AI search","2. Clean the list"],"url":"/docs/tutorials/track-competitors-in-ai-search#2-clean-the-list","text":"Remove adjacent categories, parent companies, and brands with the same name. They distort Share of voice, which divides the conversation among all the listed brands.","keywords":""},{"kind":"section","title":"Track competitors in AI search","heading":"3. Read the grid, not the ranking","section":"Documentation","crumbs":["Tutorials","Track competitors in AI search","3. Read the grid, not the ranking"],"url":"/docs/tutorials/track-competitors-in-ai-search#3-read-the-grid-not-the-ranking","text":"In Competitor comparison, the useful signal is the combination of the two scores: - High visibility, low recommendation: answer engines know the brand, but they do not prefer it. - Low visibility, high recommendation: dangerous. When answer engines show this brand, they recommend it convincingly. - High on both: the brand to beat.","keywords":""},{"kind":"section","title":"Track competitors in AI search","heading":"4. Find where competitors win and you are absent","section":"Documentation","crumbs":["Tutorials","Track competitors in AI search","4. Find where competitors win and you are absent"],"url":"/docs/tutorials/track-competitors-in-ai-search#4-find-where-competitors-win-and-you-are-absent","text":"Find the topics where a competitor is visible and you are not. Each topic is a question that your buyers ask, and you have no presence in it. This is the most actionable list that you will get.","keywords":""},{"kind":"section","title":"Track competitors in AI search","heading":"5. Find why they win","section":"Documentation","crumbs":["Tutorials","Track competitors in AI search","5. Find why they win"],"url":"/docs/tutorials/track-competitors-in-ai-search#5-find-why-they-win","text":"Open the conversations and read the cited sources. Frequently, one or two sources support a competitor that answer engines recommend in thirty answers. That is a specific item that you can respond to, not a general disadvantage.","keywords":""},{"kind":"section","title":"Track competitors in AI search","heading":"6. Monitor changes, not levels","section":"Documentation","crumbs":["Tutorials","Track competitors in AI search","6. Monitor changes, not levels"],"url":"/docs/tutorials/track-competitors-in-ai-search#6-monitor-changes-not-levels","text":"The direction of the score of a competitor is more important than its absolute value. If you examine the comparison regularly, set it as a Business Scorecard goal.","keywords":""},{"kind":"section","title":"Track competitors in AI search","heading":"What not to do","section":"Documentation","crumbs":["Tutorials","Track competitors in AI search","What not to do"],"url":"/docs/tutorials/track-competitors-in-ai-search#what-not-to-do","text":"Do not respond to each move of a competitor. Answers change from run to run, and one change is usually noise. Act only on a continuous direction that has a visible cause in the citations.","keywords":""},{"kind":"page","title":"Generate content that LLMs cite","heading":"","section":"Documentation","crumbs":["Documentation","Tutorials","Generate content that LLMs cite"],"url":"/docs/tutorials/generate-content-that-llms-cite","text":"A six-step tutorial: start from a real content gap, brief from query fanouts, edit each paragraph to be quotable, publish for crawlers, and measure citations. This tutorial goes from end to end: from a gap in the data to a published piece that answer engines quote.","keywords":""},{"kind":"section","title":"Generate content that LLMs cite","heading":"1. Start from a real gap","section":"Documentation","crumbs":["Tutorials","Generate content that LLMs cite","1. Start from a real gap"],"url":"/docs/tutorials/generate-content-that-llms-cite#1-start-from-a-real-gap","text":"Use Content opportunities. A content opportunity is a question that users ask, where a competitor is present and you are absent. Apply an honest filter: can you credibly be the answer? If the question is about a capability that you do not have, visibility cannot help.","keywords":""},{"kind":"section","title":"Generate content that LLMs cite","heading":"2. Use the query fanouts, not the title","section":"Documentation","crumbs":["Tutorials","Generate content that LLMs cite","2. Use the query fanouts, not the title"],"url":"/docs/tutorials/generate-content-that-llms-cite#2-use-the-query-fanouts-not-the-title","text":"Open the query fanouts for the topic. They show what retrieval actually looked for. They tell you the sub-questions that the piece must answer and the words to use. Group the query fanouts by intent. One group is one piece.","keywords":""},{"kind":"section","title":"Generate content that LLMs cite","heading":"3. Write a good brief","section":"Documentation","crumbs":["Tutorials","Generate content that LLMs cite","3. Write a good brief"],"url":"/docs/tutorials/generate-content-that-llms-cite#3-write-a-good-brief","text":"The brief holds the analysis: the questions, the audience, and the sources. A brief from query fanouts gives the writer a list of questions, not only a subject.","keywords":""},{"kind":"section","title":"Generate content that LLMs cite","heading":"4. Generate the article, then edit it as an editor","section":"Documentation","crumbs":["Tutorials","Generate content that LLMs cite","4. Generate the article, then edit it as an editor"],"url":"/docs/tutorials/generate-content-that-llms-cite#4-generate-the-article-then-edit-it-as-an-editor","text":"Generate the article. Then, apply one test to each paragraph: does the paragraph answer a question by itself, without the paragraph before it? If it does not, the paragraph is not quotable, also when it reads well. Read Structuring content for LLMs.","keywords":""},{"kind":"section","title":"Generate content that LLMs cite","heading":"5. Publish so that answer engines can read it","section":"Documentation","crumbs":["Tutorials","Generate content that LLMs cite","5. Publish so that answer engines can read it"],"url":"/docs/tutorials/generate-content-that-llms-cite#5-publish-so-that-answer-engines-can-read-it","text":"- Do not put the content behind a gate. - Do not publish it only in a PDF or a video. - Publish it on a URL that crawlers can read. - Include the schema markup. Read Publishing content.","keywords":""},{"kind":"section","title":"Generate content that LLMs cite","heading":"6. Measure with patience","section":"Documentation","crumbs":["Tutorials","Generate content that LLMs cite","6. Measure with patience"],"url":"/docs/tutorials/generate-content-that-llms-cite#6-measure-with-patience","text":"Record the date, and continue to run the topic. Then, find if the page occurs in Most cited sources. Citations come some weeks after publication. One run is not a verdict. If the page is still absent after many runs, read the sources that answer engines cite for that question. Then, compare them honestly with your page.","keywords":""},{"kind":"page","title":"Monitor brand reputation in AI","heading":"","section":"Documentation","crumbs":["Documentation","Tutorials","Monitor brand reputation in AI"],"url":"/docs/tutorials/monitor-brand-reputation-in-ai","text":"A tutorial to monitor what answer engines say about your brand: sort claims into three types, track perceptions, and trace each false claim to its source. In this context, reputation is not general sentiment. It is the specific claims that answer engines make about you, and if these claims are true.","keywords":""},{"kind":"section","title":"Monitor brand reputation in AI","heading":"1. Find what answer engines say","section":"Documentation","crumbs":["Tutorials","Monitor brand reputation in AI","1. Find what answer engines say"],"url":"/docs/tutorials/monitor-brand-reputation-in-ai#1-find-what-answer-engines-say","text":"Start with AI perception summary: the claims that answer engines make about you. Expect a mix of positive, neutral, and incorrect claims.","keywords":""},{"kind":"section","title":"Monitor brand reputation in AI","heading":"2. Separate the three types","section":"Documentation","crumbs":["Tutorials","Monitor brand reputation in AI","2. Separate the three types"],"url":"/docs/tutorials/monitor-brand-reputation-in-ai#2-separate-the-three-types","text":"- True but not helpful: for example, \"enterprise-focused\" when you want mid-market customers. This is a positioning problem. - False: for example, a missing feature that you actually have, or an old price. This is a source problem: a cited source is incorrect. - True and damaging: an honest weakness. Content cannot correct this type. Possibly, the product can. It is important to know the type, because the correct response is completely different for each type.","keywords":""},{"kind":"section","title":"Monitor brand reputation in AI","heading":"3. Monitor the important claims","section":"Documentation","crumbs":["Tutorials","Monitor brand reputation in AI","3. Monitor the important claims"],"url":"/docs/tutorials/monitor-brand-reputation-in-ai#3-monitor-the-important-claims","text":"Use Monitor perceptions for the claims that are important. A tracked perception keeps a history. Thus, you can prove that a change had an effect. Also, you can find a reversal in weeks, not at the next review.","keywords":""},{"kind":"section","title":"Monitor brand reputation in AI","heading":"4. Find the source of a false claim","section":"Documentation","crumbs":["Tutorials","Monitor brand reputation in AI","4. Find the source of a false claim"],"url":"/docs/tutorials/monitor-brand-reputation-in-ai#4-find-the-source-of-a-false-claim","text":"Open the conversations that contain the claim, and read the cited sources. Almost always, a false claim comes from a cited source that contains an error. The fix is to correct that page. Do not publish a rebuttal on your own site, because answer engines give it less weight.","keywords":""},{"kind":"section","title":"Monitor brand reputation in AI","heading":"5. Use sentiment as an early warning","section":"Documentation","crumbs":["Tutorials","Monitor brand reputation in AI","5. Use sentiment as an early warning"],"url":"/docs/tutorials/monitor-brand-reputation-in-ai#5-use-sentiment-as-an-early-warning","text":"Sentiment tells you the general mood, not the claim. A decrease usually means that a new source has an effect on the answers. Find that source in Most cited sources.","keywords":""},{"kind":"section","title":"Monitor brand reputation in AI","heading":"6. Make it a routine","section":"Documentation","crumbs":["Tutorials","Monitor brand reputation in AI","6. Make it a routine"],"url":"/docs/tutorials/monitor-brand-reputation-in-ai#6-make-it-a-routine","text":"- Run conversations weekly. - Track a small number of perceptions. - Record what you changed, and when. Damage to your reputation in AI answers grows quietly. Nobody gets a notification when an answer engine starts to describe you differently.","keywords":""},{"kind":"page","title":"FAQ","heading":"","section":"Documentation","crumbs":["Documentation","FAQ","FAQ"],"url":"/docs/faq","text":"Answers to the questions that teams ask in their first month with Genezio: run frequency, changes in LLM results, missing citations, accuracy and competitors. These pages answer the questions that teams ask in their first month. Read these two pages before you make a conclusion from your data: - How often should I run conversations? - Why do LLM results change? Why is my brand not cited? gives the four usual causes, in the order of probability. Read How accurate are AI visibility scores? before you show a number to leadership. How does Genezio detect competitors? explains why an unknown brand can appear on your list.","keywords":""},{"kind":"page","title":"How often should I run conversations?","heading":"","section":"Documentation","crumbs":["Documentation","FAQ","How often should I run conversations?"],"url":"/docs/faq/how-often-should-i-run-conversations","text":"For most brands, weekly runs are best. Learn when to run conversations daily, why monthly is too slow, and how to use your plan budget for breadth or depth. For most brands, weekly runs are correct. Weekly runs show a real change in some days. They are also not so frequent that the noise is larger than the signal.","keywords":""},{"kind":"section","title":"How often should I run conversations?","heading":"Why not daily runs","section":"Documentation","crumbs":["FAQ","How often should I run conversations?","Why not daily runs"],"url":"/docs/faq/how-often-should-i-run-conversations#why-not-daily-runs","text":"Answer engines give different results from run to run. Most changes from day to day come from the medium, not from your market. When you monitor daily, you learn to react to changes that have no meaning. Daily runs are useful in two conditions: - In the week after a large launch. - When you test whether a change had an effect.","keywords":""},{"kind":"section","title":"How often should I run conversations?","heading":"Why not monthly runs","section":"Documentation","crumbs":["FAQ","How often should I run conversations?","Why not monthly runs"],"url":"/docs/faq/how-often-should-i-run-conversations#why-not-monthly-runs","text":"In one month, a competitor can publish content, get citations, and take a position before you see it. When you see the change, it is difficult to find its cause.","keywords":""},{"kind":"section","title":"How often should I run conversations?","heading":"The real limit","section":"Documentation","crumbs":["FAQ","How often should I run conversations?","The real limit"],"url":"/docs/faq/how-often-should-i-run-conversations#the-real-limit","text":"Your plan sets the number of conversations that you can run. Thus, the question is usually not \"How often?\" but \"How do I use the budget?\": - More topics, less frequently. This gives breadth. It is better when you make a map of a category. - Fewer topics, more frequently. This gives a reliable trend on the important questions. It is better when you work on a specific problem. Start with breadth while you learn the structure of your category. Then make the set smaller.","keywords":""},{"kind":"section","title":"How often should I run conversations?","heading":"A practical pattern","section":"Documentation","crumbs":["FAQ","How often should I run conversations?","A practical pattern"],"url":"/docs/faq/how-often-should-i-run-conversations#a-practical-pattern","text":"Run conversations weekly across your monitored topics. Add a run after each event that you expect to change the numbers. Compare runs of the same type. A run after a publication has meaning only when you compare it with a run before the publication.","keywords":""},{"kind":"page","title":"Why do LLM results change?","heading":"","section":"Documentation","crumbs":["Documentation","FAQ","Why do LLM results change?"],"url":"/docs/faq/why-do-llm-results-change","text":"Answer engines are not deterministic, so the same question gets different answers. Learn the causes and how to measure AI visibility when the results change. LLM results change because answer engines are not deterministic. When you ask the same question two times, you get different wording. Sometimes you also get different brands. This is a property of the medium. It is not a fault in the measurement.","keywords":""},{"kind":"section","title":"Why do LLM results change?","heading":"What causes the changes","section":"Documentation","crumbs":["FAQ","Why do LLM results change?","What causes the changes"],"url":"/docs/faq/why-do-llm-results-change#what-causes-the-changes","text":"Sampling. Models make text by probability. Two runs of the same prompt can follow different paths. Retrieval. Some answer engines search before they answer. They can find different sources on different days, because the web changed or the retriever ranked the results in a different order. Changes to the answer engines. The providers update models, tune prompts, and rebuild retrieval. They do not announce these changes, and the changes can move a full category at the same time.","keywords":""},{"kind":"section","title":"Why do LLM results change?","heading":"How to measure in a medium that changes","section":"Documentation","crumbs":["FAQ","Why do LLM results change?","How to measure in a medium that changes"],"url":"/docs/faq/why-do-llm-results-change#how-to-measure-in-a-medium-that-changes","text":"Read trends, not single readings. A single run is an estimate. A direction over several runs is a finding. For more information, see Trend tracking. Do not change the questions. If you change scenarios between runs, you measure two different things. Keep a stable set, and change it only when you decide to. Find out if your competitors also changed. If all brands changed at the same time, the answer engine changed, not your brand. This one check prevents most unnecessary investigations. Run sufficient conversations. More conversations for each topic give less noise in each reading.","keywords":""},{"kind":"section","title":"Why do LLM results change?","heading":"When a change is real","section":"Documentation","crumbs":["FAQ","Why do LLM results change?","When a change is real"],"url":"/docs/faq/why-do-llm-results-change#when-a-change-is-real","text":"A change is real when these three conditions are true: - The change continues over several runs. - The change applies to specific topics, not to all topics. - The citations explain the change. For example, a source appeared or a source disappeared.","keywords":""},{"kind":"page","title":"Why is my brand not cited?","heading":"","section":"Documentation","crumbs":["Documentation","FAQ","Why is my brand not cited?"],"url":"/docs/faq/why-is-my-brand-not-cited","text":"Four usual reasons why answer engines do not cite your brand, in order of probability, with a check for each reason and what to change on your pages. There are almost always four possible reasons. This page gives them in the order of probability.","keywords":""},{"kind":"section","title":"Why is my brand not cited?","heading":"1. There is nothing to cite","section":"Documentation","crumbs":["FAQ","Why is my brand not cited?","1. There is nothing to cite"],"url":"/docs/faq/why-is-my-brand-not-cited#1-there-is-nothing-to-cite","text":"Answer engines cite sources that answer the question. Your site possibly positions your brand and tries to persuade. If it never answers the specific question, the answer engine has nothing to use. Check: Read the conversations for a topic where your brand is absent. Look at the sources that the answer engine did cite. Compare them with your equivalent page. Usually, their page answers the question and your page sells a product.","keywords":""},{"kind":"section","title":"Why is my brand not cited?","heading":"2. All content about your brand comes from your brand","section":"Documentation","crumbs":["FAQ","Why is my brand not cited?","2. All content about your brand comes from your brand"],"url":"/docs/faq/why-is-my-brand-not-cited#2-all-content-about-your-brand-comes-from-your-brand","text":"Answer engines give more weight to independent sources, because a description of a brand by the brand itself is weak evidence. A citation profile of only your own pages is fragile and easy to replace. Check: Look at Most cited sources. If your brand appears only on your own domain, that is the finding.","keywords":""},{"kind":"section","title":"Why is my brand not cited?","heading":"3. An answer engine cannot quote your page","section":"Documentation","crumbs":["FAQ","Why is my brand not cited?","3. An answer engine cannot quote your page"],"url":"/docs/faq/why-is-my-brand-not-cited#3-an-answer-engine-cannot-quote-your-page","text":"Answer engines quote passages. On some pages, the important content is divided across marketing text. On other pages, it is in an image, a PDF, or a video with no transcript. Such a page gives no quotable text, even when the information is on the page. Check: Find one paragraph that answers the question without other context. If you cannot copy such a paragraph, an answer engine also cannot.","keywords":""},{"kind":"section","title":"Why is my brand not cited?","heading":"4. You answer a different question","section":"Documentation","crumbs":["FAQ","Why is my brand not cited?","4. You answer a different question"],"url":"/docs/faq/why-is-my-brand-not-cited#4-you-answer-a-different-question","text":"The question that you optimized for and the question that buyers ask are frequently different. Query fanouts show what the retrieval actually searches for.","keywords":""},{"kind":"section","title":"Why is my brand not cited?","heading":"What not to conclude","section":"Documentation","crumbs":["FAQ","Why is my brand not cited?","What not to conclude"],"url":"/docs/faq/why-is-my-brand-not-cited#what-not-to-conclude","text":"No citation is not a penalty, and you cannot appeal it. Answer engines do not keep your brand out on purpose. They found better material for that question. Your work is to make the better material.","keywords":""},{"kind":"page","title":"How does Genezio detect competitors?","heading":"","section":"Documentation","crumbs":["Documentation","FAQ","How does Genezio detect competitors?"],"url":"/docs/faq/how-does-genezio-detect-competitors","text":"Genezio finds competitors in the AI answers: the brands that answer engines name as alternatives to yours. Learn why unknown brands appear and what to remove. Genezio detects competitors from the answers. It reads each conversation to find the brands that the answer names as alternatives to your brand. Then it collects these brands across the runs. For information about the mechanism, see Extracting competitors.","keywords":""},{"kind":"section","title":"How does Genezio detect competitors?","heading":"Why does Genezio not use only a list?","section":"Documentation","crumbs":["FAQ","How does Genezio detect competitors?","Why does Genezio not use only a list?"],"url":"/docs/faq/how-does-genezio-detect-competitors#why-does-genezio-not-use-only-a-list","text":"You can use a list. Genezio monitors each competitor that you add. But a list alone does not show the most useful part. The brands that an answer engine shows next to your brand are not always your commercial competitors. Answer engines show brands according to how well sources support them. Thus, the list frequently contains a name that your sales team does not know. This is a finding, not an error. A brand that appears in AI answers today usually appears in your sales deals later.","keywords":""},{"kind":"section","title":"How does Genezio detect competitors?","heading":"Why does a brand that is not a competitor appear?","section":"Documentation","crumbs":["FAQ","How does Genezio detect competitors?","Why does a brand that is not a competitor appear?"],"url":"/docs/faq/how-does-genezio-detect-competitors#why-does-a-brand-that-is-not-a-competitor-appear","text":"There are three usual causes: - Adjacent categories. An answer engine that gives a general answer can name tools for a related problem. - Parent or sibling brands. The answer names a product together with its parent company. - Name collisions. The name of a brand is also a common word. Thus, Genezio counts it too many times. Remove these brands. They make Share of Voice incorrect, because Share of Voice divides the conversation among all brands on the list.","keywords":""},{"kind":"section","title":"How does Genezio detect competitors?","heading":"How often must I examine the list?","section":"Documentation","crumbs":["FAQ","How does Genezio detect competitors?","How often must I examine the list?"],"url":"/docs/faq/how-does-genezio-detect-competitors#how-often-must-i-examine-the-list","text":"For most categories, one time each quarter is sufficient. The set of brands that answer engines show as alternatives changes slowly. A list that you do not change after the setup slowly becomes an incorrect description of your real market.","keywords":""},{"kind":"page","title":"How accurate are AI visibility scores?","heading":"","section":"Documentation","crumbs":["Documentation","FAQ","How accurate are AI visibility scores?"],"url":"/docs/faq/how-accurate-are-ai-visibility-scores","text":"AI visibility scores are accurate for your questions, answer engines and period. Learn what the score depends on and how to compare it in a correct way. The score is accurate as a measurement of the questions that you asked, on the answer engines that you selected, in the period of the runs. It is not accurate as a universal statement about your brand. No number of that type exists.","keywords":""},{"kind":"section","title":"How accurate are AI visibility scores?","heading":"What the score represents","section":"Documentation","crumbs":["FAQ","How accurate are AI visibility scores?","What the score represents"],"url":"/docs/faq/how-accurate-are-ai-visibility-scores#what-the-score-represents","text":"The score comes from real answers, from real answer engines, to questions that you defined. Genezio does not model or estimate any part of it. Genezio makes each data point from conversations that you can open and read. This is the strength of the score. When a number surprises you, you can examine the answers that it comes from.","keywords":""},{"kind":"section","title":"How accurate are AI visibility scores?","heading":"What the score depends on","section":"Documentation","crumbs":["FAQ","How accurate are AI visibility scores?","What the score depends on"],"url":"/docs/faq/how-accurate-are-ai-visibility-scores#what-the-score-depends-on","text":"Your set of questions. When you change the scenarios, the number changes. This behavior is correct, because you measure a different thing. But it also means that you cannot compare the score across brands that have different topics. Your answer engines. Visibility changes much from one answer engine to a different one. A combined number across four answer engines is an average of four different positions. Sampling. Answer engines give different results from run to run. Thus, each single reading has noise. For more information, see Why do LLM results change?.","keywords":""},{"kind":"section","title":"How accurate are AI visibility scores?","heading":"How to use the score correctly","section":"Documentation","crumbs":["FAQ","How accurate are AI visibility scores?","How to use the score correctly"],"url":"/docs/faq/how-accurate-are-ai-visibility-scores#how-to-use-the-score-correctly","text":"Compare the score with your own past scores. This is the most reliable use. Use the same questions and the same answer engines in different weeks. Compare the score with competitors on the same set of questions. This comparison is valid, because Genezio measures each brand on the same questions. Do not compare the score with an industry benchmark. In a crowded category, 40% can be strong. In a niche category, 40% can be weak. A number from the topic set of a different brand is not a benchmark for your brand.","keywords":""},{"kind":"section","title":"How accurate are AI visibility scores?","heading":"Summary","section":"Documentation","crumbs":["FAQ","How accurate are AI visibility scores?","Summary"],"url":"/docs/faq/how-accurate-are-ai-visibility-scores#summary","text":"Use the score as an instrument reading, not as a fact. Its direction is reliable, and its precision is a few points. You can always explain it when you open the conversations that it comes from.","keywords":""},{"kind":"page","title":"Integrations","heading":"","section":"Documentation","crumbs":["Documentation","Integrations","Integrations"],"url":"/docs/integrations","text":"Connect Genezio to Google Search Console, Google Analytics 4, and your CDN logs, and send your AI visibility data to exports, BI tools, and the API. The integrations connect Genezio with the other tools of your stack.","keywords":""},{"kind":"section","title":"Integrations","heading":"Bring data into Genezio","section":"Documentation","crumbs":["Integrations","Integrations","Bring data into Genezio"],"url":"/docs/integrations#bring-data-into-genezio","text":"- Google Search Console is the most useful connection for most teams. Its queries show real demand, and Genezio can make topics from them. - Google Analytics adds the behavior of visitors after the click. This integration has one important limit, and its page explains that limit clearly. - CDN log integration shows which crawlers of answer engines come to your site.","keywords":""},{"kind":"section","title":"Integrations","heading":"Get data out of Genezio","section":"Documentation","crumbs":["Integrations","Integrations","Get data out of Genezio"],"url":"/docs/integrations#get-data-out-of-genezio","text":"- Data exports give you PDF and CSV files for a task that you do one time. - External reporting tools are for a task that you repeat. For these tasks, use the API.","keywords":""},{"kind":"page","title":"Google Search Console","heading":"","section":"Documentation","crumbs":["Documentation","Integrations","Google Search Console"],"url":"/docs/integrations/google-search-console","text":"Connect Google Search Console to Genezio, generate topics from the real search queries of your site, and compare search demand with AI questions. When you connect Search Console, Genezio can use what persons really searched for on your site, together with what they ask answer engines.","keywords":""},{"kind":"section","title":"Google Search Console","heading":"What you get","section":"Documentation","crumbs":["Integrations","Google Search Console","What you get"],"url":"/docs/integrations/google-search-console#what-you-get","text":"Topics from real demand. Genezio can read your Search Console queries and generate topics from them. You do not guess what to track. You start from the phrases that already bring persons to your site. This is important, because the difference between the two data sets gives the most useful information. A query that brings you traffic from search but never appears in AI answers shows exactly where you lose position. Your search data next to your AI data. The queries are next to the questions that persons ask answer engines. Thus, you can see where the two data sets agree and where they are different.","keywords":""},{"kind":"section","title":"Google Search Console","heading":"Connect Search Console","section":"Documentation","crumbs":["Integrations","Google Search Console","Connect Search Console"],"url":"/docs/integrations/google-search-console#connect-search-console","text":"1. Open Settings Integrations for the brand. 2. Connect the Google account that has access to the property. 3. Select the property for this brand. A domain property reads the full site. A URL-prefix property reads only the addresses below that prefix. Select the property that agrees with the brand.","keywords":""},{"kind":"section","title":"Google Search Console","heading":"Generate topics from queries","section":"Documentation","crumbs":["Integrations","Google Search Console","Generate topics from queries"],"url":"/docs/integrations/google-search-console#generate-topics-from-queries","text":"After you connect Search Console, you can generate topics from the queries that Search Console reports. By default, Genezio filters the queries and keeps only the queries that are worth an action. You do not get each long-tail string that your site matched in the past. Examine the proposed topics before you accept them. Search queries and AI questions do not have the same shape. \"crm pricing\" is a search. \"what does a CRM cost for a team of five\" is what a person asks an assistant. The generated topics connect the two shapes, and a quick review makes them better.","keywords":""},{"kind":"section","title":"Google Search Console","heading":"Google Analytics 4","section":"Documentation","crumbs":["Integrations","Google Search Console","Google Analytics 4"],"url":"/docs/integrations/google-search-console#google-analytics-4","text":"You can connect Google Analytics 4 from the same location. It shows the behavior of a website, not what persons ask. Thus, it adds to the data, but it does not make topics. Refer to Google Analytics integration.","keywords":""},{"kind":"page","title":"Google Analytics integration","heading":"","section":"Documentation","crumbs":["Documentation","Integrations","Google Analytics integration"],"url":"/docs/integrations/google-analytics-integration","text":"Connect Google Analytics 4 to Genezio to compare what answer engines say about your brand with what visitors do on your site, and learn the attribution limit. When you connect Google Analytics 4, Genezio can show what answer engines say about your brand next to what visitors do when they come to your site.","keywords":""},{"kind":"section","title":"Google Analytics integration","heading":"What the integration adds","section":"Documentation","crumbs":["Integrations","Google Analytics integration","What the integration adds"],"url":"/docs/integrations/google-analytics-integration#what-the-integration-adds","text":"Analytics shows the behavior on your site: which pages persons go to, and what they do next. Genezio shows what occurs before that: if your brand was in the answer. Together, they close a gap that neither tool can close alone. A page that AI answers cite frequently but that gets no conversions is a different problem from a page that gets no citations.","keywords":""},{"kind":"section","title":"Google Analytics integration","heading":"Connect Google Analytics","section":"Documentation","crumbs":["Integrations","Google Analytics integration","Connect Google Analytics"],"url":"/docs/integrations/google-analytics-integration#connect-google-analytics","text":"1. Open Settings Integrations for the brand. 2. Connect the Google account that has access to the property. 3. Select the property for this brand.","keywords":""},{"kind":"section","title":"Google Analytics integration","heading":"What the integration does not do","section":"Documentation","crumbs":["Integrations","Google Analytics integration","What the integration does not do"],"url":"/docs/integrations/google-analytics-integration#what-the-integration-does-not-do","text":"Analytics cannot tell you why a person came to your site. Most answer engines send little or no referrer information. Thus, traffic that started as an AI recommendation frequently looks like direct traffic. Make this clear in your team: you cannot attribute AI visibility through Analytics. A team that expects this attribution can think that the work does not give results. In fact, the data is not available. Use the two tools as measures that complete each other, not as one funnel.","keywords":""},{"kind":"section","title":"Google Analytics integration","heading":"The more useful connection","section":"Documentation","crumbs":["Integrations","Google Analytics integration","The more useful connection"],"url":"/docs/integrations/google-analytics-integration#the-more-useful-connection","text":"For most teams, Google Search Console is the stronger integration. Its queries show demand: what persons really ask. Genezio can change these queries into topics. Analytics shows the behavior only after the visit.","keywords":""},{"kind":"page","title":"CDN log integration","heading":"","section":"Documentation","crumbs":["Documentation","Integrations","CDN log integration"],"url":"/docs/integrations/cdn-log-integration","text":"The CDN log analysis of Genezio uses your server access logs to show which pages AI crawlers fetch, and if they agree with the pages that AI answers cite. The CDN log analysis uses your own server-side access logs to track AI crawler activity. It shows how answer engines really use your website: - Which pages they fetch. - When they fetch them. - If these pages agree with the pages that they cite. This analysis adds first-party evidence from your own infrastructure to the citation data of Genezio. Genezio does the log ingestion on request, for each customer. Refer to Request the log ingestion.","keywords":""},{"kind":"section","title":"CDN log integration","heading":"Why CDN logs show how answer engines use your site","section":"Documentation","crumbs":["Integrations","CDN log integration","Why CDN logs show how answer engines use your site"],"url":"/docs/integrations/cdn-log-integration#why-cdn-logs-show-how-answer-engines-use-your-site","text":"The citation tracking tells you which of your pages appear in the answers of answer engines. Your logs tell you the other half: what answer engines do on your site first. The logs show two different behaviors. The main purpose of the analysis is to keep these two behaviors separate.","keywords":""},{"kind":"section","title":"CDN log integration","heading":"1. Training crawls","section":"Documentation","crumbs":["Integrations","CDN log integration","1. Training crawls"],"url":"/docs/integrations/cdn-log-integration#1-training-crawls","text":"Answer engines crawl the web to make and update the models behind their answers. Your logs show which of your pages they visit for training. Thus, you see which parts of your site the model knows and which parts it ignored.","keywords":""},{"kind":"section","title":"CDN log integration","heading":"2. Real-time answer enrichment","section":"Documentation","crumbs":["Integrations","CDN log integration","2. Real-time answer enrichment"],"url":"/docs/integrations/cdn-log-integration#2-real-time-answer-enrichment","text":"This signal has more value. Answer engines also fetch pages in real time, while they answer a specific question. They use these pages to add current data to the answer. When a page has this type of access, an answer engine adds your content to a live answer at that time. Thus, you see which pages are important for the answers of answer engines now. This data is very different from \"a crawler visited this page at some time\", and you can act on it more easily.","keywords":""},{"kind":"section","title":"CDN log integration","heading":"Questions that the analysis answers","section":"Documentation","crumbs":["Integrations","CDN log integration","Questions that the analysis answers"],"url":"/docs/integrations/cdn-log-integration#questions-that-the-analysis-answers","text":"- Do answer engines crawl my most important content, or do they ignore it? - Which pages do answer engines fetch in real time to support live answers? - Which high-value pages do AI bots never visit? - Which pages do crawlers visit frequently, but answer engines never cite? The analysis uses your own server logs. Thus, it is direct, first-party evidence, not an estimate.","keywords":""},{"kind":"section","title":"CDN log integration","heading":"Use the CDN logs with GA4 data","section":"Documentation","crumbs":["Integrations","CDN log integration","Use the CDN logs with GA4 data"],"url":"/docs/integrations/cdn-log-integration#use-the-cdn-logs-with-ga4-data","text":"The log data is most useful when you read it together with your Google Analytics (GA4) data. Your CDN logs show what the answer engines did: the crawls and the real-time fetches. GA4 shows what persons did. When you compare the two, you get a much fuller view of these items: - How answer engines use the content that you publish - What the persons who come to your site from these answers do","keywords":""},{"kind":"section","title":"CDN log integration","heading":"What log data Genezio ingests","section":"Documentation","crumbs":["Integrations","CDN log integration","What log data Genezio ingests"],"url":"/docs/integrations/cdn-log-integration#what-log-data-genezio-ingests","text":"Genezio ingests your CDN or web server access logs. These are the standard logs that your CDN or your origin server already records for each request: - The URLs and the paths of the requests - The user agents that identify the source of each request, which include AI crawlers - The time of each request, which makes the real-time answer enrichment visible Usually, your infrastructure team can export these logs from your CDN or your web server. You do not have to add special instrumentation.","keywords":""},{"kind":"section","title":"CDN log integration","heading":"What Genezio analyzes in the logs","section":"Documentation","crumbs":["Integrations","CDN log integration","What Genezio analyzes in the logs"],"url":"/docs/integrations/cdn-log-integration#what-genezio-analyzes-in-the-logs","text":"Raw logs are long lists of separate requests. Alone, they tell you very little. Genezio analyzes them and gives you reports on three items.","keywords":""},{"kind":"section","title":"CDN log integration","heading":"Clustering by keyword","section":"Documentation","crumbs":["Integrations","CDN log integration","Clustering by keyword"],"url":"/docs/integrations/cdn-log-integration#clustering-by-keyword","text":"Genezio clusters your log entries by keyword. It puts the requested paths into themes. Thus, Genezio reports the activity at the level of topics, not of single URLs. When a different view is useful, Genezio can cluster the same logs in more than one way.","keywords":""},{"kind":"section","title":"CDN log integration","heading":"Crawler breakdown","section":"Documentation","crumbs":["Integrations","CDN log integration","Crawler breakdown"],"url":"/docs/integrations/cdn-log-integration#crawler-breakdown","text":"Genezio makes a crawler breakdown. It identifies the bots and the crawlers that access your site. It gives special attention to AI crawlers, which are the crawlers that answer engines use to fetch content. This keeps the traffic of answer engines separate from search engines, scrapers, and usual visitors.","keywords":""},{"kind":"section","title":"CDN log integration","heading":"AI-crawler traffic compared with citations","section":"Documentation","crumbs":["Integrations","CDN log integration","AI-crawler traffic compared with citations"],"url":"/docs/integrations/cdn-log-integration#ai-crawler-traffic-compared-with-citations","text":"The analysis with the most value connects the two halves. It compares the AI-crawler traffic with your citations. It examines if the pages that answer engines fetch agree with the pages that they really cite in their answers. This analysis shows the gaps. Examples are important pages that answer engines ignore, or pages that they fetch frequently but never cite.","keywords":""},{"kind":"section","title":"CDN log integration","heading":"Request the log ingestion","section":"Documentation","crumbs":["Integrations","CDN log integration","Request the log ingestion"],"url":"/docs/integrations/cdn-log-integration#request-the-log-ingestion","text":"The CDN log ingestion is not self-serve. Genezio does it on request for each customer. Contact Genezio and tell us that you want us to ingest your CDN logs or server logs. Genezio works with you and your infrastructure team to agree on how you send the logs. Genezio analyzes the data and gives you custom reports from your CDN data and your GA4 data. The reports show what answer engines crawl, what they fetch in real time, and how this compares with the pages that they cite.","keywords":""},{"kind":"section","title":"CDN log integration","heading":"Related pages","section":"Documentation","crumbs":["Integrations","CDN log integration","Related pages"],"url":"/docs/integrations/cdn-log-integration#related-pages","text":"- Citations - Actionable insights - How LLMs select sources","keywords":""},{"kind":"page","title":"Data exports","heading":"","section":"Documentation","crumbs":["Documentation","Integrations","Data exports"],"url":"/docs/integrations/data-exports","text":"Export Genezio reports and content analyses as PDF, and tabular views as CSV. Learn when to use the API in place of a manual export, and how to compare exports. You can export Genezio reports from the product, for example for a board pack, a client report, or a data warehouse.","keywords":""},{"kind":"section","title":"Data exports","heading":"PDF","section":"Documentation","crumbs":["Integrations","Data exports","PDF"],"url":"/docs/integrations/data-exports#pdf","text":"You can download reports and content analyses as PDF files. Use this format for a document that a person reads. Examples are a monthly client report, an appendix for a slide deck, or a record of the status at a point in time.","keywords":""},{"kind":"section","title":"Data exports","heading":"CSV","section":"Documentation","crumbs":["Integrations","Data exports","CSV"],"url":"/docs/integrations/data-exports#csv","text":"You can export the tabular views as CSV files for a spreadsheet or a data warehouse. This format is useful when you want to combine the Genezio numbers with data that Genezio does not have. Examples are the pipeline, the revenue, and the campaign spend.","keywords":""},{"kind":"section","title":"Data exports","heading":"The API, for each export that you repeat","section":"Documentation","crumbs":["Integrations","Data exports","The API, for each export that you repeat"],"url":"/docs/integrations/data-exports#the-api-for-each-export-that-you-repeat","text":"If you do an export each month, use the API in place of a manual export. The API gives access to these data: - Brands - Topics - Conversations - Citations - Visibility - Content - The Business Scorecard The API returns the same numbers that the dashboard shows. The practical rule: use a manual export for a task that you do one time, and use the API for a routine task. Over time, a person can forget to do a monthly manual export.","keywords":""},{"kind":"section","title":"Data exports","heading":"Compare exports","section":"Documentation","crumbs":["Integrations","Data exports","Compare exports"],"url":"/docs/integrations/data-exports#compare-exports","text":"An export is a snapshot of a measurement that changes. Two exports from different weeks are different, partly because answer engines change their answers between runs. When you compare exports, use the same topics and the same answer engines. Look at the direction, not at small differences. Refer to Trend tracking.","keywords":""},{"kind":"page","title":"External reporting tools","heading":"","section":"Documentation","crumbs":["Documentation","Integrations","External reporting tools"],"url":"/docs/integrations/external-reporting-tools","text":"Send Genezio AI visibility numbers to Looker Studio, a BI tool, a data warehouse, or a client dashboard through the API, with a pattern for scheduled pulls. Most teams want the Genezio numbers together with all their other reports. These reports can be in Looker Studio, a BI tool, a data warehouse, or a client dashboard.","keywords":""},{"kind":"section","title":"External reporting tools","heading":"How the data gets to a different tool","section":"Documentation","crumbs":["Integrations","External reporting tools","How the data gets to a different tool"],"url":"/docs/integrations/external-reporting-tools#how-the-data-gets-to-a-different-tool","text":"The API sends Genezio data to a different tool. It gives the same numbers that the dashboard shows: - Visibility and recommendation - Visibility for each topic - Share against competitors - Citations - Conversations - Sentiment and perceptions - The Business Scorecard You can generate a typed client from the OpenAPI document. You do not have to write the client yourself. Refer to the API overview.","keywords":""},{"kind":"section","title":"External reporting tools","heading":"A workable pattern","section":"Documentation","crumbs":["Integrations","External reporting tools","A workable pattern"],"url":"/docs/integrations/external-reporting-tools#a-workable-pattern","text":"1. Create an API key for only the brands that the report needs. Refer to Authentication. 2. Get the data on a schedule that agrees with how frequently you run conversations. When you get data each hour but run conversations each week, you get noise, not newer data. 3. Keep each reading with its date. Thus, you can show the direction, not only a single number. 4. Stay in the rate limit of 200 requests each minute for each key. Give each system its own key.","keywords":""},{"kind":"section","title":"External reporting tools","heading":"What to put on a shared dashboard","section":"Documentation","crumbs":["Integrations","External reporting tools","What to put on a shared dashboard"],"url":"/docs/integrations/external-reporting-tools#what-to-put-on-a-shared-dashboard","text":"Show the direction, not a single number. A visibility number without a trend and without a competitor next to it causes a question that nobody can answer. These three items are the most useful on a shared dashboard: - Your visibility over time. - Your position against named competitors on the same questions. - The progress on the Business Scorecard goals that the team agreed.","keywords":""},{"kind":"section","title":"External reporting tools","heading":"What not to build","section":"Documentation","crumbs":["Integrations","External reporting tools","What not to build"],"url":"/docs/integrations/external-reporting-tools#what-not-to-build","text":"Do not build the conversation reader again in a different tool. A Genezio number has value because you can open the answers below it. An export removes this possibility. Report the numbers outside Genezio, and do the investigation in Genezio.","keywords":""},{"kind":"page","title":"Security and Data","heading":"","section":"Documentation","crumbs":["Documentation","Security","Security and Data"],"url":"/docs/security","text":"How Genezio controls access, identity and the data in your account: enterprise SSO and SCIM, access by brand, API keys by brand, and the security contact. This section tells how Genezio controls access, identity, and the data in your account.","keywords":""},{"kind":"section","title":"Security and Data","heading":"Identity","section":"Documentation","crumbs":["Security","Security and Data","Identity"],"url":"/docs/security#identity","text":"Enterprise SSO and SCIM tells how your users sign in through your own identity provider. It also tells how your identity provider adds members automatically.","keywords":""},{"kind":"section","title":"Security and Data","heading":"Access in an account","section":"Documentation","crumbs":["Security","Security and Data","Access in an account"],"url":"/docs/security#access-in-an-account","text":"You can control the access in an account in two ways: - You can give a member access to only specified brands, not to the full account. - You can give an API key access to only specified brands. Thus, an integration gets access only to the brands that it needs.","keywords":""},{"kind":"section","title":"Security and Data","heading":"Policy pages","section":"Documentation","crumbs":["Security","Security and Data","Policy pages"],"url":"/docs/security#policy-pages","text":"For the policy pages of this section, contact your account manager or security@genezio.com. These pages are about data collection, data retention, privacy, certifications, and compliance. They are formal commitments. Thus, Genezio answers them directly and does not give a summary here.","keywords":""},{"kind":"page","title":"Enterprise SSO and SCIM","heading":"","section":"Documentation","crumbs":["Documentation","Security","Enterprise SSO and SCIM"],"url":"/docs/security/enterprise-sso-and-scim","text":"Configure enterprise SSO with SAML 2.0 and SCIM provisioning for Genezio with Okta, Microsoft Entra ID or another IdP, and learn how roles and offboarding work. This page explains how to configure enterprise SSO (SAML) and SCIM provisioning for Genezio, and how the two operate after the setup. With enterprise SSO, your team signs in to Genezio through your own identity provider (IdP). Examples are Okta, Microsoft Entra ID, OneLogin, Ping, or any SAML 2.0 IdP. SCIM provisioning does more. It makes your IdP the source of truth for who has a Genezio account and what each person can do. You configure the two for each customer, together with the Genezio team. They are not self-serve. This page tells you these items: - What you must do - What Genezio does - How SSO and SCIM operate after you start to use them.","keywords":""},{"kind":"section","title":"Enterprise SSO and SCIM","heading":"What enterprise SSO and SCIM give you","section":"Documentation","crumbs":["Security","Enterprise SSO and SCIM","What enterprise SSO and SCIM give you"],"url":"/docs/security/enterprise-sso-and-scim#what-enterprise-sso-and-scim-give-you","text":"Single sign-on (SAML): Your users sign in with your corporate credentials, under your own MFA policies and session policies. You do not manage a separate Genezio password. SCIM provisioning: When a person joins the correct group in your IdP, Genezio automatically creates the account of that person. When the person leaves, Genezio removes the access. You do not send invitations manually, and no accounts stay active after offboarding. Then, an account Owner sets the roles in Genezio. Refer to Roles are assigned in Genezio, not in your IdP. Most organizations want the two. They solve different problems: - SSO answers this question: \"Is this person the person that they say?\" - SCIM answers this question: \"Must this person have an account, and what can the person do with it?\"","keywords":""},{"kind":"section","title":"Enterprise SSO and SCIM","heading":"SSO does not create accounts","section":"Documentation","crumbs":["Security","Enterprise SSO and SCIM","SSO does not create accounts"],"url":"/docs/security/enterprise-sso-and-scim#sso-does-not-create-accounts","text":"SSO does not create accounts. A successful sign-in proves the identity of the person. It does not give access to your Genezio account. A first-time user can authenticate through your IdP but have no Genezio account and no pending invitation. Genezio then refuses the user and shows a message that the user needs an invitation. Thus, a user must first exist in Genezio. SCIM or an invitation from an account Owner creates the user. If you use SSO without SCIM, you must invite each user manually.","keywords":""},{"kind":"section","title":"Enterprise SSO and SCIM","heading":"How the SSO and SCIM setup operates","section":"Documentation","crumbs":["Security","Enterprise SSO and SCIM","How the SSO and SCIM setup operates"],"url":"/docs/security/enterprise-sso-and-scim#how-the-sso-and-scim-setup-operates","text":"Ask your Genezio representative to start the process. Genezio then does these actions: - It registers your email domains and creates the connection to the identity provider for your organization. - It sends you the two values that you need to configure the SAML application in your IdP: the Identifier (Entity ID) and the Reply URL (Assertion Consumer Service URL). - It enables SCIM on your account. Until Genezio enables SCIM, your account settings do not show the SCIM entry. Then, do these steps: Configure SAML in your IdP with the two values. Send Genezio the SAML metadata XML of your IdP. Test with one pilot user. This user must already have a Genezio account or a pending invitation. Connect SCIM with a token that you make in Genezio. Provision the same pilot user. Look for email conflicts before you provision all users. Refer to Troubleshooting. Provision your team. Then, an Owner gives a different role to each user who needs more than the default Member role. Deactivate a test user. Make sure that the user has no access.","keywords":""},{"kind":"section","title":"Enterprise SSO and SCIM","heading":"Connect your identity provider with SAML 2.0","section":"Documentation","crumbs":["Security","Enterprise SSO and SCIM","Connect your identity provider with SAML 2.0"],"url":"/docs/security/enterprise-sso-and-scim#connect-your-identity-provider-with-saml-20","text":"The SSO setup is an exchange in two directions. Genezio sends you two values, and you send Genezio one file. What Genezio gives you. Create a standard SAML 2.0 application in your IdP and enter these values: Identifier (Entity ID): Audience URI, SP Entity ID. Reply URL (Assertion Consumer Service URL): ACS URL, Single sign-on URL The two values are specific to the connection of your organization. Use the values that Genezio sends you. Do not copy them from a different tenant. What you send to Genezio. Export the SAML metadata XML of your IdP and send it to your Genezio representative. Genezio uses it to complete the federation. Nothing occurs until Genezio has this file. Thus, this step usually causes the delays in a rollout. If your IdP can give you only a metadata URL, tell Genezio. Genezio can use the URL. Send the email as an attribute. The email is mandatory. Send it as the NameID in the emailAddress format, as an explicit email attribute, or as the two. The display name is optional, but we recommend it. Without it, Genezio uses the part of the email before the @. The email must be exactly the same. Genezio compares the address from your IdP with the address that the user typed on the sign-in page. Upper case and lower case are not important, but all other differences are. These differences cause the sign-in to fail: - An alias - A UPN that is different from the mail attribute - A legacy domain Make sure that the attribute that you send is the address that your users know as their work email. Give Genezio all domains. SSO uses email domains as the key. Thus, include secondary domains and the domains of acquired companies. Each domain connects to one identity provider. You cannot divide one domain between two identity providers. The sign-in starts at Genezio. Users go to the Genezio sign-in page and enter their work email. Genezio then sends them to your IdP. Genezio does not support a sign-in that starts at the IdP. If you want a tile in your app catalog, make the tile a bookmark to the Genezio sign-in page.","keywords":""},{"kind":"section","title":"Enterprise SSO and SCIM","heading":"Enable SCIM provisioning","section":"Documentation","crumbs":["Security","Enterprise SSO and SCIM","Enable SCIM provisioning"],"url":"/docs/security/enterprise-sso-and-scim#enable-scim-provisioning","text":"You must have the Owner role, and SCIM must be enabled on your account. Go to Settings - SCIM. Create a token. Give it a name that you can identify later, for example \"Okta production\". Optionally, set an expiry date. Copy the token immediately. Genezio shows it only one time, and you cannot get it again later. Only the first characters stay visible. Thus, you can identify each token. In the provisioning settings of your IdP, paste the token and the base URL that Genezio shows next to it. Use bearer-token authentication. You can do these actions on a token: - Rotate it: The token gets a new secret with the same name and expiry date. The old secret stops immediately. - Disable it and enable it again. - Delete it permanently. Each token operates for only one Genezio account. It can act only on the users that it provisioned. Treat the token as a high-value credential and keep it in the secret store of your IdP. When you disable or delete a token, provisioning stops. But no user loses access. To offboard users, first deprovision them through your IdP. Then remove the connection.","keywords":""},{"kind":"section","title":"Enterprise SSO and SCIM","heading":"Roles are assigned in Genezio, not in your IdP","section":"Documentation","crumbs":["Security","Enterprise SSO and SCIM","Roles are assigned in Genezio, not in your IdP"],"url":"/docs/security/enterprise-sso-and-scim#roles-are-assigned-in-genezio-not-in-your-idp","text":"Genezio does not support SCIM groups. The SCIM implementation of Genezio includes only the lifecycle of users: it creates, updates, deactivates, and removes users. It does not have a Group resource. Thus, your IdP cannot send groups or assign roles. Each user that SCIM creates has the Member role. To give a user a different role, an account Owner changes it in the Genezio dashboard under Users. You change roles only in the application. Your IdP cannot change roles. Genezio has three account roles: Owner: Full control: manage brands, invite and remove users, change roles, billing. Member: A collaborator with access to the data. A Member can invite other users, but cannot create or delete brands. Viewer: Read-only access: a Viewer can see reports and data, but cannot change anything. Thus, provisioning and permissions are two separate tasks. Your IdP sets who has an account. An Owner sets what each user can do. For most teams, this is one step for each person, one time only. Member is the correct role for most users.","keywords":""},{"kind":"section","title":"Enterprise SSO and SCIM","heading":"SCIM connector settings","section":"Documentation","crumbs":["Security","Enterprise SSO and SCIM","SCIM connector settings"],"url":"/docs/security/enterprise-sso-and-scim#scim-connector-settings","text":"Genezio uses SCIM 2.0. Thus, each connector that agrees with the standard operates. Configure these items: - Keep group provisioning and role provisioning off. Genezio shows only the User resource. Thus, a group push has no target. Be careful with Microsoft Entra ID, because it enables group mapping by default. Disable it. - Disable bulk operations and sorting. Genezio provisions one user at a time. - Keep the filters simple. Genezio supports only equality filters on a single attribute. - Do not synchronize passwords. SCIM users authenticate through your IdP, not with a Genezio password. The provisioning validator of Entra shows bulk, sort, and etag as not supported. This is expected and does not stop the provisioning.","keywords":""},{"kind":"section","title":"Enterprise SSO and SCIM","heading":"What occurs when a person leaves the organization","section":"Documentation","crumbs":["Security","Enterprise SSO and SCIM","What occurs when a person leaves the organization"],"url":"/docs/security/enterprise-sso-and-scim#what-occurs-when-a-person-leaves-the-organization","text":"When you deactivate or delete a user in your IdP, Genezio immediately removes the membership of the user in the Genezio account. It also removes all permissions of the user. If this was the only Genezio account of the user, Genezio also revokes the active session. A person who is also in a different Genezio account keeps the access to that account. Know this limit: SSO does not disable the sign-in with a password. SSO gives your users one more method to sign in. It does not remove a password that a user already has, and it does not prevent its use. Today, Genezio has no setting that permits only SSO sign-in for a domain. If your security review requires this, tell Genezio before you plan the rollout. This function needs development work from Genezio. A change of the configuration is not sufficient.","keywords":""},{"kind":"section","title":"Enterprise SSO and SCIM","heading":"Troubleshooting","section":"Documentation","crumbs":["Security","Enterprise SSO and SCIM","Troubleshooting"],"url":"/docs/security/enterprise-sso-and-scim#troubleshooting","text":"\"SSO is not configured for this email domain\": The domain is not registered, or it is disabled. Ask Genezio to register the domain or to enable it again. \"You need an invitation before signing in with SSO\": The sign-in was successful, but this person has no Genezio account. Provision the person through SCIM, or ask an Owner to invite the person. \"Does not match the email you entered\": Your IdP sent an address that is different from the address that the user typed. Make the email attribute of your IdP the same as the work email of the user. \"Your SSO session expired\": An old sign-in, usually from many open login tabs. Close the other tabs and try again. SCIM: invalid or expired token: The token is incorrect, disabled, or after its expiry date. Rotate the token and update your IdP. SCIM: SCIM is not enabled for this account: The token is valid, but the function is not enabled. Contact Genezio. SCIM: user already exists in another account: This email belongs to a user in a different Genezio account. Contact Genezio support to detach the user. Your IdP shows failures when it sends groups: Genezio has no Group resource. It does not support group provisioning. Disable group provisioning in your connector. Set the roles in Genezio under Users. The SCIM entry is not in Settings: SCIM is not enabled yet, or you are not an Owner. Contact Genezio, and examine your role. Look for email conflicts. Some persons registered for Genezio themselves before your enterprise rollout. They already have an account with their work email, and their provisioning fails. Find these persons and solve the conflicts with Genezio support before you provision all users.","keywords":""},{"kind":"section","title":"Enterprise SSO and SCIM","heading":"What to send to Genezio for the setup","section":"Documentation","crumbs":["Security","Enterprise SSO and SCIM","What to send to Genezio for the setup"],"url":"/docs/security/enterprise-sso-and-scim#what-to-send-to-genezio-for-the-setup","text":"For a quick setup, prepare these items: - All email domains that your users sign in with - The name of your identity provider - The SAML metadata XML of your IdP (Genezio cannot continue without this file) - The attribute that you send as the email, with a confirmation that it is the same as the work emails of the users - The Genezio account that SCIM must manage, and the email of its Owner","keywords":""},{"kind":"section","title":"Enterprise SSO and SCIM","heading":"Related pages","section":"Documentation","crumbs":["Security","Enterprise SSO and SCIM","Related pages"],"url":"/docs/security/enterprise-sso-and-scim#related-pages","text":"- Sign-in and SSO options - Users","keywords":""},{"kind":"page","title":"API overview","heading":"","section":"API","crumbs":["API","Guides","API overview"],"url":"/docs/api","text":"The Genezio API is a REST API for AI visibility data. Read the visibility, citations and sentiment of your brands in answer engines from your own code. The Genezio API is the REST API of the Genezio platform. It gives your code the AI visibility data of your brands: how answer engines mention, recommend and cite each brand. Make a key in the dashboard. A key works only on the public API. Make your first request and read your first report.","keywords":""},{"kind":"section","title":"API overview","heading":"What data the Genezio API gives you","section":"API","crumbs":["Guides","API overview","What data the Genezio API gives you"],"url":"/docs/api#what-data-the-genezio-api-gives-you","text":"Visibility and recommendation: How often each engine mentions your brand, and how often it recommends it, period by period. Citations: The sources that the answers cite, by domain and by page, and who owns them. Sentiment and perceptions: What the answers say about your brand, and which of it your facts confirm. Brands, topics and competitors: The brands of your account, their topics and scenarios, and their competitors.","keywords":""},{"kind":"section","title":"API overview","heading":"Your first API request","section":"API","crumbs":["Guides","API overview","Your first API request"],"url":"/docs/api#your-first-api-request","text":"The Quickstart makes your key, and the API reference gives each route with a playground.","keywords":"curl https://app.backend.genezio.ai/customer/v1/brands \\ -H X-API-Key: $GENEZIO_API_KEY import os import requests answer = requests.get( https://app.backend.genezio.ai/customer/v1/brands headers= X-API-Key : os.environ GENEZIO_API_KEY ) print(answer.json() data ) const answer = await fetch( https://app.backend.genezio.ai/customer/v1/brands headers: X-API-Key : process.env.GENEZIO_API_KEY ); console.log((await answer.json()).data);"},{"kind":"section","title":"API overview","heading":"Statistics and errors that you can rely on","section":"API","crumbs":["Guides","API overview","Statistics and errors that you can rely on"],"url":"/docs/api#statistics-and-errors-that-you-can-rely-on","text":"- Math, not noise. Each rate has its interval of confidence, and each change says if it is significant. Alert on a real change only. - The headline is ready. info.summary gives the last period of your brand and its rank among your competitors. - Errors with a stable code. Compare error.code with a constant, and send request_id to the support.","keywords":""},{"kind":"page","title":"Quickstart","heading":"","section":"API","crumbs":["API","Get started","Quickstart"],"url":"/docs/api/quickstart","text":"Make a Genezio API key, list your brands, and read the AI visibility report of a brand in three steps, with curl examples and sample responses. This quickstart for the Genezio API makes an API key, lists your brands, and reads the AI visibility report of one brand. For the details of the key, refer to Authentication. In the dashboard, open the menu of your account and select API Keys. Select Create API key, give the key a name, and select Create. Copy the key now. The dashboard shows it one time only. A key starts with gnz-. Keep the key in a secret store or in an environment variable. Do not put it in the code of a page or in a repository. Each row is one brand in one week. info.summary gives the last week of your brand and its rank:.","keywords":"export GENEZIO_API_KEY= gnz-... curl https://app.backend.genezio.ai/customer/v1/brands \\ -H X-API-Key: $GENEZIO_API_KEY data : id : 12 name : Acme industry : Retail websites : https://acme.com info : total_rows : 1 page : 1 page_size : 50 curl https://app.backend.genezio.ai/customer/v1/brands/12/reports/visibility?start_date=2026-09-01&end_date=2026-09-30&granularity=week \\ -H X-API-Key: $GENEZIO_API_KEY info : summary : date : 2026-09-28 visibility : 0.57 change : 0.03 significant : true rank : 2 brands_ranked : 12"},{"kind":"section","title":"Quickstart","heading":"Next steps with the Genezio API","section":"API","crumbs":["Get started","Quickstart","Next steps with the Genezio API"],"url":"/docs/api/quickstart#next-steps-with-the-genezio-api","text":"- Read a response: the envelope of each answer. - Date ranges: how the API reads start_date and end_date. - API reference: try each route with your key.","keywords":""},{"kind":"page","title":"Authentication","heading":"","section":"API","crumbs":["API","Get started","Authentication"],"url":"/docs/api/authentication","text":"Genezio API authentication uses API keys in the X-API-Key header, or OAuth 2.0 client credentials tokens. Learn how to limit, rotate and revoke a key. This page explains Genezio API authentication: how to send an API key, how to limit it, and how to exchange it for an OAuth 2.0 token. Each request sends an API key. A key belongs to one account and reads the data of that account only.","keywords":""},{"kind":"section","title":"Authentication","heading":"Send the API key in a header","section":"API","crumbs":["Get started","Authentication","Send the API key in a header"],"url":"/docs/api/authentication#send-the-api-key-in-a-header","text":"Send the key in the X-API-Key header: The Authorization header also works, for an HTTP client that has a field for a bearer token only: A request with no key, or with a key that is incorrect, expired or revoked, gets 401.","keywords":"curl https://app.backend.genezio.ai/customer/v1/brands -H X-API-Key: gnz-... curl https://app.backend.genezio.ai/customer/v1/brands -H Authorization: Bearer gnz-..."},{"kind":"section","title":"Authentication","heading":"The limits of an API key: access, brands and expiration","section":"API","crumbs":["Get started","Authentication","The limits of an API key: access, brands and expiration"],"url":"/docs/api/authentication#the-limits-of-an-api-key-access-brands-and-expiration","text":"The owner of the account sets the limits of a key when they make it. Access: Read only, or read and write. A key that reads only gets 403 insufficient_scope on a write. Brands: All brands of the account, also the brands that you add later, or some brands. A key of some brands gets 403 brand_access_denied on each other brand. Expiration: 30, 90 or 180 days, 1 year, or never. Give each program its own key, with the fewest brands that it needs. Then a key that leaks reads less data, and you revoke the key of one program only.","keywords":""},{"kind":"section","title":"Authentication","heading":"API keys and your plan","section":"API","crumbs":["Get started","Authentication","API keys and your plan"],"url":"/docs/api/authentication#api-keys-and-your-plan","text":"The plan of your account gives the number of API keys that the account can hold at the same time. The API Keys page of the dashboard shows the keys that you use and the limit of your plan. - When the account holds each key that the plan permits, a new key gets 403. Revoke a key, or change the plan. - When the plan does not include API access, each key of the account gets 403 forbidden, also a key that you made before.","keywords":""},{"kind":"section","title":"Authentication","heading":"Revoke an API key","section":"API","crumbs":["Get started","Authentication","Revoke an API key"],"url":"/docs/api/authentication#revoke-an-api-key","text":"In API Keys, select the trash icon of the key. The key and each token of the key stop at that moment.","keywords":""},{"kind":"section","title":"Authentication","heading":"OAuth 2.0 tokens with client credentials","section":"API","crumbs":["Get started","Authentication","OAuth 2.0 tokens with client credentials"],"url":"/docs/api/authentication#oauth-20-tokens-with-client-credentials","text":"A program that uses OAuth 2.0 can exchange the key for a token of 30 minutes. The dashboard shows the client_id of the key beside the key. Send the token as Authorization: Bearer. The token has the limits of its key.","keywords":"curl -X POST https://app.backend.genezio.ai/oauth/token \\ -d grant_type=client_credentials \\ -d client_id=gzo_svc_... \\ -d client_secret=gnz-..."},{"kind":"page","title":"Responses","heading":"","section":"API","crumbs":["API","Concepts","Responses"],"url":"/docs/api/guides/responses","text":"Each Genezio API response has the rows in data and the facts of the query in info. Learn the fields of info, the pagination, the list parameters and the scales. This page explains how to read a response of the Genezio API. Each answer has the same envelope: - data holds the rows. Each route gives the shape of its rows in the API reference. - info holds the facts of the query.","keywords":"data : ... info : ..."},{"kind":"section","title":"Responses","heading":"The fields of info","section":"API","crumbs":["Concepts","Responses","The fields of info"],"url":"/docs/api/guides/responses#the-fields-of-info","text":"total_rows: Each list The rows that match the query, on each page together. page, page_size: Each list with pages The page of this answer. query: Each report Your request, with the value of each default: the range of dates, the engines, and each filter. summary: Each report The headline numbers, ready to show.","keywords":""},{"kind":"section","title":"Responses","heading":"Pages of a list","section":"API","crumbs":["Concepts","Responses","Pages of a list"],"url":"/docs/api/guides/responses#pages-of-a-list","text":"A list with pages takes page (from 1) and page_size (at most 100, or 200 for the brands). Read the next page while page page_size < total_rows.","keywords":"page = 1 while True: answer = get( /customer/v1/brands/12/citations/urls page=page page_size=100) handle(answer data ) if page answer info page_size = answer info total_rows : break page += 1"},{"kind":"section","title":"Responses","heading":"Lists in the query parameters","section":"API","crumbs":["Concepts","Responses","Lists in the query parameters"],"url":"/docs/api/guides/responses#lists-in-the-query-parameters","text":"Repeat a parameter to send a list:","keywords":"?models=chatgpt.com&models=claude&topic_ids=5&topic_ids=7"},{"kind":"section","title":"Responses","heading":"Scales of the values","section":"API","crumbs":["Concepts","Responses","Scales of the values"],"url":"/docs/api/guides/responses#scales-of-the-values","text":"- A rate or a share is a number from 0 to 1, for example 0.57 for 57%. - A count is a whole number. - A date is YYYY-MM-DD, the first day of the period.","keywords":""},{"kind":"section","title":"Responses","heading":"Unknown parameters get an error","section":"API","crumbs":["Concepts","Responses","Unknown parameters get an error"],"url":"/docs/api/guides/responses#unknown-parameters-get-an-error","text":"A parameter with an incorrect name, such as from in place of start_date, gets 422 validation_error. The API does not ignore it. Thus, a typing error does not give a report of the incorrect range.","keywords":""},{"kind":"page","title":"Date ranges","heading":"","section":"API","crumbs":["API","Concepts","Date ranges"],"url":"/docs/api/guides/date-ranges","text":"Learn how the Genezio API reads start_date and end_date: UTC days, inclusive ends, the 366-day limit, and the day, week and month periods of granularity. This page explains how the Genezio API reads a date range. Each report reads one range of dates, with start_date and end_date.","keywords":""},{"kind":"section","title":"Date ranges","heading":"The rules of a date range","section":"API","crumbs":["Concepts","Date ranges","The rules of a date range"],"url":"/docs/api/guides/date-ranges#the-rules-of-a-date-range","text":"- A date is YYYY-MM-DD, or a datetime of ISO 8601 such as 2026-09-01T08:00:00Z. - A date is a day in UTC. - Both ends are inclusive. An end_date of 2026-09-30 covers that whole day, up to 23:59:59.999999 UTC. - A datetime with no time zone is in UTC. - With no start_date, the range starts 30 days before its end. - With no end_date, the range ends now. - A range holds at most 366 days.","keywords":""},{"kind":"section","title":"Date ranges","heading":"Examples of date ranges","section":"API","crumbs":["Concepts","Date ranges","Examples of date ranges"],"url":"/docs/api/guides/date-ranges#examples-of-date-ranges","text":"start_date=2026-09-01&end_date=2026-09-30: The whole of September, in UTC. start_date=2026-09-01T08:00:00&end_date=2026-09-01T17:00:00: From 08:00 to 17:00 UTC on 1 September. start_date=2026-09-01T08:00:00+02:00: From 06:00 UTC. No dates: The last 30 days. info.query gives the range that the API read, with each end as a full datetime:","keywords":"query : start_date : 2026-09-01T00:00:00+00:00 end_date : 2026-09-30T23:59:59.999999+00:00"},{"kind":"section","title":"Date ranges","heading":"Periods and granularity","section":"API","crumbs":["Concepts","Date ranges","Periods and granularity"],"url":"/docs/api/guides/date-ranges#periods-and-granularity","text":"A report with granularity divides the range into periods: day: Each day. week: Each Monday. month: The first day of each month. The date of a row is the first day of its period. The first and the last period can be partial: a range that starts on a Wednesday gives a first week that starts on the Monday before it. A range that ends now gives a last period that is not complete. Compare the last period with care, or end the range at the last complete period.","keywords":""},{"kind":"page","title":"Errors","heading":"","section":"API","crumbs":["API","Concepts","Errors"],"url":"/docs/api/guides/errors","text":"Each Genezio API error has one JSON shape and a stable code. Read the list of error codes, from 400 to 500, and an example of error handling in Python. This page lists the errors of the Genezio API and tells you how to handle them. Each error has this shape: code: Stable. Compare it with a constant in your code. message: For a person. It can change at any time. request_id: Send it to support@genez.io. The support finds the request in the logs with it. details: Only with validation_error: one item for each field that is not valid.","keywords":"error : code : brand_access_denied message : This API key has no access to this brand request_id : 4f3c2e1a-9b7d-4c1e-8f2a-1d3e5b7c9a0f"},{"kind":"section","title":"Errors","heading":"The error codes of the API","section":"API","crumbs":["Concepts","Errors","The error codes of the API"],"url":"/docs/api/guides/errors#the-error-codes-of-the-api","text":"400: invalid_request The request names a thing that the brand does not hold, such as a topic of another brand. 401: unauthorized The request has no key. 401: invalid_api_key The key is wrong, expired or revoked. 403: insufficient_scope The key reads only, and the request writes. 403: brand_access_denied The key has no access to this brand. 403: forbidden The key has no access to this thing, or the plan of the account does not include API access. 403: plan_limit_reached The plan of the account holds no more of this thing, such as more scenarios. Nothing was written. 404: not_found The thing, or the route, does not exist. 409: conflict The thing already exists, or it is not in a state for this request, such as a restore of a scenario that is not deleted. 409: idempotency_in_progress A request with the same Idempotency-Key did not finish yet. Send it again in a few seconds. 422: validation_error A parameter is not valid. details names each one. 422: idempotency_key_reused The Idempotency-Key came with another request. Use a new key for each new write. 429: rate_limited The key sent too many requests. See Rate limits. 500: internal_error An error of the API. Try again, and send request_id to the support if it continues.","keywords":""},{"kind":"section","title":"Errors","heading":"An example of a validation error","section":"API","crumbs":["Concepts","Errors","An example of a validation error"],"url":"/docs/api/guides/errors#an-example-of-a-validation-error","text":"","keywords":"error : code : validation_error message : The request has fields that are not valid. request_id : 9147c93b-ca37-41a8-ac69-71706e03866d details : field : page_size message : Input should be less than or equal to 100"},{"kind":"section","title":"Errors","heading":"Handle the errors in your code","section":"API","crumbs":["Concepts","Errors","Handle the errors in your code"],"url":"/docs/api/guides/errors#handle-the-errors-in-your-code","text":"","keywords":"answer = requests.get(url headers= X-API-Key : key ) if answer.status_code = 400: error = answer.json() error if error code == rate_limited : time.sleep(int(answer.headers Retry-After )) elif error code in ( unauthorized invalid_api_key ): raise RuntimeError( Make a new API key. ) else: raise RuntimeError(f error 'code' : error 'message' ( error 'request_id' ) )"},{"kind":"page","title":"Rate limits","heading":"","section":"API","crumbs":["API","Concepts","Rate limits"],"url":"/docs/api/guides/rate-limits","text":"Each Genezio API key can send 200 requests in one minute. Learn the X-RateLimit headers, the 429 response with Retry-After, and how to stay under the limit. This page explains the rate limit of the Genezio API. Each API key can send 200 requests in one minute. A token of the key counts against the same key.","keywords":""},{"kind":"section","title":"Rate limits","heading":"The rate limit headers","section":"API","crumbs":["Concepts","Rate limits","The rate limit headers"],"url":"/docs/api/guides/rate-limits#the-rate-limit-headers","text":"Each answer has the headers of the limit: X-RateLimit-Limit: The requests of one minute: 200. X-RateLimit-Remaining: The requests that remain in this minute. X-RateLimit-Reset: The seconds until the minute ends.","keywords":""},{"kind":"section","title":"Rate limits","heading":"What occurs when you send too many requests","section":"API","crumbs":["Concepts","Rate limits","What occurs when you send too many requests"],"url":"/docs/api/guides/rate-limits#what-occurs-when-you-send-too-many-requests","text":"The request after the limit gets 429 with the code rate_limited, and the Retry-After header gives the seconds to wait:","keywords":"HTTP/1.1 429 Too Many Requests Retry-After: 23 X-RateLimit-Limit: 200 X-RateLimit-Remaining: 0 X-RateLimit-Reset: 23"},{"kind":"section","title":"Rate limits","heading":"Good practice for the rate limit","section":"API","crumbs":["Concepts","Rate limits","Good practice for the rate limit"],"url":"/docs/api/guides/rate-limits#good-practice-for-the-rate-limit","text":"- Read X-RateLimit-Remaining, and send fewer requests before it gets to 0. - After a 429, wait for Retry-After seconds, then try again. - Ask for a whole range in one report, in place of one request for each day. - For a higher limit, write to support@genez.io.","keywords":""},{"kind":"page","title":"Write data","heading":"","section":"API","crumbs":["API","Concepts","Write data"],"url":"/docs/api/guides/writes","text":"Change the setup of a brand with the Genezio API: topics, scenarios and competitors. Learn idempotency keys, dry runs, partial writes and the writes that cost. This page explains how to change the setup of a brand with the Genezio API. The routes that write use POST, PATCH and DELETE. They need a key with the write scope. A key that reads only gets 403 insufficient_scope. Each write does what the same change in the dashboard does: it checks the plan of the account, and it goes to the audit log of the account. Each entry gives the key and the address of the request: To see what one key changed, search for its name on the Audit logs page of the dashboard.","keywords":"via : public_api api_key_id : 3 api_key_name : Nightly sync api_key_prefix : gnz-ab12 ip : 203.0.113.9"},{"kind":"section","title":"Write data","heading":"Send a write again safely with an idempotency key","section":"API","crumbs":["Concepts","Write data","Send a write again safely with an idempotency key"],"url":"/docs/api/guides/writes#send-a-write-again-safely-with-an-idempotency-key","text":"A network can fail after the API writes and before your program gets the answer. Send an Idempotency-Key header with each POST and PATCH, and send the same key again when you retry: First request with the key: The API writes and answers. It keeps the answer for 24 hours. Same key, same request: The first answer again, with Idempotent-Replayed: true. Nothing is written a second time. Same key, another body or route: 422 idempotency_key_reused. Use a new key for each new write. Same key while the first request runs: 409 idempotency_in_progress. Send it again in a few seconds. The first request got an error: The API does not keep an error. Thus, you can send the key again. Use a new UUID for each write. A DELETE takes no key: a second delete gets 404.","keywords":"curl -X POST https://app.backend.genezio.ai/customer/v1/brands/12/scenarios \\ -H X-API-Key: $GENEZIO_API_KEY \\ -H Idempotency-Key: 6f1c2a52-6a4b-4b52-9d6e-3d1c0f3b8f11 \\ -H Content-Type: application/json \\ -d ' scenarios : topic_id : 3 text : Which headphones are best for a flight? '"},{"kind":"section","title":"Write data","heading":"Try a write first with a dry run","section":"API","crumbs":["Concepts","Write data","Try a write first with a dry run"],"url":"/docs/api/guides/writes#try-a-write-first-with-a-dry-run","text":"Each write of many items takes \"dry_run\": true. The API checks each item and gives the answer, but it writes nothing:","keywords":"data : rejected : index : 1 code : not_found message : Brand 12 has no topic 99. info : dry_run : true written : 2 rejected : 1"},{"kind":"section","title":"Write data","heading":"Writes of many items and rejected items","section":"API","crumbs":["Concepts","Write data","Writes of many items and rejected items"],"url":"/docs/api/guides/writes#writes-of-many-items-and-rejected-items","text":"A write of many items writes each item on its own. An item that is not valid is in rejected, with its index in your request and a code, and the other items are written: The plan of the account is checked for the items together. When the plan cannot hold them, the API writes nothing and answers 403 plan_limit_reached.","keywords":"data : id : 812 topic_id : 3 text : Which headphones are best for a flight? status : active created_at : 2026-10-06T09:12:44Z rejected : index : 1 code : conflict message : Topic 3 already has this scenario. info : dry_run : false written : 1 rejected : 1"},{"kind":"section","title":"Write data","heading":"Create a topic with its scenarios","section":"API","crumbs":["Concepts","Write data","Create a topic with its scenarios"],"url":"/docs/api/guides/writes#create-a-topic-with-its-scenarios","text":"One request makes a topic and its scenarios: A topic with the name, the type and the language of another topic of the brand is rejected with conflict. A topic that the plan cannot hold is rejected with plan_limit_reached, and the topics before it stay written.","keywords":"topics : name : Headphones for travel type : prompter language : english answer_engines : chatgpt.com perplexity scenarios : Which headphones are best for a long flight? Are noise-cancelling earbuds good for a train?"},{"kind":"section","title":"Write data","heading":"Change only what you send with PATCH","section":"API","crumbs":["Concepts","Write data","Change only what you send with PATCH"],"url":"/docs/api/guides/writes#change-only-what-you-send-with-patch","text":"A PATCH changes only the fields of its body. The other fields keep their values. For the engines of a topic: not in the body: The engines do not change. [\"chatgpt.com\"]: The topic uses these engines. An engine that the plan does not hold gets 403 plan_limit_reached. []: The topic uses the engines of the brand again.","keywords":""},{"kind":"section","title":"Write data","heading":"What you can write with the API","section":"API","crumbs":["Concepts","Write data","What you can write with the API"],"url":"/docs/api/guides/writes#what-you-can-write-with-the-api","text":"Brand: Change the websites and the answer engines. GET /brands/{brand_id}/models gives each engine, in the plan or not. Topics: Create many, each with its scenarios. Change one, pause or activate many, delete, restore. Scenarios: Create many, change one, pause or activate many, delete, restore. Competitors: Track many, change one, stop the tracking of one, delete. A name of a suggestion of the platform tracks the suggestion, with its history. Brand groups: Create, change, give the whole list of members with one PUT, delete. A competitor is in one group at most: a second group gets 409. Tags: Create, change, delete, and give a topic its tags with one PUT. Each topic gives its tags in tags. Personas: Create, change, delete. Give a persona to a topic with persona_id. A persona that a topic uses gets 409 on a delete. Tracked URLs: Track many pages (a page with no title gets the title of the page), change one, stop the tracking of one. Put the pages in tracked URL groups: create, rename, delete. An address that the brand tracks gets conflict. Knowledge bases: Create from a text, a site or files, crawl a site again, stop the indexing, delete. See below: these writes cost. Scorecard: Change a goal, archive a goal, restore it. A new scenario runs with the next run of its topic. A write does not start a run, thus it costs nothing. The API reads the articles, the briefs and the templates, but it does not write them. The dashboard writes them with a model. A tracked URL group is a folder for the tracked URLs of a brand, such as \"Blog\" or \"Q4 campaign\". It is not a brand group. A brand group joins competitors into one brand.","keywords":""},{"kind":"section","title":"Write data","heading":"Writes that have a cost","section":"API","crumbs":["Concepts","Write data","Writes that have a cost"],"url":"/docs/api/guides/writes#writes-that-have-a-cost","text":"Most writes only change the setup, thus they cost nothing. These writes start work that costs, as the same action in the dashboard does: Create a knowledge base: An embedding of each chunk. A site also costs a crawl of each page, up to max_pages. Crawl a knowledge base again: A crawl of each page and an embedding of each chunk. A schedule does it by itself. Search the knowledge bases: One embedding of the query. The plan of the account limits the knowledge bases of a brand. A write over the limit gets 403 plan_limit_reached, and nothing is written.","keywords":""},{"kind":"page","title":"Visibility and recommendation reports","heading":"","section":"API","crumbs":["API","Reports","Visibility and recommendation reports"],"url":"/docs/api/reports/visibility","text":"The brand visibility API of Genezio tells you how often answer engines mention and recommend your brand, with confidence intervals, ranks and per-engine series. The visibility and recommendation reports of the Genezio API measure how often answer engines mention and recommend your brand. Two reports read the presence of a brand in the answers of the engines: Visibility: GET /customer/v1/brands/{brand_id}/reports/visibility visibility: the share of the answers that mention the brand. Recommendation: GET /customer/v1/brands/{brand_id}/reports/recommendation recommendation: the share of the answers that recommend the brand. The two reports have the same shape. Only the name of the value changes.","keywords":""},{"kind":"section","title":"Visibility and recommendation reports","heading":"A row of the visibility report","section":"API","crumbs":["Reports","Visibility and recommendation reports","A row of the visibility report"],"url":"/docs/api/reports/visibility#a-row-of-the-visibility-report","text":"Each row is one brand in one period: your brand and, with competitors=true (the default), each competitor. visibility: From 0 to 1. 0.565 means that 56.5% of the answers mention the brand. confidence_interval: The true value is in this interval, at the confidence level of your brand. change: The value minus the value of the previous period of the same brand. significant: true when the intervals of the two periods do not overlap: the change is real and not noise. share_of_voice: From 0 to 1: the part of this brand among each brand of the period. conversations: The answers that the value reads.","keywords":"date : 2026-09-28 brand : name : Acme website : https://acme.com is_own_brand : true model : null visibility : 0.565 confidence_interval : lower : 0.552 upper : 0.578 change : 0.031 significant : true share_of_voice : 0.091 conversations : 24074"},{"kind":"section","title":"Visibility and recommendation reports","heading":"The summary and the rank of your brand","section":"API","crumbs":["Reports","Visibility and recommendation reports","The summary and the rank of your brand"],"url":"/docs/api/reports/visibility#the-summary-and-the-rank-of-your-brand","text":"info.summary gives the last period of your brand, and its place among each brand of that period:","keywords":"summary : model : null date : 2026-10-05 visibility : 0.565 change : 0.031 significant : true rank : 2 brands_ranked : 17"},{"kind":"section","title":"Visibility and recommendation reports","heading":"One series for each answer engine","section":"API","crumbs":["Reports","Visibility and recommendation reports","One series for each answer engine"],"url":"/docs/api/reports/visibility#one-series-for-each-answer-engine","text":"Add group_by=model for one series for each answer engine. Each row then has model, and info.summary has one item for each engine.","keywords":"curl https://app.backend.genezio.ai/customer/v1/brands/12/reports/visibility?granularity=week&group_by=model&competitors=false \\ -H X-API-Key: $GENEZIO_API_KEY"},{"kind":"section","title":"Visibility and recommendation reports","heading":"Filters of the visibility reports","section":"API","crumbs":["Reports","Visibility and recommendation reports","Filters of the visibility reports"],"url":"/docs/api/reports/visibility#filters-of-the-visibility-reports","text":"start_date, end_date: The range. See Date ranges. granularity: day, week or month. models: The answer engines, such as chatgpt.com, perplexity, claude. With no list, the engines of the brand. topic_ids: The topics of the brand. A topic of another brand gets 400. group_by: model, for one series for each engine. competitors: false gives the rows of your brand only. The rank still reads each brand.","keywords":""},{"kind":"section","title":"Visibility and recommendation reports","heading":"Visibility and recommendation by topic","section":"API","crumbs":["Reports","Visibility and recommendation reports","Visibility and recommendation by topic"],"url":"/docs/api/reports/visibility#visibility-and-recommendation-by-topic","text":"Two more routes give one value for each topic, and for each scenario of the topic, in place of one value for each period: Visibility by topic: GET /customer/v1/brands/{brand_id}/reports/visibility/topics. Recommendation by topic: GET /customer/v1/brands/{brand_id}/reports/recommendation/topics These are the numbers of the Topics page of the dashboard. Each row is one topic that has answers in the range: visibility: Your brand in the topic, from 0 to 1. 0 when the answers mention only competitors. null when the topic has nothing to read. days: The length of the window of days that the value reads. The window ends on the last day with answers. change: The value minus the value of the window that ends on the previous day with answers. brands: Your brand and the five competitors with the highest value. scenarios: The same values for each scenario of the topic. A scenario with no answers has null. The filters are start_date, end_date, models and topic_ids, as above. Add scenarios=false to get the topics only. That report is faster.","keywords":"topic_id : 1 name : Best noise-cancelling headphones visibility : 0.549 confidence_interval : lower : 0.452 upper : 0.647 change : 0.037 share_of_voice : 0.109 conversations : 71 days : 12 brands : name : Acme is_own_brand : true visibility : 0.549 change : 0.037 ... : ... name : Rival is_own_brand : false visibility : 0.845 change : -0.019 ... : ... scenarios : scenario_id : 1 text : What are the best noise-cancelling headphones? visibility : 0.542 ... : ..."},{"kind":"section","title":"Visibility and recommendation reports","heading":"The report of one scenario, day by day","section":"API","crumbs":["Reports","Visibility and recommendation reports","The report of one scenario, day by day"],"url":"/docs/api/reports/visibility#the-report-of-one-scenario-day-by-day","text":"Two routes give the report of one scenario, with one row for each day and each brand: Visibility of a scenario: GET /customer/v1/brands/{brand_id}/scenarios/{scenario_id}/reports/visibility. Recommendation of a scenario: GET /customer/v1/brands/{brand_id}/scenarios/{scenario_id}/reports/recommendation The rows and info.summary have the shape of the report of the brand, thus the same code reads the two reports. change and significant compare a day with the previous day with answers. The filters are start_date, end_date, models and competitors. A scenario has one period for each day, thus these routes have no granularity, no group_by and no topic_ids. A scenario of another brand gets 404.","keywords":"curl https://app.backend.genezio.ai/customer/v1/brands/12/scenarios/420/reports/visibility?start_date=2026-09-01&competitors=false \\ -H X-API-Key: $GENEZIO_API_KEY"},{"kind":"section","title":"Visibility and recommendation reports","heading":"How the platform calculates the value of a period","section":"API","crumbs":["Reports","Visibility and recommendation reports","How the platform calculates the value of a period"],"url":"/docs/api/reports/visibility#how-the-platform-calculates-the-value-of-a-period","text":"The value of a day is not the share of the answers of that day only. The platform reads a sliding window: the newest answers up to the end of that day. The dashboard shows the same values.","keywords":""},{"kind":"section","title":"Visibility and recommendation reports","heading":"The sliding window of one day","section":"API","crumbs":["Reports","Visibility and recommendation reports","The sliding window of one day"],"url":"/docs/api/reports/visibility#the-sliding-window-of-one-day","text":"For each day with answers, the platform does these steps: 1. For each scenario and each answer engine, the platform starts at the newest answer of the day. Then it adds older answers, one at a time. 2. It stops when the confidence interval is narrow enough, or when the range has no older answer. The confidence level and the maximum width of the interval are settings of your brand. 3. The value of the scenario on the engine is the share of the answers in the window that mention the brand. 4. The value of the day is the average of the engines of each scenario, and then the average of the scenarios. A scenario with many answers each day gets a short window, of one or two days. A scenario with few answers gets a longer window.","keywords":""},{"kind":"section","title":"Visibility and recommendation reports","heading":"An example of a sliding window","section":"API","crumbs":["Reports","Visibility and recommendation reports","An example of a sliding window"],"url":"/docs/api/reports/visibility#an-example-of-a-sliding-window","text":"A scenario runs on ChatGPT once each day, with 8 answers each time. The window needs 24 answers for a narrow interval. 10 September: The answers of 8, 9 and 10 September 12 of 24 0.5. 11 September: The answers of 9, 10 and 11 September 15 of 24 0.625 The two windows share the answers of 9 and 10 September. Thus the value changes slowly from one day to the next day. One bad day does not make the value fall to zero.","keywords":""},{"kind":"section","title":"Visibility and recommendation reports","heading":"Weeks and months","section":"API","crumbs":["Reports","Visibility and recommendation reports","Weeks and months"],"url":"/docs/api/reports/visibility#weeks-and-months","text":"day: The sliding window of the day. week: The average of the sliding windows of each day with answers in the week, from Monday to Sunday. month: The average of the sliding windows of each day with answers in the last week of the month. The earlier weeks of the month do not change the value. conversations gives the answers of the sliding window of the last day of the period. A day with no answers has no row.","keywords":""},{"kind":"section","title":"Visibility and recommendation reports","heading":"The first days of a range","section":"API","crumbs":["Reports","Visibility and recommendation reports","The first days of a range"],"url":"/docs/api/reports/visibility#the-first-days-of-a-range","text":"The sliding window does not read an answer before start_date. Thus the first days of a range read fewer answers, and their intervals are wider. To get a full window for each day that you need, start the range some days earlier. To alert on a change, use significant and not a fixed threshold. A drop of 5 points can be noise on a small brand and real on a large one.","keywords":""},{"kind":"page","title":"Citations report","heading":"","section":"API","crumbs":["API","Reports","Citations report"],"url":"/docs/api/reports/citations","text":"The citations API of Genezio gives the domains and the pages that answer engines cite for your topics, with their owners, content types and sentiment. The citations routes of the Genezio API give the pages that answer engines cite when they speak of your topics. An answer engine gives the pages that it read as the sources of its answer. These routes read those pages. GET /customer/v1/brands/{brand_id}/reports/citations: The occurrences period by period, and the share of each kind of owner. GET /customer/v1/brands/{brand_id}/citations/domains: The domains, the most frequent first. GET /customer/v1/brands/{brand_id}/citations/domains/{domain}: One domain: over time, by topic, by scenario and by engine. GET /customer/v1/brands/{brand_id}/citations/urls: The pages, the most frequent or the newest first.","keywords":""},{"kind":"section","title":"Citations report","heading":"Two counts of citations","section":"API","crumbs":["Reports","Citations report","Two counts of citations"],"url":"/docs/api/reports/citations#two-counts-of-citations","text":"occurrences: The answers that list the page as a source. cited_occurrences: The answers whose text also quotes the page. A page can be a source and not be in the text.","keywords":""},{"kind":"section","title":"Citations report","heading":"The citations report","section":"API","crumbs":["Reports","Citations report","The citations report"],"url":"/docs/api/reports/citations#the-citations-report","text":"- Each row with \"domain\": null counts each domain together. - Each domain of domains gets its own series. - info.summary.by_ownership gives the share of each kind of owner of the pages, such as first_party (your pages), competitor_owned and third_party_editorial.","keywords":"curl https://app.backend.genezio.ai/customer/v1/brands/12/reports/citations?granularity=week&domains=g2.com&domains=reddit.com \\ -H X-API-Key: $GENEZIO_API_KEY"},{"kind":"section","title":"Citations report","heading":"Filters of the citations routes","section":"API","crumbs":["Reports","Citations report","Filters of the citations routes"],"url":"/docs/api/reports/citations#filters-of-the-citations-routes","text":"Each route of the citations takes these filters: start_date, end_date: The range. See Date ranges. models, topic_ids: The engines and the topics. ownership: Who owns the page: first_party, competitor_owned, third_party_editorial, industry_trade_publication, government_regulatory, academic_research, aggregator_marketplace, social_ugc_platform, other. content_types: What the page is, such as review_editorial, listicle_best_of, comparison_vs_page, forum_thread_qa. sentiment: positive, neutral or negative: how the page speaks of your brand. first_party: true for your pages only, false for the pages of others. Find the pages that work against you: citations/urls?sentiment=negative&first_party=false&sort=occurrences.","keywords":""},{"kind":"page","title":"Sentiment and perceptions reports","heading":"","section":"API","crumbs":["API","Reports","Sentiment and perceptions reports"],"url":"/docs/api/reports/sentiment","text":"The sentiment API of Genezio gives what answer engines say about your brand: the sentiment by period and by topic, and the fact check of each perception. The sentiment and perception routes of the Genezio API give what answer engines say about your brand. The platform keeps each statement of the answers about your brand, with its sentiment, and it groups the statements that say the same thing into one perception. GET /customer/v1/brands/{brand_id}/reports/sentiment: The statements by sentiment, period by period. GET /customer/v1/brands/{brand_id}/reports/sentiment/topics: The sentiment of each topic and of each scenario. GET /customer/v1/brands/{brand_id}/perceptions: Each perception, with the check of its facts.","keywords":""},{"kind":"section","title":"Sentiment and perceptions reports","heading":"The sentiment report by period","section":"API","crumbs":["Reports","Sentiment and perceptions reports","The sentiment report by period"],"url":"/docs/api/reports/sentiment#the-sentiment-report-by-period","text":"Each row gives the statements of one period as counts and as shares: info.summary.range gives the shares of the whole range, and info.summary.last_period gives the last period.","keywords":"date : 2026-09-29 positive : 16450 neutral : 32888 negative : 17027 total : 66365 positive_share : 0.2479 neutral_share : 0.4956 negative_share : 0.2566 positive_share_change : 0.0015"},{"kind":"section","title":"Sentiment and perceptions reports","heading":"The sentiment report by topic","section":"API","crumbs":["Reports","Sentiment and perceptions reports","The sentiment report by topic"],"url":"/docs/api/reports/sentiment#the-sentiment-report-by-topic","text":"Each topic gives two kinds of sentiment: - statements: the sentiment of what the answers say. - citations: the sentiment of the pages that the answers cite. Each kind has positive, neutral, negative, competitor_only (about a competitor and not about your brand) and none_mentioned (about no brand).","keywords":""},{"kind":"section","title":"Sentiment and perceptions reports","heading":"The perceptions of your brand","section":"API","crumbs":["Reports","Sentiment and perceptions reports","The perceptions of your brand"],"url":"/docs/api/reports/sentiment#the-perceptions-of-your-brand","text":"statements: The statements of the answers that say this thing. fact_check: confirmed: the facts of your brand support it. contradicted: they contradict it. unconfirmed: not checked yet. in_knowledge_base: Whether your knowledge base holds it. null when not checked yet. Filter with sentiment, fact_check, in_knowledge_base and search. The perceptions to fix first: perceptions?fact_check=contradicted&in_knowledge_base=true. The answers say a thing that your own knowledge base says is false.","keywords":"id : 1000001 text : Answers criticise Acme for battery life in working from home. sentiment : negative statements : 6233 fact_check : contradicted in_knowledge_base : false"},{"kind":"page","title":"Fact check report","heading":"","section":"API","crumbs":["API","Reports","Fact check report"],"url":"/docs/api/reports/fact-check","text":"The fact check API of Genezio tells you how often answer engines give the correct facts about your brand, with the pass rate by period, by topic and by engine. The fact check reports of the Genezio API measure how often answer engines give the correct facts about your brand. A topic of the type fact_checker asks the engines about facts of your brand, such as a price, a feature or a policy. Each scenario has the expected answer, thus each answer passes or fails. Over time: GET /customer/v1/brands/{brand_id}/reports/fact-check. By topic: GET /customer/v1/brands/{brand_id}/reports/fact-check/topics. By engine: GET /customer/v1/brands/{brand_id}/reports/fact-check/models.","keywords":""},{"kind":"section","title":"Fact check report","heading":"A row of the fact check report","section":"API","crumbs":["Reports","Fact check report","A row of the fact check report"],"url":"/docs/api/reports/fact-check#a-row-of-the-fact-check-report","text":"pass_rate: From 0 to 1: passed divided by total. null when there is no answer. confidence_interval: The Wilson interval of pass_rate, at 90%. change: Over time only: the rate minus the rate of the previous period. significant: Over time only: true when the intervals of the two periods do not overlap.","keywords":"date : 2026-09-28 passed : 41 failed : 9 total : 50 pass_rate : 0.82 confidence_interval : lower : 0.72 upper : 0.89 change : 0.11 significant : true"},{"kind":"section","title":"Fact check report","heading":"Find the facts to correct","section":"API","crumbs":["Reports","Fact check report","Find the facts to correct"],"url":"/docs/api/reports/fact-check#find-the-facts-to-correct","text":"The report by topic gives the scenarios of each topic with the lowest rate first, and the report by engine gives the engine with the lowest rate first. Thus, the first row tells you which fact, or which engine, to correct first.","keywords":""},{"kind":"section","title":"Fact check report","heading":"Filters of the fact check reports","section":"API","crumbs":["Reports","Fact check report","Filters of the fact check reports"],"url":"/docs/api/reports/fact-check#filters-of-the-fact-check-reports","text":"start_date, end_date, models and topic_ids, as in the other reports. The report over time also takes granularity: day, week or month.","keywords":""},{"kind":"page","title":"Products report","heading":"","section":"API","crumbs":["API","Reports","Products report"],"url":"/docs/api/reports/products","text":"The products API of Genezio gives the products that answer engines show to shoppers, their visibility and recommendation, their retailers and the UCP audit. The products routes of the Genezio API give the products that answer engines show to a person who wants to buy, and the retailers that sell them. For the dashboard view, refer to Product Visibility. When a person asks to buy something, an answer engine can show a carousel of products. The platform reads each product of the carousel. A product is the brand's (label brand), a competitor's (label competitor), or neither (label neutral). A product has variants (SKUs), such as a color or a size, and offers of retailers. Which of my products do the engines show?: GET /customer/v1/brands/{brand_id}/products. Which product is the strongest, or the weakest?: GET../products/ranking (order=asc for the weakest). How does each product change over time?: GET../reports/products. How many products do I and each competitor have?: GET../products/stats. What does one product look like in the answers?: GET../products/{product_id}. Which variant of a product do the engines show?: GET../products/{product_id}/skus. Which products compete with one product?: GET../products/{product_id}/co-mentions and /series. What do the answers say about a product?: GET../products/{product_id}/perception. Which pages do the answers read about it?: GET../products/{product_id}/sources. Which retailers sell in the answers?: GET../products/retailers.","keywords":""},{"kind":"section","title":"Products report","heading":"The values of the products report","section":"API","crumbs":["Reports","Products report","The values of the products report"],"url":"/docs/api/reports/products#the-values-of-the-products-report","text":"Each value is from 0 to 1, as in the other reports: visibility: The share of the shopping answers that show the product. recommendation: The share of the shopping answers that recommend it. share_of_voice: Its part of each product of the same answers. confidence_interval: The true value is in this interval. With no models, a report reads the engines of the brand. The filters are start_date, end_date, models and topic_ids, as in the other reports.","keywords":""},{"kind":"section","title":"Products report","heading":"Can an AI agent buy on your site?","section":"API","crumbs":["Reports","Products report","Can an AI agent buy on your site?"],"url":"/docs/api/reports/products#can-an-ai-agent-buy-on-your-site","text":"An agent of an answer engine can buy on a site that declares its catalog and its checkout in a manifest at /.well-known/ucp (the Universal Commerce Protocol). For the concept, refer to Agentic Commerce Readiness. The e-commerce audit reads that manifest and gives a score from 0 to 1: What is my score, and which rules fail?: GET../ecommerce-audit. How does my score change?: GET../ecommerce-audit/history. What are the scores of my competitors?: GET../ecommerce-audit/competitors. Audit my site again now.: POST../ecommerce-audit/runs The audit reads the sites only, thus it costs no run of an engine.","keywords":""},{"kind":"page","title":"Query fanouts report","heading":"","section":"API","crumbs":["API","Reports","Query fanouts report"],"url":"/docs/api/reports/fanouts","text":"The query fanouts API of Genezio gives the web searches that answer engines run before they answer, the pages that each search finds, and their relevance. A query fanout is a web search that an answer engine runs before it answers. The fanout routes of the Genezio API give these searches. To answer a scenario, an engine first searches the web, such as best headphones for flights 2026. Each search is a fanout. The pages that a search finds are the pages that the answer can cite, thus the searches tell you which words bring the engines to your pages. For the concept, refer to Query Fanouts. Which searches do the engines run for my scenarios?: GET /customer/v1/brands/{brand_id}/fanouts. Which searches find a page or a domain that the answers cite?: GET../fanouts/cited?url=.. or?domain=.. Which words do many searches share?: GET../fanouts/phrases. Which searches hold one of these phrases?: GET../fanouts/by-phrase?phrase=.. Which answers ran one search?: GET../fanouts/{fanout_id}/conversations. Which pages did one search find?: GET../fanouts/{fanout_id}/sources relevance gives, from 0 to 1, the share of the answers of a search that name your brand or a tracked competitor. A search with a high count and a low relevance is a search where no brand of your market appears yet. With no models, a route reads the engines of the brand. With no dates, the last 30 days.","keywords":""},{"kind":"page","title":"API reference","heading":"","section":"API","crumbs":["API","API reference","API reference"],"url":"/docs/api-reference/introduction","text":"The reference of each route of the Genezio API: brands, topics, scenarios, reports and content, with a playground that sends a real request. https://app.backend.genezio.ai/customer/v1 Send your key in the X-API-Key header. 200 requests in one minute for each key.","keywords":""},{"kind":"section","title":"API reference","heading":"Setup","section":"API","crumbs":["API reference","API reference","Setup"],"url":"/docs/api-reference/introduction#setup","text":"Read and change what the platform measures. A route that changes data needs a key with the write scope. See Write data. The brands of the account, their websites and their answer engines. Create topics with their scenarios in one request. The scenarios: create, pause, change and delete many at once. Track competitors, or the suggestions of the platform. Group topics, such as by campaign. Who asks the scenarios, and from where.","keywords":""},{"kind":"section","title":"API reference","heading":"Analytics","section":"API","crumbs":["API reference","API reference","Analytics"],"url":"/docs/api-reference/introduction#analytics","text":"Read the numbers. Each route only reads, thus a key with the read scope can send it. Visibility and recommendation, by period, topic and scenario. The sources that the answers cite, by domain and by page. The sentiment of the answers, by period and by topic. What the answers say about the brand, with the check of its facts. Each answer of an engine: its text, the brands in order, and its sources.","keywords":""},{"kind":"section","title":"API reference","heading":"Use the playground","section":"API","crumbs":["API reference","API reference","Use the playground"],"url":"/docs/api-reference/introduction#use-the-playground","text":"The playground sends each request to the Genezio API at https://app.backend.genezio.ai. In the dashboard, select API Keys in the menu of the account. Open a route, paste the key in X-API-Key, and select Send. Read Read a response, Date ranges and Errors first. Each route follows those rules.","keywords":""},{"kind":"page","title":"MCP server","heading":"","section":"MCP","crumbs":["MCP","MCP server","MCP server"],"url":"/docs/mcp","text":"Connect Claude, ChatGPT, Cursor or another MCP client to Genezio, and ask about the AI visibility, citations and competitors of your brands. The Genezio MCP server connects an AI assistant to your Genezio data with the Model Context Protocol (MCP). After you connect it, you can ask Claude, ChatGPT, Cursor or another MCP client about the visibility, the citations and the competitors of your brands, in your own words.","keywords":""},{"kind":"section","title":"MCP server","heading":"What you can do with the Genezio MCP server","section":"MCP","crumbs":["MCP server","MCP server","What you can do with the Genezio MCP server"],"url":"/docs/mcp#what-you-can-do-with-the-genezio-mcp-server","text":"- Read your AI visibility. Ask for the AI Visibility %, the AI Recommendations % and the share of voice of a brand, by topic, by scenario and by answer engine. - Find the cause of a change. Ask why a metric dropped. The assistant reads the conversations, the citations and the query fanouts that are behind the number. - Study competitors. Read the presence of each competitor by answer engine, and their SWOT analysis. - Read the citations. See the domains and the pages that answer engines cite, and the statements that they take from each source. - Change the setup. With the write scope, create topics, scenarios, personas and competitors, and change them. - Use the Business Scorecard. Read and build the goals of a brand. Refer to Example questions for questions that work well, and to the MCP tools reference for each tool.","keywords":""},{"kind":"section","title":"MCP server","heading":"The MCP server URL","section":"MCP","crumbs":["MCP server","MCP server","The MCP server URL"],"url":"/docs/mcp#the-mcp-server-url","text":"Use this URL in your MCP client: The server uses the Streamable HTTP transport. It does not use the old SSE transport.","keywords":"https://app.backend.genezio.ai/mcp"},{"kind":"section","title":"MCP server","heading":"Requirements","section":"MCP","crumbs":["MCP server","MCP server","Requirements"],"url":"/docs/mcp#requirements","text":"- A Genezio user. The assistant sees only the brands that this user can open, with the same permissions as in the dashboard. - An active subscription on at least one account of the user. Without one, Genezio refuses the connection. - An MCP client that supports remote servers and OAuth, for example Claude, ChatGPT, Cursor, VS Code or Windsurf.","keywords":""},{"kind":"section","title":"MCP server","heading":"How the connection works","section":"MCP","crumbs":["MCP server","MCP server","How the connection works"],"url":"/docs/mcp#how-the-connection-works","text":"Add the MCP server URL to your client. Refer to Connect an MCP client. The client opens Genezio in your browser. Sign in, if you are not signed in. Genezio shows the name of the client and two permissions: Read and Write. Select the permissions, and click Allow access. Go back to the client, and ask a question about your brand.","keywords":""},{"kind":"section","title":"MCP server","heading":"MCP server or REST API","section":"MCP","crumbs":["MCP server","MCP server","MCP server or REST API"],"url":"/docs/mcp#mcp-server-or-rest-api","text":"Who uses it: A person, through an AI assistant A program. Authentication: OAuth, with your Genezio user An API key (gnz-..). How you ask: A question in your own words An HTTP request to a route. Best for: Analysis, reports and quick changes Dashboards, exports and automations An API key does not work with the MCP server, and an MCP token does not work with the REST API. Refer to the Genezio API for the REST API.","keywords":""},{"kind":"page","title":"Connect a client","heading":"","section":"MCP","crumbs":["MCP","MCP server","Connect a client"],"url":"/docs/mcp/connect","text":"Add the Genezio MCP server to Claude, Claude Code, ChatGPT, Cursor, VS Code, Windsurf or another MCP client, step by step. To connect an MCP client to Genezio, add the Genezio MCP server URL to the client, and then sign in to Genezio in your browser. Each client keeps its list of MCP servers in a different place. This page gives the procedure for the most frequent clients. The URL of the server is: Each client uses OAuth: you do not copy a token or an API key. After you add the server, the client opens Genezio, and you approve the access. Refer to Authentication and scopes.","keywords":"https://app.backend.genezio.ai/mcp"},{"kind":"section","title":"Connect a client","heading":"Claude (web and desktop)","section":"MCP","crumbs":["MCP server","Connect a client","Claude (web and desktop)"],"url":"/docs/mcp/connect#claude-web-and-desktop","text":"In Claude, open Settings, then Connectors. Click Add custom connector. Enter the name Genezio and the MCP server URL. Click Connect. Sign in to Genezio, and click Allow access. On a Team or Enterprise plan of Claude, an owner of the organization can add the connector for all the members.","keywords":""},{"kind":"section","title":"Connect a client","heading":"Claude Code","section":"MCP","crumbs":["MCP server","Connect a client","Claude Code"],"url":"/docs/mcp/connect#claude-code","text":"Add the server from the terminal: Then start Claude Code, enter /mcp, select genezio, and authenticate in the browser.","keywords":"claude mcp add --transport http genezio https://app.backend.genezio.ai/mcp"},{"kind":"section","title":"Connect a client","heading":"ChatGPT","section":"MCP","crumbs":["MCP server","Connect a client","ChatGPT"],"url":"/docs/mcp/connect#chatgpt","text":"In ChatGPT, open Settings, then the settings of the apps and connectors. A custom MCP connector can need the developer mode. Create a new connector. Enter the name Genezio and the MCP server URL, and select OAuth as the authentication. Sign in to Genezio, and click Allow access.","keywords":""},{"kind":"section","title":"Connect a client","heading":"Cursor","section":"MCP","crumbs":["MCP server","Connect a client","Cursor"],"url":"/docs/mcp/connect#cursor","text":"Add the server to ~/.cursor/mcp.json for all your projects, or to.cursor/mcp.json for one project: Open the settings of Cursor, find the genezio server in the MCP section, and click the button that connects it.","keywords":"mcpServers : genezio : url : https://app.backend.genezio.ai/mcp"},{"kind":"section","title":"Connect a client","heading":"VS Code","section":"MCP","crumbs":["MCP server","Connect a client","VS Code"],"url":"/docs/mcp/connect#vs-code","text":"Add the server to.vscode/mcp.json in your project: Start the server from the file, and authenticate in the browser when VS Code asks.","keywords":"servers : genezio : type : http url : https://app.backend.genezio.ai/mcp"},{"kind":"section","title":"Connect a client","heading":"Windsurf","section":"MCP","crumbs":["MCP server","Connect a client","Windsurf"],"url":"/docs/mcp/connect#windsurf","text":"Add the server to ~/.codeium/windsurf/mcp_config.json:","keywords":"mcpServers : genezio : serverUrl : https://app.backend.genezio.ai/mcp"},{"kind":"section","title":"Connect a client","heading":"Other MCP clients","section":"MCP","crumbs":["MCP server","Connect a client","Other MCP clients"],"url":"/docs/mcp/connect#other-mcp-clients","text":"A client can connect to Genezio when it supports these items: - A remote MCP server with the Streamable HTTP transport. - OAuth 2.1 with PKCE and dynamic client registration. The client finds the OAuth endpoints by itself, from the 401 answer of the server and from /.well-known/oauth-protected-resource. You only give it the MCP server URL.","keywords":""},{"kind":"section","title":"Connect a client","heading":"Troubleshooting the connection","section":"MCP","crumbs":["MCP server","Connect a client","Troubleshooting the connection"],"url":"/docs/mcp/connect#troubleshooting-the-connection","text":"Genezio gives the MCP access only to a user who has an active subscription on at least one account. Ask the owner of your account to check the subscription. The assistant sees only the brands that your Genezio user can open. Open the dashboard with the same user, and check that you can see the brand. The client has only the Read permission. Remove the server from the client, add it again, and select Write on the page of the permissions. An access token is valid for 12 hours, and the client gets a new one by itself. If the client does not use the server for 30 days, you must sign in again.","keywords":""},{"kind":"page","title":"Authentication and scopes","heading":"","section":"MCP","crumbs":["MCP","MCP server","Authentication and scopes"],"url":"/docs/mcp/authentication","text":"How the Genezio MCP server uses OAuth 2.1: the consent page, the read and write scopes, the lifetime of the tokens, and the endpoints. The Genezio MCP server uses OAuth 2.1 to authenticate an MCP client. The client acts as your Genezio user, with the permissions that you give it on the consent page: Read, Write, or both.","keywords":""},{"kind":"section","title":"Authentication and scopes","heading":"The consent page","section":"MCP","crumbs":["MCP server","Authentication and scopes","The consent page"],"url":"/docs/mcp/authentication#the-consent-page","text":"When a client connects, Genezio opens the page Connect client to Genezio in your browser. The page shows the name of the client and two permissions: Read: mcp:read Read the brands, the competitors, the citations, the query fanouts, the insights and the other data. Write: mcp:write Create and change scenarios, topics, personas, competitors and the other setup. You can clear a permission that the client asks for, but one permission must stay selected. Click Allow access to connect, or Deny to refuse.","keywords":""},{"kind":"section","title":"Authentication and scopes","heading":"What each scope permits","section":"MCP","crumbs":["MCP server","Authentication and scopes","What each scope permits"],"url":"/docs/mcp/authentication#what-each-scope-permits","text":"- A tool that only reads needs mcp:read. - A tool that changes data needs mcp:write. - A client does not see the tools that its scopes do not permit. If it calls one, Genezio tells it to connect again with the required permission. - The scopes do not give access to more brands. The client sees only the brands that your Genezio user can open. The MCP tools reference gives the scope of each tool.","keywords":""},{"kind":"section","title":"Authentication and scopes","heading":"Token lifetime","section":"MCP","crumbs":["MCP server","Authentication and scopes","Token lifetime"],"url":"/docs/mcp/authentication#token-lifetime","text":"Authorization code: 10 minutes. Access token: 12 hours. Refresh token: 30 days. Each use gives a new refresh token. The client refreshes the access token by itself. After 30 days with no use, you sign in again.","keywords":""},{"kind":"section","title":"Authentication and scopes","heading":"OAuth endpoints","section":"MCP","crumbs":["MCP server","Authentication and scopes","OAuth endpoints"],"url":"/docs/mcp/authentication#oauth-endpoints","text":"The client finds these endpoints by itself. You need them only to write your own client. Protected resource metadata: /.well-known/oauth-protected-resource. Authorization server metadata: /.well-known/oauth-authorization-server. Dynamic client registration: POST /oauth/register. Authorization: GET /oauth/authorize. Token: POST /oauth/token Each path is on the host of the MCP server. The server supports the authorization_code and refresh_token grants, with PKCE (S256). A public client has no client secret.","keywords":""},{"kind":"section","title":"Authentication and scopes","heading":"API keys and the MCP server","section":"MCP","crumbs":["MCP server","Authentication and scopes","API keys and the MCP server"],"url":"/docs/mcp/authentication#api-keys-and-the-mcp-server","text":"An API key of Genezio (gnz-...) works only with the REST API. The MCP server does not accept it. Use OAuth to connect an MCP client.","keywords":""},{"kind":"page","title":"Example questions","heading":"","section":"MCP","crumbs":["MCP","MCP server","Example questions"],"url":"/docs/mcp/examples","text":"Questions to ask Claude, ChatGPT or Cursor through the Genezio MCP server: visibility, competitors, citations, content and setup. After you connect an MCP client to Genezio, you ask questions in your own words. The assistant selects the Genezio tools that answer them. These example questions show what the Genezio MCP server can do. Replace the names with the names of your brand, topics and competitors.","keywords":""},{"kind":"section","title":"Example questions","heading":"Start with a brief","section":"MCP","crumbs":["MCP server","Example questions","Start with a brief"],"url":"/docs/mcp/examples#start-with-a-brief","text":"The prompt genezio_exec_brief gives a short executive brief of one brand, with the evidence and the next step. In most clients, enter /genezio_exec_brief and select the brand.","keywords":""},{"kind":"section","title":"Example questions","heading":"Questions about AI visibility","section":"MCP","crumbs":["MCP server","Example questions","Questions about AI visibility"],"url":"/docs/mcp/examples#questions-about-ai-visibility","text":"- \"What is the AI Visibility % of Acme in the last 30 days, by answer engine?\" - \"In which topics did the recommendation rate of Acme drop this month?\" - \"Which scenarios mention Acme the least on Perplexity?\" - \"Compare the share of voice of Acme and its competitors on ChatGPT.\"","keywords":""},{"kind":"section","title":"Example questions","heading":"Questions about competitors","section":"MCP","crumbs":["MCP server","Example questions","Questions about competitors"],"url":"/docs/mcp/examples#questions-about-competitors","text":"- \"Which competitor gains the most visibility in the topic CRM for startups?\" - \"Show the SWOT analysis of Acme against Globex.\" - \"On which answer engine is Globex stronger than Acme?\"","keywords":""},{"kind":"section","title":"Example questions","heading":"Questions about citations and sources","section":"MCP","crumbs":["MCP server","Example questions","Questions about citations and sources"],"url":"/docs/mcp/examples#questions-about-citations-and-sources","text":"- \"Which domains do the answer engines cite the most for Acme?\" - \"Which pages of acme.com get cited, and in which answers?\" - \"Which query fanouts lead the answer engines to the pages of our competitors?\" - \"What do the answers say about Acme that comes from reddit.com?\"","keywords":""},{"kind":"section","title":"Example questions","heading":"Questions about perceptions and facts","section":"MCP","crumbs":["MCP server","Example questions","Questions about perceptions and facts"],"url":"/docs/mcp/examples#questions-about-perceptions-and-facts","text":"- \"What do the answer engines say about the pricing of Acme?\" - \"Which claims about Acme do the answer engines contradict?\" - \"Show the fact check results of Acme by answer engine.\"","keywords":""},{"kind":"section","title":"Example questions","heading":"Questions about shopping and products","section":"MCP","crumbs":["MCP server","Example questions","Questions about shopping and products"],"url":"/docs/mcp/examples#questions-about-shopping-and-products","text":"- \"Which products of Acme do the shopping answers show the most?\" - \"Which retailers offer our products in the AI shopping answers?\" - \"Which products of competitors appear next to our best product?\"","keywords":""},{"kind":"section","title":"Example questions","heading":"Changes to the setup","section":"MCP","crumbs":["MCP server","Example questions","Changes to the setup"],"url":"/docs/mcp/examples#changes-to-the-setup","text":"These questions need the Write permission. Refer to Authentication and scopes. - \"Create a topic CRM for agencies with three scenarios for a marketing agency owner.\" - \"Add Globex as a competitor of Acme, with the website globex.com.\" - \"Pause the scenarios of the topic Enterprise CRM.\" - \"Create a persona for a CFO in a mid-market company in Germany.\" Ask for the evidence: \"show the conversations behind this number\". The assistant then reads the answers of the engines, and you can check each conclusion.","keywords":""},{"kind":"page","title":"Scorecards over MCP","heading":"","section":"MCP","crumbs":["MCP","MCP server","Scorecards over MCP"],"url":"/docs/mcp/scorecards-over-mcp","text":"Create, read and change Business Scorecard goals from Claude or another MCP client. The same agent as in the dashboard builds each goal from your words. You can create and change Business Scorecard goals from an MCP client. You do not have to open the board.","keywords":""},{"kind":"section","title":"Scorecards over MCP","heading":"Why you use an MCP client for goals","section":"MCP","crumbs":["MCP server","Scorecards over MCP","Why you use an MCP client for goals"],"url":"/docs/mcp/scorecards-over-mcp#why-you-use-an-mcp-client-for-goals","text":"The board is the correct place to read goals. But you usually decide on a goal during a different conversation. For example, a planning session, a quarterly review, or a competitor analysis that you already do in your AI client. When you create the goal where you make the decision, the goal exists before people forget it. You say: Add a goal: be recommended for \"project management for agencies\" on ChatGPT and Perplexity, UK only, measured against Asana and Monday. Then the goal is on the board, and Genezio tracks it from that day.","keywords":""},{"kind":"section","title":"Scorecards over MCP","heading":"What occurs","section":"MCP","crumbs":["MCP server","Scorecards over MCP","What occurs"],"url":"/docs/mcp/scorecards-over-mcp#what-occurs","text":"The agent that writes goals in the dashboard also writes them here. It reads your brand, decides what to measure, and builds the goal. It tells you in words what it will track. Then you accept the goal or continue to change it, as in the conversation on the board. A goal that you create from an MCP client is a usual goal: - It shows on the board. - Genezio measures it every day. - You can open its card and follow the number back to the conversations behind it. - You can tag it, change its position, or archive it, as any other goal.","keywords":""},{"kind":"section","title":"Scorecards over MCP","heading":"What you can do","section":"MCP","crumbs":["MCP server","Scorecards over MCP","What you can do"],"url":"/docs/mcp/scorecards-over-mcp#what-you-can-do","text":"- Create a goal. Describe the goal. - Read the board. See what each goal is and where it stands. - Change a goal. Make the market wider, replace a competitor, or change the answer engines. For the MCP tools of the Business Scorecard, refer to the MCP tools reference.","keywords":""},{"kind":"section","title":"Scorecards over MCP","heading":"Where to go next","section":"MCP","crumbs":["MCP server","Scorecards over MCP","Where to go next"],"url":"/docs/mcp/scorecards-over-mcp#where-to-go-next","text":"- Business Scorecard: what goals are and why they exist. - Creating a Goal: create a goal from the dashboard.","keywords":""},{"kind":"page","title":"MCP tools reference","heading":"","section":"MCP","crumbs":["MCP","Reference","MCP tools reference"],"url":"/docs/mcp/tools","text":"The full list of the tools and prompts of the Genezio MCP server: what each tool reads or changes, and the scope that it needs. {/ This page comes from scripts/mcp_tools_doc.py in apps/api. Do not edit it. /} The Genezio MCP server has 155 tools and 2 prompts. 113 tools read data and need the mcp:read scope. 42 tools change data and need the mcp:write scope. An AI assistant selects the tools by itself, thus you do not call them by name. Ask a question in your own words, and the assistant calls the tools that answer it. A tool shows only the brands that your Genezio user can open. A client with only the mcp:read scope does not see the write tools. Refer to Authentication and scopes.","keywords":""},{"kind":"section","title":"MCP tools reference","heading":"Prompts","section":"MCP","crumbs":["Reference","MCP tools reference","Prompts"],"url":"/docs/mcp/tools#prompts","text":"A prompt is a ready workflow. In most clients, you start it with a slash command, for example /genezio_exec_brief. genezio_exec_brief: brand, focus, window_hours Single safe executive MCP brief. One-tool diagnostics for a brand with evidence and deterministic next-step call. genezio_build_business_goal: brand_id, goal Build one Business Scorecard goal for a brand, in a conversation with the agent that writes one. The agent settles the number, reads the real data, and saves the card.","keywords":""},{"kind":"section","title":"MCP tools reference","heading":"Business Scorecard","section":"MCP","crumbs":["Reference","MCP tools reference","Business Scorecard"],"url":"/docs/mcp/tools#business-scorecard","text":"archive_scorecard_goal: write Take one goal off the board. build_scorecard_goal: write Make a goal of the Business Scorecard of a brand, in a conversation with the agent that builds one. get_scorecard_authoring_guide: read The rules that a Business Scorecard controller must keep, the reads of the API that it can call, and the contract of each view type that the board draws. get_scorecard_goal: read One goal of the Business Scorecard, with its datapoints and the history of its runs. list_archived_scorecard_goals: read The goals of the Business Scorecard of one brand that a person archived, last archived first. list_scorecard_goals: read The goals of the Business Scorecard of one brand, each with the datapoints of its last 90 days. probe_scorecard_data: write Run a short Python script against the real data of one brand, in a sandbox, and read what it printed. restore_scorecard_goal: write Put an archived goal back on the board. revise_scorecard_goal: write Write a new version of the controller of a goal that exists, with the sentences that moved with it. run_scorecard_goal: write Compute the last date again for one goal, in the same way as the night does. update_scorecard_goal: write Change the title, the category, or the dimensions of one goal.","keywords":""},{"kind":"section","title":"MCP tools reference","heading":"CDN logs","section":"MCP","crumbs":["Reference","MCP tools reference","CDN logs"],"url":"/docs/mcp/tools#cdn-logs","text":"analyze_cdn_vs_citations: read Compare pages public-LLM crawlers fetched (from a CDN log) against pages cited in the brand's simulated scenarios, matched by URL. cluster_cdn_logs_by_keywords: write Cluster a brand's CDN log of public-LLM crawler traffic into keyword-defined buckets and get back per-bucket traffic volumes. get_cdn_crawler_breakdown: read Per-crawler traffic totals for one CDN log file: row counts and occurrence sums for each AI-assistant crawler (ChatGPT, Claude, Perplexity, Google, Bing, etc.), sorted by volume. get_saved_cdn_log_clusters: read Fetch ONE named clustering snapshot for a brand+period. list_cdn_log_clusterings: read List every named clustering snapshot saved for a brand+period. list_cdn_log_inventory: read Catalog of CDN log files available for a brand in S3: period folders, CSV filenames, file sizes, and saved clustering counts. list_cdn_log_paths: read Return every unique URL path that public-LLM crawlers requested from the brand during the given period.","keywords":""},{"kind":"section","title":"MCP tools reference","heading":"E-commerce audit","section":"MCP","crumbs":["Reference","MCP tools reference","E-commerce audit"],"url":"/docs/mcp/tools#e-commerce-audit","text":"get_competitor_ucp_readiness: read Get latest crawled UCP readiness scores for tracked competitors. get_ecommerce_audit: read Get the latest UCP readiness audit for a brand, including rule results. get_ecommerce_audit_history: read Get UCP readiness score history for drift tracking. run_ecommerce_audit: write Queue a UCP readiness audit for a brand and its tracked competitors.","keywords":""},{"kind":"section","title":"MCP tools reference","heading":"Shopping and products","section":"MCP","crumbs":["Reference","MCP tools reference","Shopping and products"],"url":"/docs/mcp/tools#shopping-and-products","text":"get_brand_product: read Catalog details for a single product (title, category, line, manufacturer, label, image). get_conversation_shopping_percent: read Percentage of conversations that include a shopping carousel for a brand, optionally filtered by status, date range, topic, and model. get_merchant: read Merchant details plus products this merchant offered in shopping conversations. get_merchant_conversations: read The distinct shopping conversations in which a merchant offered a product, newest first. get_product_co_mentions: read Paginated products co-mentioned with a focus product in the same shopping carousels, ordered by the chosen metric (visibility or recommendation) in the latest period, descending. get_product_conversations: read The conversations a product appeared in. get_product_ownership_totals: read Distinct products with at least one appearance in the window, split into our brand products, competitor products, and the grand total. get_product_perception: read How the AI perceives a product, split into three sentiment columns (positive/negative/neutral), each ordered by how often the perception was repeated (occurrence_count desc) and paginated independently. get_product_query_conversations: read The distinct shopping conversations in which a product-search query was run, newest first. get_product_query_products: read Products returned for a product-search query. get_product_retailer_offers: read Paginated product offers for one retailer at one product level. get_product_series: read Per-period visibility and recommendation (each with share-of-voice) for one product and its co-mentioned products, for the product drawer chart. get_product_sku_retailers: read Paginated retailers for one level of a product: a specific SKU when sku_id is given, otherwise the general (variant-less) level. get_product_sku_visibility: read A single product's aggregate shopping visibility plus a per-SKU (variant) breakdown over the range. get_product_sources: read Paginated citation sources (URL, answer engine, ownership type, sentiment, occurrences, percent_cited) from the conversations a product appeared in, with first-party stats. get_product_visibility_metrics: read Per-product shopping time series for a brand. get_retailer_share: read Merchants ranked by their share of the brand's shopping conversations. list_actual_competitor_products: read Actual competitors for a brand that have at least one product appearing in the filtered shopping conversations, each with a product_count (distinct products that appeared in the topic/model/date window). list_brand_product_categories: read Distinct product categories for a brand, sorted alphabetically. list_brand_product_queries: read Paginated LLM-generated product-search queries for a brand (deduplicated per brand), sorted by occurrence count descending. list_brand_product_query_fanouts: read Paginated query fanouts from conversations where brand products appeared (via product_appearances). list_brand_products: read Paginated list of the brand's products that appeared in shopping conversations matching the filters, sorted by visibility descending. rank_brand_products: read Paginated ranking of the brand's products by a single metric, for the top/worst products widget.","keywords":""},{"kind":"section","title":"MCP tools reference","heading":"Brand groups","section":"MCP","crumbs":["Reference","MCP tools reference","Brand groups"],"url":"/docs/mcp/tools#brand-groups","text":"get_alias_group_by_id: read Fetch detailed configuration and membership for a specific alias group by its ID. list_all_alias_groups: read Retrieve configured brand groups for a specific brand.","keywords":""},{"kind":"section","title":"MCP tools reference","heading":"Search Console and integrations","section":"MCP","crumbs":["Reference","MCP tools reference","Search Console and integrations"],"url":"/docs/mcp/tools#search-console-and-integrations","text":"get_search_console_queries: read Read the top Google Search Console queries of the property that a brand reads, with the clicks, the impressions, the CTR, and the position of each one. list_brand_integrations: read List the external providers of a brand and whether each one is connected: Google Search Console and Google Analytics 4. list_search_console_properties: read List the Google Search Console properties of the account that a brand connected, and name the one that the brand reads. select_search_console_property: write Store the Google Search Console property that a brand reads.","keywords":""},{"kind":"section","title":"MCP tools reference","heading":"Tracked URLs","section":"MCP","crumbs":["Reference","MCP tools reference","Tracked URLs"],"url":"/docs/mcp/tools#tracked-urls","text":"assign_tracked_url_group: write Move tracked URLs into one custom group. create_tracked_url_group: write Create a custom group that collects the tracked URLs of a brand. delete_tracked_url_group: write Delete a custom group of tracked URLs. list_brand_tracked_urls: read Fetch a paginated list of all URLs being actively tracked for brand citations, with optional filtering by date, topic, model, and search term. list_tracked_url_groups: read List the custom groups that collect the tracked URLs of a brand, each one with the number of its members. update_tracked_url_group: write Rename a custom group of tracked URLs.","keywords":""},{"kind":"section","title":"MCP tools reference","heading":"Articles and briefs","section":"MCP","crumbs":["Reference","MCP tools reference","Articles and briefs"],"url":"/docs/mcp/tools#articles-and-briefs","text":"get_article_details: read Fetch the full content and metadata of a specific article by its ID. get_brief_details: read Fetch details and content of a specific brief by ID. update_article_schema_markup: write Replace the schema.org markup (JSON-LD) of an article with a full edited JSON-LD object.","keywords":""},{"kind":"section","title":"MCP tools reference","heading":"Competitors and SWOT","section":"MCP","crumbs":["Reference","MCP tools reference","Competitors and SWOT"],"url":"/docs/mcp/tools#competitors-and-swot","text":"create_competitor: write Create a new competitor for a brand. delete_competitor: write Soft-delete a competitor by ID. demote_competitor: write Demote an actual competitor back to suggested. get_brand_competitor_presence_by_model: read Presence or recommendation percentages per AI model for the brand and its competitors. get_competitor_details: read Retrieve detailed information about a specific competitor by its ID. get_competitor_swot_detail: read Get full detail for a single SWOT entry, including paginated child statements. get_competitor_swot_overview: read Get the AI-generated head-to-head SWOT overview comparing the brand against a specific competitor. get_competitor_swots: read Get clustered SWOT statements for a brand, grouped by competitor. list_competitors_by_brand: read List confirmed competitor names and websites for a brand. update_competitor: write Update competitor details by competitor ID.","keywords":""},{"kind":"section","title":"MCP tools reference","heading":"Perceptions, sentiment and fact check","section":"MCP","crumbs":["Reference","MCP tools reference","Perceptions, sentiment and fact check"],"url":"/docs/mcp/tools#perceptions-sentiment-and-fact-check","text":"get_brand_perception_confirmation_analytics: read Confirmed, contradicted, and unconfirmed perception occurrence counts for pie charts (weighted by cluster size). get_brand_perception_summaries: read Get cached or freshly generated perception summaries for a brand. get_brand_statement_sentiment_analytics: read Get sentiment analytics for a brand's statements over time, including positive/neutral/negative percentages and latest snapshot. get_perception_monitoring_run: read Read the run of a monitor that start_perception_monitoring started. get_statement_details: read Retrieve detailed information about a specific statement by its ID, including all its mentions across different conversations. get_stored_perception_verdicts: read Give the verdicts that the runs of a monitor saved, with the filters of the monitor. get_topic_fact_check_stats: read Retrieve fact-check pass/fail counts for fact_checker topics under a brand, grouped by topic and scenario. get_topic_fact_check_stats_by_model: read Retrieve claimed-rate percentages grouped by LLM model for fact_checker topics. get_topic_fact_check_stats_over_time: read Retrieve claimed vs not-claimed counts over time for fact_checker topics. get_topic_sentiment_stats: read Retrieve sentiment statistics (positive, negative, neutral) for introspector topics under a specific brand. list_brand_perceptions: read Fetch a paginated list of brand perceptions for a specific brand ID. list_citation_sources_by_statement: read Paginated citation sources linked to a specific statement ID via statement-source associations. list_monitored_perceptions: read List the monitors of the brand, from the last run to the oldest. list_statements_by_citation_domain: read Paginated statements linked to citation sources from a specific domain via statement-source associations. list_statements_by_citation_source: read Paginated statements linked to a specific citation source ID via statement-source associations. start_perception_monitoring: read Count how many answers of the models agree with a perception, and how many disagree. stop_perception_monitoring_run: write Stop the run of a monitor after the step that the worker reads now.","keywords":""},{"kind":"section","title":"MCP tools reference","heading":"Query fanouts","section":"MCP","crumbs":["Reference","MCP tools reference","Query fanouts"],"url":"/docs/mcp/tools#query-fanouts","text":"get_fanout_by_id: read Metadata and total occurrence count for a query fanout by ID: keyword, llm_model, and replay_index. get_fanouts_by_citation: read Retrieve query fanouts extracted from a specific citation source (by ID or URL). get_fanouts_for_phrase: read Return a paginated list of full fanout queries linked to the exact given sub-query phrase. get_sub_query_fanouts_for_brand: read Return a paginated list of sub-query phrases for a brand. list_brand_query_fanouts: read Fetch a paginated list of query fanouts associated with a brand, with filters for date, topics, and models. list_citation_sources_by_fanout: read Paginated citation sources that contain a specific query fanout. list_conversations_by_fanout: read Paginated conversations where a query fanout was mentioned.","keywords":""},{"kind":"section","title":"MCP tools reference","heading":"Citations","section":"MCP","crumbs":["Reference","MCP tools reference","Citations"],"url":"/docs/mcp/tools#citations","text":"get_brand_citations_analytics: read Aggregated citation analytics across a brand: domain ranking, occurrence trends over time for selected root_domains, period totals, and source-ownership distribution. get_citation_metrics_by_domain: read Citation occurrence counts for a single domain: total count, daily time series, breakdowns by topic, scenario, and LLM model, and how many citations name a channel or an author (channel_occurrences, author_occurrences). get_source_citation_insights: read Fetch qualitative and quantitative insights for a specific citation source ID. get_top_citation_sources: read Retrieve the most frequently cited sources for a brand, ordered by occurrence count. get_total_brand_citations: read Fetch the total number of citations/references identified for a specific brand, with optional filters for topics, date range, and models. list_brand_citation_channels_and_authors: read Which communities and people the answer engines cite for a brand. list_brand_citation_domains: read Retrieve citation sources grouped by domain, with total_occurrences and cited_occurrences per domain and per source. list_brand_citation_sources: read Fetch a detailed, paginated list of all citation sources for a brand, including URL, sentiment, total_occurrences, and cited_occurrences (sources with matched_texts in the LLM response). list_domain_citation_channels_and_authors: read The channels or the authors behind the cited URLs of one domain, most cited first, with citation counts. list_recent_conversations_by_citation: read Fetch the most recent conversation for a specific normalized URL citation, with optional date, topic, and model filters. list_scenario_sources: read List all citation sources (websites/platforms) identified in conversations for a specific scenario. list_source_url_occurrences: read Retrieve a list of citation URLs with total_occurrences and cited_occurrences across all conversations for a brand.","keywords":""},{"kind":"section","title":"MCP tools reference","heading":"Insights","section":"MCP","crumbs":["Reference","MCP tools reference","Insights"],"url":"/docs/mcp/tools#insights","text":"get_insight_details: read Fetch the full content and metadata of a specific data-driven insight by its ID. list_brand_insights: read Retrieve all insights generated for a specific brand ID, with optional date range and frequency filters. list_insights_by_status: read Fetch insights filtered by status (todo, done, declined), optionally scoped to a specific brand ID.","keywords":""},{"kind":"section","title":"MCP tools reference","heading":"Conversations","section":"MCP","crumbs":["Reference","MCP tools reference","Conversations"],"url":"/docs/mcp/tools#conversations","text":"get_conversation_aggregates: read Fetch aggregate counts of conversations for a brand, optionally filtered by status, date range, topic, and model. get_conversation_details: read Fetch a complete log of a specific conversation by its ID, including AI messages, brand rankings, and citation sources. get_conversation_progress: read Get conversation status counts for a brand (total, success, in_progress, error). list_recent_conversations_for_brand: read Fetch the most recent conversations (max 5) for a specific brand. list_scenario_conversations: read Fetch detailed conversation logs and metadata for a specific scenario. list_topic_conversations: read Paginated topic conversations: filter by date, llm_models (Answer Engine ids as stored, e.g. search_conversations: read Find what the answers of the models say about a subject, such as 'Vercel pricing' or 'credit for young people'.","keywords":""},{"kind":"section","title":"MCP tools reference","heading":"Scenarios","section":"MCP","crumbs":["Reference","MCP tools reference","Scenarios"],"url":"/docs/mcp/tools#scenarios","text":"create_scenario: write Create a single scenario for a topic. delete_scenario: write Delete a scenario by ID. get_scenario_details: read Retrieve detailed information about a specific scenario by its ID. get_scenario_performance_metrics: read Fetch quantitative metrics (mentions/recommendations) for a specific scenario over time. list_all_scenarios_by_brand: read Fetch all scenarios for a specific brand ID. list_recently_deleted_scenarios_by_topic: read Fetch scenarios soft-deleted within the last N days for a topic. list_scenarios_for_topic: read Retrieve all scenarios associated with a given topic ID. restore_scenario: write Restore a soft-deleted scenario by ID. start_scenario: write Start scenario runs for a specific scenario ID. update_scenario: write Update a single scenario by ID.","keywords":""},{"kind":"section","title":"MCP tools reference","heading":"Topics and tags","section":"MCP","crumbs":["Reference","MCP tools reference","Topics and tags"],"url":"/docs/mcp/tools#topics-and-tags","text":"change_topics_status: write Bulk update status for multiple topics. create_topic: write Create a new topic for a brand. create_topic_custom_tag: write Create a new custom tag for organizing topics within a brand. delete_topic: write Soft-delete a topic by ID. delete_topic_custom_tag: write Soft-delete a custom topic tag for a brand and remove it from all topics. get_topic_details: read Fetch detailed information about a specific topic by its ID. get_topic_performance_metrics: read Top competitors by visibility or recommendation (dashboard Top 10 widget). list_recently_deleted_topics_by_brand: read Fetch topics soft-deleted within the last N days for a brand, including their scenarios and custom tags. list_topic_custom_tags: read List brand-scoped custom topic tag definitions (names and ids) for filtering and assignment. list_topics_by_brand: read List all topics for a brand, including nested scenarios and brand-scoped custom_tags (tag ids and names). replace_topic_custom_tags: write Replace the full set of custom tags assigned to a topic (body: tag_ids). restore_topic: write Restore a soft-deleted topic by ID. start_topics: write Start one or more topics for execution. update_topic: write Update topic metadata by topic ID. update_topic_custom_tag: write Update a custom topic tag of a brand: its name, its color, or the personas of a master filter.","keywords":""},{"kind":"section","title":"MCP tools reference","heading":"Personas","section":"MCP","crumbs":["Reference","MCP tools reference","Personas"],"url":"/docs/mcp/tools#personas","text":"create_persona: write Create a new persona for a brand. delete_persona: write Delete an existing persona by ID. get_persona_details: read Fetch full details for a specific persona, including demographics, interests, and behavior profile. list_personas: read Retrieve the research personas of one brand. update_persona: write Update an existing persona by ID.","keywords":""},{"kind":"section","title":"MCP tools reference","heading":"Brands and visibility","section":"MCP","crumbs":["Reference","MCP tools reference","Brands and visibility"],"url":"/docs/mcp/tools#brands-and-visibility","text":"create_brand: write Create a new brand in a specific account. delete_brand: write Soft-delete a brand by ID. get_brand_overview: read Read-only: get a brand by ID with nested competitors, topics, and scenarios in one response. get_brand_visibility_metrics: read Brand visibility/recommendation time series. list_all_available_brands: read Fetch a paginated list of all brands the user has access to with optional filtering. set_brand_llm_models: write Set which LLM/AI models (answer engines) a brand monitors. update_brand_websites: write Replace or toggle websites for a specific brand.","keywords":""},{"kind":"section","title":"MCP tools reference","heading":"Web research","section":"MCP","crumbs":["Reference","MCP tools reference","Web research"],"url":"/docs/mcp/tools#web-research","text":"crawl_website: read Crawls a single given URL using Firecrawl and returns the extracted content. google_search_ui: read Searches Google using Firecrawl and returns a list of search results (URLs, titles, and descriptions).","keywords":""},{"kind":"route","title":"List Brands","heading":"","section":"API reference","crumbs":["API reference","Brands","List Brands"],"url":"/docs/api-reference/brands/list-brands","method":"GET","path":"/customer/v1/brands","text":"List the brands of the account of the API key, by name.","keywords":"/customer/v1/brands page The page, from 1. page_size The rows of a page."},{"kind":"route","title":"Get Brand","heading":"","section":"API reference","crumbs":["API reference","Brands","Get Brand"],"url":"/docs/api-reference/brands/get-brand","method":"GET","path":"/customer/v1/brands/{brand_id}","text":"One brand, with its websites and its answer engines.","keywords":"/customer/v1/brands/{brand_id} brand_id"},{"kind":"route","title":"Update Brand","heading":"","section":"API reference","crumbs":["API reference","Brands","Update Brand"],"url":"/docs/api-reference/brands/update-brand","method":"PATCH","path":"/customer/v1/brands/{brand_id}","text":"Change the websites or the answer engines of the brand. websites replaces each website: the platform reads a page of these websites as a page of the brand. An engine that the plan does not hold gets 403 plan_limit_reached. A topic with its own engines keeps them, and topics_with_other_engines names each topic whose engines the brand does not have now.","keywords":"/customer/v1/brands/{brand_id} brand_id Idempotency-Key Any unique string, such as a UUID. Send the same key again to get the first answer and not a second write. The API keeps an answer for 24 hours."},{"kind":"route","title":"List Articles","heading":"","section":"API reference","crumbs":["API reference","Articles","List Articles"],"url":"/docs/api-reference/articles/list-articles","method":"GET","path":"/customer/v1/brands/{brand_id}/articles","text":"The articles of the brand, the newest first, with no text. info counts the articles in each status.","keywords":"/customer/v1/brands/{brand_id}/articles brand_id topic_ids page page_size"},{"kind":"route","title":"Get Article","heading":"","section":"API reference","crumbs":["API reference","Articles","Get Article"],"url":"/docs/api-reference/articles/get-article","method":"GET","path":"/customer/v1/brands/{brand_id}/articles/{article_id}","text":"One article, with its text in Markdown, its sources and its schema.org markup. Send version for an older text.","keywords":"/customer/v1/brands/{brand_id}/articles/{article_id} article_id An article of the brand. brand_id version With no version, the newest."},{"kind":"route","title":"List Brand Groups","heading":"","section":"API reference","crumbs":["API reference","Brand groups","List Brand Groups"],"url":"/docs/api-reference/brand-groups/list-brand-groups","method":"GET","path":"/customer/v1/brands/{brand_id}/brand-groups","text":"The brand groups of the brand, by name, each with its members.","keywords":"/customer/v1/brands/{brand_id}/brand-groups brand_id"},{"kind":"route","title":"Create Brand Group","heading":"","section":"API reference","crumbs":["API reference","Brand groups","Create Brand Group"],"url":"/docs/api-reference/brand-groups/create-brand-group","method":"POST","path":"/customer/v1/brands/{brand_id}/brand-groups","text":"Create a brand group. Each member must be a competitor of the brand that is in no other group, and the name must be new for the brand.","keywords":"/customer/v1/brands/{brand_id}/brand-groups brand_id Idempotency-Key Any unique string, such as a UUID. Send the same key again to get the first answer and not a second write. The API keeps an answer for 24 hours."},{"kind":"route","title":"Delete Brand Group","heading":"","section":"API reference","crumbs":["API reference","Brand groups","Delete Brand Group"],"url":"/docs/api-reference/brand-groups/delete-brand-group","method":"DELETE","path":"/customer/v1/brands/{brand_id}/brand-groups/{group_id}","text":"Delete a brand group. Its competitors stay, and the reports count each one alone again.","keywords":"/customer/v1/brands/{brand_id}/brand-groups/{group_id} group_id A brand group of the brand. brand_id"},{"kind":"route","title":"Get Brand Group","heading":"","section":"API reference","crumbs":["API reference","Brand groups","Get Brand Group"],"url":"/docs/api-reference/brand-groups/get-brand-group","method":"GET","path":"/customer/v1/brands/{brand_id}/brand-groups/{group_id}","text":"One brand group, with its members.","keywords":"/customer/v1/brands/{brand_id}/brand-groups/{group_id} group_id A brand group of the brand. brand_id"},{"kind":"route","title":"Update Brand Group","heading":"","section":"API reference","crumbs":["API reference","Brand groups","Update Brand Group"],"url":"/docs/api-reference/brand-groups/update-brand-group","method":"PATCH","path":"/customer/v1/brands/{brand_id}/brand-groups/{group_id}","text":"Change the fields that the request sends. Change the members with PUT /members.","keywords":"/customer/v1/brands/{brand_id}/brand-groups/{group_id} group_id A brand group of the brand. brand_id"},{"kind":"route","title":"Set Brand Group Members","heading":"","section":"API reference","crumbs":["API reference","Brand groups","Set Brand Group Members"],"url":"/docs/api-reference/brand-groups/set-brand-group-members","method":"PUT","path":"/customer/v1/brands/{brand_id}/brand-groups/{group_id}/members","text":"Give the group its whole list of members, with their ranks. A competitor that is not in the list leaves the group. Send the same list two times, and the second request changes nothing.","keywords":"/customer/v1/brands/{brand_id}/brand-groups/{group_id}/members group_id A brand group of the brand. brand_id"},{"kind":"route","title":"List Briefs","heading":"","section":"API reference","crumbs":["API reference","Briefs","List Briefs"],"url":"/docs/api-reference/briefs/list-briefs","method":"GET","path":"/customer/v1/brands/{brand_id}/briefs","text":"The briefs of the brand, the newest first, with no text.","keywords":"/customer/v1/brands/{brand_id}/briefs brand_id topic_ids page page_size"},{"kind":"route","title":"Get Brief","heading":"","section":"API reference","crumbs":["API reference","Briefs","Get Brief"],"url":"/docs/api-reference/briefs/get-brief","method":"GET","path":"/customer/v1/brands/{brand_id}/briefs/{brief_id}","text":"One brief, with its text in Markdown. Send version for an older text.","keywords":"/customer/v1/brands/{brand_id}/briefs/{brief_id} brief_id A brief of the brand. brand_id version With no version, the newest."},{"kind":"route","title":"Citation Domains","heading":"","section":"API reference","crumbs":["API reference","Citations","Citation Domains"],"url":"/docs/api-reference/citations/citation-domains","method":"GET","path":"/customer/v1/brands/{brand_id}/citations/domains","text":"The domains of the pages that the answers cite, the most frequent first.","keywords":"/customer/v1/brands/{brand_id}/citations/domains brand_id start_date YYYY-MM-DD (a day in UTC) or an ISO 8601 datetime. Both ends of the range are inclusive: an end date covers its whole day. end_date YYYY-MM-DD (a day in UTC) or an ISO 8601 datetime. Both ends of the range are inclusive: an end date covers its whole day. models The answer engines. With no list, each engine. topic_ids The topics of the brand. With no list, each topic. ownership Who owns the page, for example competitor_owned. content_types What the page is, for example review_editorial. sentiment How the page speaks of the brand. first_party true: only the pages of the brand. false: only the pages of others. No value: each page. page The page, from 1. page_size The rows of a page."},{"kind":"route","title":"Citation Domain","heading":"","section":"API reference","crumbs":["API reference","Citations","Citation Domain"],"url":"/docs/api-reference/citations/citation-domain","method":"GET","path":"/customer/v1/brands/{brand_id}/citations/domains/{domain}","text":"One domain: its occurrences over time, by topic, by scenario and by engine. A domain with no citation in the range gives 404.","keywords":"/customer/v1/brands/{brand_id}/citations/domains/{domain} domain A domain of the list, such as www.g2.com. brand_id start_date YYYY-MM-DD (a day in UTC) or an ISO 8601 datetime. Both ends of the range are inclusive: an end date covers its whole day. end_date YYYY-MM-DD (a day in UTC) or an ISO 8601 datetime. Both ends of the range are inclusive: an end date covers its whole day. models The answer engines. With no list, each engine. topic_ids The topics of the brand. With no list, each topic. ownership Who owns the page, for example competitor_owned. content_types What the page is, for example review_editorial. sentiment How the page speaks of the brand. first_party true: only the pages of the brand. false: only the pages of others. No value: each page."},{"kind":"route","title":"Citation Urls","heading":"","section":"API reference","crumbs":["API reference","Citations","Citation Urls"],"url":"/docs/api-reference/citations/citation-urls","method":"GET","path":"/customer/v1/brands/{brand_id}/citations/urls","text":"The pages that the answers cite, the most frequent or the newest first.","keywords":"/customer/v1/brands/{brand_id}/citations/urls brand_id start_date YYYY-MM-DD (a day in UTC) or an ISO 8601 datetime. Both ends of the range are inclusive: an end date covers its whole day. end_date YYYY-MM-DD (a day in UTC) or an ISO 8601 datetime. Both ends of the range are inclusive: an end date covers its whole day. models The answer engines. With no list, each engine. topic_ids The topics of the brand. With no list, each topic. ownership Who owns the page, for example competitor_owned. content_types What the page is, for example review_editorial. sentiment How the page speaks of the brand. first_party true: only the pages of the brand. false: only the pages of others. No value: each page. page The page, from 1. page_size The rows of a page. sort occurrences: the most frequent pages first. latest: the newest first. search A part of the address of the page."},{"kind":"route","title":"List Competitors","heading":"","section":"API reference","crumbs":["API reference","Competitors","List Competitors"],"url":"/docs/api-reference/competitors/list-competitors","method":"GET","path":"/customer/v1/brands/{brand_id}/competitors","text":"List the competitors of a brand.","keywords":"/customer/v1/brands/{brand_id}/competitors brand_id actual_only True gives the competitors that the account confirmed. False also gives the competitors that the platform suggests."},{"kind":"route","title":"Create Competitors","heading":"","section":"API reference","crumbs":["API reference","Competitors","Create Competitors"],"url":"/docs/api-reference/competitors/create-competitors","method":"POST","path":"/customer/v1/brands/{brand_id}/competitors","text":"Track up to 50 competitors of the brand. A name of a suggestion of the platform makes the suggestion a tracked competitor, with its history. A name that the brand already tracks is rejected with conflict. The case of a name does not matter.","keywords":"/customer/v1/brands/{brand_id}/competitors brand_id Idempotency-Key Any unique string, such as a UUID. Send the same key again to get the first answer and not a second write. The API keeps an answer for 24 hours."},{"kind":"route","title":"Delete Competitor","heading":"","section":"API reference","crumbs":["API reference","Competitors","Delete Competitor"],"url":"/docs/api-reference/competitors/delete-competitor","method":"DELETE","path":"/customer/v1/brands/{brand_id}/competitors/{competitor_id}","text":"Delete a competitor. The reports stop showing it.","keywords":"/customer/v1/brands/{brand_id}/competitors/{competitor_id} competitor_id A competitor of the brand. brand_id"},{"kind":"route","title":"Update Competitor","heading":"","section":"API reference","crumbs":["API reference","Competitors","Update Competitor"],"url":"/docs/api-reference/competitors/update-competitor","method":"PATCH","path":"/customer/v1/brands/{brand_id}/competitors/{competitor_id}","text":"Change the name or the website, or track or stop tracking a competitor. is_actual: false sends a competitor back to the suggestions, and is_actual: true tracks a suggestion. Only the fields of the body change.","keywords":"/customer/v1/brands/{brand_id}/competitors/{competitor_id} competitor_id A competitor of the brand. brand_id Idempotency-Key Any unique string, such as a UUID. Send the same key again to get the first answer and not a second write. The API keeps an answer for 24 hours."},{"kind":"route","title":"List Content Analyses","heading":"","section":"API reference","crumbs":["API reference","Content analyses","List Content Analyses"],"url":"/docs/api-reference/content-analyses/list-content-analyses","method":"GET","path":"/customer/v1/brands/{brand_id}/content-analyses","text":"The content analyses of the brand, the newest first, with no result.","keywords":"/customer/v1/brands/{brand_id}/content-analyses brand_id limit The newest first."},{"kind":"route","title":"Get Content Analysis","heading":"","section":"API reference","crumbs":["API reference","Content analyses","Get Content Analysis"],"url":"/docs/api-reference/content-analyses/get-content-analysis","method":"GET","path":"/customer/v1/brands/{brand_id}/content-analyses/{analysis_id}","text":"One content analysis, with its result. While status is pending or running, the result has the parts of the steps that ended. Read it again until status is success or failed.","keywords":"/customer/v1/brands/{brand_id}/content-analyses/{analysis_id} analysis_id A content analysis of the brand. brand_id"},{"kind":"route","title":"List Conversations","heading":"","section":"API reference","crumbs":["API reference","Conversations","List Conversations"],"url":"/docs/api-reference/conversations/list-conversations","method":"GET","path":"/customer/v1/brands/{brand_id}/conversations","text":"List the answers of the engines, the newest first. Each answer gives whether it mentions and recommends the brand, the place of the brand, and each brand that it names in order. Filter by topic, scenario, engine, mentioned and recommended.","keywords":"/customer/v1/brands/{brand_id}/conversations brand_id start_date YYYY-MM-DD (a day in UTC) or an ISO 8601 datetime. Both ends of the range are inclusive: an end date covers its whole day. end_date YYYY-MM-DD (a day in UTC) or an ISO 8601 datetime. Both ends of the range are inclusive: an end date covers its whole day. models The answer engines. With no list, each engine. topic_id Only the answers of a topic. scenario_id Only the answers of a scenario. mentioned true: only the answers that mention the brand. false: only the others. recommended true: only the answers that recommend the brand. false: only the others. order page page_size"},{"kind":"route","title":"Search Conversations","heading":"","section":"API reference","crumbs":["API reference","Conversations","Search Conversations"],"url":"/docs/api-reference/conversations/search-conversations","method":"GET","path":"/customer/v1/brands/{brand_id}/conversations/search","text":"Find what the answers say about a subject, such as battery life. The search matches by meaning and by the words of q, and a model orders the paragraphs by how well they match. Each answer gives the paragraphs that match, with relevance from 0 to 1.","keywords":"/customer/v1/brands/{brand_id}/conversations/search brand_id start_date YYYY-MM-DD (a day in UTC) or an ISO 8601 datetime. Both ends of the range are inclusive: an end date covers its whole day. end_date YYYY-MM-DD (a day in UTC) or an ISO 8601 datetime. Both ends of the range are inclusive: an end date covers its whole day. models The answer engines. With no list, each engine. topic_ids The topics of the brand. With no list, each topic. q Words or a question, such as price of the premium plan. limit The paragraphs to read."},{"kind":"route","title":"Conversation Stats","heading":"","section":"API reference","crumbs":["API reference","Conversations","Conversation Stats"],"url":"/docs/api-reference/conversations/conversation-stats","method":"GET","path":"/customer/v1/brands/{brand_id}/conversations/stats","text":"How many answers the range has, how many show products to buy, and the state of the runs of the brand now (in progress and failed).","keywords":"/customer/v1/brands/{brand_id}/conversations/stats brand_id start_date YYYY-MM-DD (a day in UTC) or an ISO 8601 datetime. Both ends of the range are inclusive: an end date covers its whole day. end_date YYYY-MM-DD (a day in UTC) or an ISO 8601 datetime. Both ends of the range are inclusive: an end date covers its whole day. models The answer engines. With no list, each engine. topic_ids The topics of the brand. With no list, each topic."},{"kind":"route","title":"Get Conversation","heading":"","section":"API reference","crumbs":["API reference","Conversations","Get Conversation"],"url":"/docs/api-reference/conversations/get-conversation","method":"GET","path":"/customer/v1/brands/{brand_id}/conversations/{conversation_id}","text":"One answer: the scenario, the text of the answer, the brands that it names and recommends in order, and the pages that it cites.","keywords":"/customer/v1/brands/{brand_id}/conversations/{conversation_id} conversation_id A conversation of the brand. brand_id"},{"kind":"route","title":"Get Ecommerce Audit","heading":"","section":"API reference","crumbs":["API reference","E-commerce audit","Get Ecommerce Audit"],"url":"/docs/api-reference/e-commerce-audit/get-ecommerce-audit","method":"GET","path":"/customer/v1/brands/{brand_id}/ecommerce-audit","text":"The newest audit: the score, each rule with the failed ones first, and each request of the audit to the site. A brand with no audit gets 404.","keywords":"/customer/v1/brands/{brand_id}/ecommerce-audit brand_id engine The surface of the engine that the audit reads."},{"kind":"route","title":"Ecommerce Audit Of Competitors","heading":"","section":"API reference","crumbs":["API reference","E-commerce audit","Ecommerce Audit Of Competitors"],"url":"/docs/api-reference/e-commerce-audit/ecommerce-audit-of-competitors","method":"GET","path":"/customer/v1/brands/{brand_id}/ecommerce-audit/competitors","text":"The newest score of each tracked competitor, the highest first.","keywords":"/customer/v1/brands/{brand_id}/ecommerce-audit/competitors brand_id engine The surface of the engine that the audit reads."},{"kind":"route","title":"Ecommerce Audit History","heading":"","section":"API reference","crumbs":["API reference","E-commerce audit","Ecommerce Audit History"],"url":"/docs/api-reference/e-commerce-audit/ecommerce-audit-history","method":"GET","path":"/customer/v1/brands/{brand_id}/ecommerce-audit/history","text":"The score of each audit, the oldest first, with the change between two audits.","keywords":"/customer/v1/brands/{brand_id}/ecommerce-audit/history brand_id engine The surface of the engine that the audit reads. limit The newest audits."},{"kind":"route","title":"Start Ecommerce Audit","heading":"","section":"API reference","crumbs":["API reference","E-commerce audit","Start Ecommerce Audit"],"url":"/docs/api-reference/e-commerce-audit/start-ecommerce-audit","method":"POST","path":"/customer/v1/brands/{brand_id}/ecommerce-audit/runs","text":"Start an audit of the site of the brand, and of its competitors. The audit reads the sites and needs no answer engine, thus it costs no run. It takes some minutes: read the newest audit after it.","keywords":"/customer/v1/brands/{brand_id}/ecommerce-audit/runs brand_id"},{"kind":"route","title":"List Fanouts","heading":"","section":"API reference","crumbs":["API reference","Query fanouts","List Fanouts"],"url":"/docs/api-reference/query-fanouts/list-fanouts","method":"GET","path":"/customer/v1/brands/{brand_id}/fanouts","text":"The web searches that the engines ran to answer the scenarios of the brand, the most frequent first. relevance gives the share of their answers that name the brand or a tracked competitor.","keywords":"/customer/v1/brands/{brand_id}/fanouts brand_id start_date YYYY-MM-DD (a day in UTC) or an ISO 8601 datetime. Both ends of the range are inclusive: an end date covers its whole day. end_date YYYY-MM-DD (a day in UTC) or an ISO 8601 datetime. Both ends of the range are inclusive: an end date covers its whole day. models The answer engines. With no list, the engines of the brand. topic_ids The topics of the brand. With no list, each topic. page page_size scenario_ids search A part of the search."},{"kind":"route","title":"List Fanouts Of Phrase","heading":"","section":"API reference","crumbs":["API reference","Query fanouts","List Fanouts Of Phrase"],"url":"/docs/api-reference/query-fanouts/list-fanouts-of-phrase","method":"GET","path":"/customer/v1/brands/{brand_id}/fanouts/by-phrase","text":"The searches that hold a phrase of the list of the phrases.","keywords":"/customer/v1/brands/{brand_id}/fanouts/by-phrase brand_id query phrase A phrase of the list of the phrases."},{"kind":"route","title":"List Cited Fanouts","heading":"","section":"API reference","crumbs":["API reference","Query fanouts","List Cited Fanouts"],"url":"/docs/api-reference/query-fanouts/list-cited-fanouts","method":"GET","path":"/customer/v1/brands/{brand_id}/fanouts/cited","text":"The searches that led the answers to a cited page or domain. Send url or domain: a page that you want the engines to cite tells you which searches find it.","keywords":"/customer/v1/brands/{brand_id}/fanouts/cited brand_id start_date YYYY-MM-DD (a day in UTC) or an ISO 8601 datetime. Both ends of the range are inclusive: an end date covers its whole day. end_date YYYY-MM-DD (a day in UTC) or an ISO 8601 datetime. Both ends of the range are inclusive: an end date covers its whole day. models The answer engines. With no list, the engines of the brand. topic_ids The topics of the brand. With no list, each topic. page page_size url A cited page. Send url or domain. domain A cited domain, such as rtings.com."},{"kind":"route","title":"List Fanout Phrases","heading":"","section":"API reference","crumbs":["API reference","Query fanouts","List Fanout Phrases"],"url":"/docs/api-reference/query-fanouts/list-fanout-phrases","method":"GET","path":"/customer/v1/brands/{brand_id}/fanouts/phrases","text":"The phrases that many searches share, such as noise cancelling, the most frequent first. They tell which words the engines use to search.","keywords":"/customer/v1/brands/{brand_id}/fanouts/phrases brand_id start_date YYYY-MM-DD (a day in UTC) or an ISO 8601 datetime. Both ends of the range are inclusive: an end date covers its whole day. end_date YYYY-MM-DD (a day in UTC) or an ISO 8601 datetime. Both ends of the range are inclusive: an end date covers its whole day. models The answer engines. With no list, the engines of the brand. topic_ids The topics of the brand. With no list, each topic. page page_size search"},{"kind":"route","title":"List Fanout Conversations","heading":"","section":"API reference","crumbs":["API reference","Query fanouts","List Fanout Conversations"],"url":"/docs/api-reference/query-fanouts/list-fanout-conversations","method":"GET","path":"/customer/v1/brands/{brand_id}/fanouts/{fanout_id}/conversations","text":"The answers that ran the search, the newest first, with the brands that each answer names.","keywords":"/customer/v1/brands/{brand_id}/fanouts/{fanout_id}/conversations fanout_id A search of the brand, from the list of the searches. brand_id start_date YYYY-MM-DD (a day in UTC) or an ISO 8601 datetime. Both ends of the range are inclusive: an end date covers its whole day. end_date YYYY-MM-DD (a day in UTC) or an ISO 8601 datetime. Both ends of the range are inclusive: an end date covers its whole day. models The answer engines. With no list, the engines of the brand. topic_ids The topics of the brand. With no list, each topic. page page_size"},{"kind":"route","title":"List Fanout Sources","heading":"","section":"API reference","crumbs":["API reference","Query fanouts","List Fanout Sources"],"url":"/docs/api-reference/query-fanouts/list-fanout-sources","method":"GET","path":"/customer/v1/brands/{brand_id}/fanouts/{fanout_id}/sources","text":"The pages that the search found and that the answers read.","keywords":"/customer/v1/brands/{brand_id}/fanouts/{fanout_id}/sources fanout_id A search of the brand, from the list of the searches. brand_id start_date YYYY-MM-DD (a day in UTC) or an ISO 8601 datetime. Both ends of the range are inclusive: an end date covers its whole day. end_date YYYY-MM-DD (a day in UTC) or an ISO 8601 datetime. Both ends of the range are inclusive: an end date covers its whole day. models The answer engines. With no list, the engines of the brand. topic_ids The topics of the brand. With no list, each topic. page page_size"},{"kind":"route","title":"List Knowledge Bases","heading":"","section":"API reference","crumbs":["API reference","Knowledge bases","List Knowledge Bases"],"url":"/docs/api-reference/knowledge-bases/list-knowledge-bases","method":"GET","path":"/customer/v1/brands/{brand_id}/knowledge-bases","text":"The knowledge bases of the brand, the newest first, each in its newest version that has no error.","keywords":"/customer/v1/brands/{brand_id}/knowledge-bases brand_id"},{"kind":"route","title":"Create Knowledge Base","heading":"","section":"API reference","crumbs":["API reference","Knowledge bases","Create Knowledge Base"],"url":"/docs/api-reference/knowledge-bases/create-knowledge-base","method":"POST","path":"/customer/v1/brands/{brand_id}/knowledge-bases","text":"Create a knowledge base from a text (type: text) or a site (type: url). The answer is 202 with status: indexing: the platform reads it in the background. Read it again until status is active. It costs an embedding of each chunk, and a site also costs a crawl of each page. A schedule crawls the site again by itself, and each crawl costs. The plan of the account limits the knowledge bases of a brand.","keywords":"/customer/v1/brands/{brand_id}/knowledge-bases brand_id Idempotency-Key Any unique string, such as a UUID. Send the same key again to get the first answer and not a second write. The API keeps an answer for 24 hours."},{"kind":"route","title":"Create Knowledge Base From Files","heading":"","section":"API reference","crumbs":["API reference","Knowledge bases","Create Knowledge Base From Files"],"url":"/docs/api-reference/knowledge-bases/create-knowledge-base-from-files","method":"POST","path":"/customer/v1/brands/{brand_id}/knowledge-bases/files","text":"Create a knowledge base from up to 20 files of up to 25 MiB each. Send the files as multipart/form-data. The answer is 202 with status: indexing, as for a text or a site. It costs an embedding of each chunk.","keywords":"/customer/v1/brands/{brand_id}/knowledge-bases/files brand_id"},{"kind":"route","title":"Search Knowledge Bases","heading":"","section":"API reference","crumbs":["API reference","Knowledge bases","Search Knowledge Bases"],"url":"/docs/api-reference/knowledge-bases/search-knowledge-bases","method":"GET","path":"/customer/v1/brands/{brand_id}/knowledge-bases/search","text":"The chunks closest to q by meaning, the closest first. Each search makes one embedding of q, thus it costs a little. A search gives no chunk when no chunk is close enough.","keywords":"/customer/v1/brands/{brand_id}/knowledge-bases/search brand_id q A question or words. knowledge_base_ids With no list, each knowledge base of the brand. limit min_similarity From 0 to 1. A chunk with no relation to the query scores about 0.5, thus 0.6 keeps the related chunks and 0.8 the close ones."},{"kind":"route","title":"Delete Knowledge Base","heading":"","section":"API reference","crumbs":["API reference","Knowledge bases","Delete Knowledge Base"],"url":"/docs/api-reference/knowledge-bases/delete-knowledge-base","method":"DELETE","path":"/customer/v1/brands/{brand_id}/knowledge-bases/{kb_id}","text":"Delete a knowledge base and each of its versions. One that a validation agent uses gets 409.","keywords":"/customer/v1/brands/{brand_id}/knowledge-bases/{kb_id} kb_id A knowledge base of the brand. brand_id"},{"kind":"route","title":"Get Knowledge Base","heading":"","section":"API reference","crumbs":["API reference","Knowledge bases","Get Knowledge Base"],"url":"/docs/api-reference/knowledge-bases/get-knowledge-base","method":"GET","path":"/customer/v1/brands/{brand_id}/knowledge-bases/{kb_id}","text":"One knowledge base, with its chunks and a link to each of its files.","keywords":"/customer/v1/brands/{brand_id}/knowledge-bases/{kb_id} kb_id A knowledge base of the brand. brand_id"},{"kind":"route","title":"Cancel Knowledge Base","heading":"","section":"API reference","crumbs":["API reference","Knowledge bases","Cancel Knowledge Base"],"url":"/docs/api-reference/knowledge-bases/cancel-knowledge-base","method":"POST","path":"/customer/v1/brands/{brand_id}/knowledge-bases/{kb_id}/cancel","text":"Stop the indexing of a knowledge base. One that does not index gets 409.","keywords":"/customer/v1/brands/{brand_id}/knowledge-bases/{kb_id}/cancel kb_id A knowledge base of the brand. brand_id"},{"kind":"route","title":"List Knowledge Base Chunks","heading":"","section":"API reference","crumbs":["API reference","Knowledge bases","List Knowledge Base Chunks"],"url":"/docs/api-reference/knowledge-bases/list-knowledge-base-chunks","method":"GET","path":"/customer/v1/brands/{brand_id}/knowledge-bases/{kb_id}/chunks","text":"The chunks of a knowledge base, as the check of the facts reads them.","keywords":"/customer/v1/brands/{brand_id}/knowledge-bases/{kb_id}/chunks kb_id A knowledge base of the brand. brand_id limit"},{"kind":"route","title":"Recrawl Knowledge Base","heading":"","section":"API reference","crumbs":["API reference","Knowledge bases","Recrawl Knowledge Base"],"url":"/docs/api-reference/knowledge-bases/recrawl-knowledge-base","method":"POST","path":"/customer/v1/brands/{brand_id}/knowledge-bases/{kb_id}/recrawl","text":"Crawl the site of a knowledge base again, as a new version. It costs a crawl of each page and an embedding of each chunk. A knowledge base that is not a site gets 400, and one that indexes now gets 409.","keywords":"/customer/v1/brands/{brand_id}/knowledge-bases/{kb_id}/recrawl kb_id A knowledge base of the brand. brand_id"},{"kind":"route","title":"List Models","heading":"","section":"API reference","crumbs":["API reference","Brands","List Models"],"url":"/docs/api-reference/brands/list-models","method":"GET","path":"/customer/v1/brands/{brand_id}/models","text":"Each answer engine: the plan holds it, and the brand runs on it. Use the id of an engine in models of a report and in answer_engines of a brand or a topic.","keywords":"/customer/v1/brands/{brand_id}/models brand_id"},{"kind":"route","title":"Perceptions","heading":"","section":"API reference","crumbs":["API reference","Perceptions","Perceptions"],"url":"/docs/api-reference/perceptions/perceptions","method":"GET","path":"/customer/v1/brands/{brand_id}/perceptions","text":"What the answer engines say about the brand, one perception a row. Each perception gives its sentiment, its statements, and the result of the check of its facts. info.summary gives the statements of the range by that result: confirmed, contradicted, or unconfirmed.","keywords":"/customer/v1/brands/{brand_id}/perceptions brand_id start_date YYYY-MM-DD (a day in UTC) or an ISO 8601 datetime. Both ends of the range are inclusive: an end date covers its whole day. end_date YYYY-MM-DD (a day in UTC) or an ISO 8601 datetime. Both ends of the range are inclusive: an end date covers its whole day. models The answer engines. With no list, each engine. topic_ids The topics of the brand. With no list, each topic. page The page, from 1. page_size The rows of a page. sentiment fact_check confirmed: the facts of the brand support it. contradicted: they go against it. unconfirmed: not checked yet. in_knowledge_base Whether the knowledge base of the brand holds it. search A part of the text."},{"kind":"route","title":"List Personas","heading":"","section":"API reference","crumbs":["API reference","Personas","List Personas"],"url":"/docs/api-reference/personas/list-personas","method":"GET","path":"/customer/v1/brands/{brand_id}/personas","text":"List the personas of the brand: who asks the scenarios, and from where.","keywords":"/customer/v1/brands/{brand_id}/personas brand_id"},{"kind":"route","title":"Create Persona","heading":"","section":"API reference","crumbs":["API reference","Personas","Create Persona"],"url":"/docs/api-reference/personas/create-persona","method":"POST","path":"/customer/v1/brands/{brand_id}/personas","text":"Create a persona. Give it to a topic with persona_id. The answer engines answer the scenarios of the topic as they answer a person in country and city. A name that the brand has gets 409 conflict, and a persona that the plan cannot hold gets 403 plan_limit_reached.","keywords":"/customer/v1/brands/{brand_id}/personas brand_id Idempotency-Key Any unique string, such as a UUID. Send the same key again to get the first answer and not a second write. The API keeps an answer for 24 hours."},{"kind":"route","title":"Delete Persona","heading":"","section":"API reference","crumbs":["API reference","Personas","Delete Persona"],"url":"/docs/api-reference/personas/delete-persona","method":"DELETE","path":"/customer/v1/brands/{brand_id}/personas/{persona_id}","text":"Delete a persona. A persona that a topic uses gets 409 conflict: give the topic another persona first.","keywords":"/customer/v1/brands/{brand_id}/personas/{persona_id} persona_id A persona of the brand. brand_id"},{"kind":"route","title":"Get Persona","heading":"","section":"API reference","crumbs":["API reference","Personas","Get Persona"],"url":"/docs/api-reference/personas/get-persona","method":"GET","path":"/customer/v1/brands/{brand_id}/personas/{persona_id}","text":"One persona of the brand.","keywords":"/customer/v1/brands/{brand_id}/personas/{persona_id} persona_id A persona of the brand. brand_id"},{"kind":"route","title":"Update Persona","heading":"","section":"API reference","crumbs":["API reference","Personas","Update Persona"],"url":"/docs/api-reference/personas/update-persona","method":"PATCH","path":"/customer/v1/brands/{brand_id}/personas/{persona_id}","text":"Change a persona. Only the fields of the body change, and properties replaces each section.","keywords":"/customer/v1/brands/{brand_id}/personas/{persona_id} persona_id A persona of the brand. brand_id Idempotency-Key Any unique string, such as a UUID. Send the same key again to get the first answer and not a second write. The API keeps an answer for 24 hours."},{"kind":"route","title":"List Products","heading":"","section":"API reference","crumbs":["API reference","Products","List Products"],"url":"/docs/api-reference/products/list-products","method":"GET","path":"/customer/v1/brands/{brand_id}/products","text":"The products of the brand, or of its competitors, that the shopping answers show. Each product gives its visibility and its recommendation, from 0 to 1, with their intervals of confidence, and the brands of the same answers.","keywords":"/customer/v1/brands/{brand_id}/products brand_id start_date YYYY-MM-DD (a day in UTC) or an ISO 8601 datetime. Both ends of the range are inclusive: an end date covers its whole day. end_date YYYY-MM-DD (a day in UTC) or an ISO 8601 datetime. Both ends of the range are inclusive: an end date covers its whole day. models The answer engines. With no list, the engines of the brand. topic_ids The topics of the brand. With no list, each topic. label The products of the brand, or of its competitors. competitor_id Only the products of this competitor. search category A value of the list of the categories. sort_by order page page_size"},{"kind":"route","title":"List Product Categories","heading":"","section":"API reference","crumbs":["API reference","Products","List Product Categories"],"url":"/docs/api-reference/products/list-product-categories","method":"GET","path":"/customer/v1/brands/{brand_id}/products/categories","text":"The categories of the products, for the filter category.","keywords":"/customer/v1/brands/{brand_id}/products/categories brand_id"},{"kind":"route","title":"Rank Products","heading":"","section":"API reference","crumbs":["API reference","Products","Rank Products"],"url":"/docs/api-reference/products/rank-products","method":"GET","path":"/customer/v1/brands/{brand_id}/products/ranking","text":"The products by visibility or by recommendation, with their price and rating. order=asc gives the lowest first: the products to improve.","keywords":"/customer/v1/brands/{brand_id}/products/ranking brand_id start_date YYYY-MM-DD (a day in UTC) or an ISO 8601 datetime. Both ends of the range are inclusive: an end date covers its whole day. end_date YYYY-MM-DD (a day in UTC) or an ISO 8601 datetime. Both ends of the range are inclusive: an end date covers its whole day. models The answer engines. With no list, the engines of the brand. topic_ids The topics of the brand. With no list, each topic. metric label category order desc: the highest first. asc: the lowest first. page page_size"},{"kind":"route","title":"List Retailers","heading":"","section":"API reference","crumbs":["API reference","Products","List Retailers"],"url":"/docs/api-reference/products/list-retailers","method":"GET","path":"/customer/v1/brands/{brand_id}/products/retailers","text":"The retailers whose offers the shopping answers show, the most offers first.","keywords":"/customer/v1/brands/{brand_id}/products/retailers brand_id start_date YYYY-MM-DD (a day in UTC) or an ISO 8601 datetime. Both ends of the range are inclusive: an end date covers its whole day. end_date YYYY-MM-DD (a day in UTC) or an ISO 8601 datetime. Both ends of the range are inclusive: an end date covers its whole day. models The answer engines. With no list, the engines of the brand. topic_ids The topics of the brand. With no list, each topic. top The retailers with the most offers."},{"kind":"route","title":"Product Stats","heading":"","section":"API reference","crumbs":["API reference","Products","Product Stats"],"url":"/docs/api-reference/products/product-stats","method":"GET","path":"/customer/v1/brands/{brand_id}/products/stats","text":"How many products of the brand and of its competitors the shopping answers show, and the products of each tracked competitor.","keywords":"/customer/v1/brands/{brand_id}/products/stats brand_id start_date YYYY-MM-DD (a day in UTC) or an ISO 8601 datetime. Both ends of the range are inclusive: an end date covers its whole day. end_date YYYY-MM-DD (a day in UTC) or an ISO 8601 datetime. Both ends of the range are inclusive: an end date covers its whole day. models The answer engines. With no list, the engines of the brand. topic_ids The topics of the brand. With no list, each topic."},{"kind":"route","title":"Get Product","heading":"","section":"API reference","crumbs":["API reference","Products","Get Product"],"url":"/docs/api-reference/products/get-product","method":"GET","path":"/customer/v1/brands/{brand_id}/products/{product_id}","text":"One product: its newest price and rating, the searches and the topics that show it, and the retailers of the product and of each variant.","keywords":"/customer/v1/brands/{brand_id}/products/{product_id} product_id A product of the brand. brand_id start_date YYYY-MM-DD (a day in UTC) or an ISO 8601 datetime. Both ends of the range are inclusive: an end date covers its whole day. end_date YYYY-MM-DD (a day in UTC) or an ISO 8601 datetime. Both ends of the range are inclusive: an end date covers its whole day. models The answer engines. With no list, the engines of the brand. topic_ids The topics of the brand. With no list, each topic."},{"kind":"route","title":"Product Co Mentions","heading":"","section":"API reference","crumbs":["API reference","Products","Product Co Mentions"],"url":"/docs/api-reference/products/product-co-mentions","method":"GET","path":"/customer/v1/brands/{brand_id}/products/{product_id}/co-mentions","text":"The products that the same answers show, the strongest first: the real competition of the product.","keywords":"/customer/v1/brands/{brand_id}/products/{product_id}/co-mentions product_id A product of the brand. brand_id start_date YYYY-MM-DD (a day in UTC) or an ISO 8601 datetime. Both ends of the range are inclusive: an end date covers its whole day. end_date YYYY-MM-DD (a day in UTC) or an ISO 8601 datetime. Both ends of the range are inclusive: an end date covers its whole day. models The answer engines. With no list, the engines of the brand. topic_ids The topics of the brand. With no list, each topic. metric sort_by page page_size"},{"kind":"route","title":"Product Perception","heading":"","section":"API reference","crumbs":["API reference","Products","Product Perception"],"url":"/docs/api-reference/products/product-perception","method":"GET","path":"/customer/v1/brands/{brand_id}/products/{product_id}/perception","text":"What the answers say about the product: positive, neutral and negative, the most frequent first.","keywords":"/customer/v1/brands/{brand_id}/products/{product_id}/perception product_id A product of the brand. brand_id start_date YYYY-MM-DD (a day in UTC) or an ISO 8601 datetime. Both ends of the range are inclusive: an end date covers its whole day. end_date YYYY-MM-DD (a day in UTC) or an ISO 8601 datetime. Both ends of the range are inclusive: an end date covers its whole day. models The answer engines. With no list, the engines of the brand. topic_ids The topics of the brand. With no list, each topic. page The page of each sentiment. page_size"},{"kind":"route","title":"Product Series","heading":"","section":"API reference","crumbs":["API reference","Products","Product Series"],"url":"/docs/api-reference/products/product-series","method":"GET","path":"/customer/v1/brands/{brand_id}/products/{product_id}/series","text":"The product and the products of the same answers, period by period, with the visibility and the recommendation of each one.","keywords":"/customer/v1/brands/{brand_id}/products/{product_id}/series product_id A product of the brand. brand_id start_date YYYY-MM-DD (a day in UTC) or an ISO 8601 datetime. Both ends of the range are inclusive: an end date covers its whole day. end_date YYYY-MM-DD (a day in UTC) or an ISO 8601 datetime. Both ends of the range are inclusive: an end date covers its whole day. models The answer engines. With no list, the engines of the brand. topic_ids The topics of the brand. With no list, each topic. granularity"},{"kind":"route","title":"Product Skus","heading":"","section":"API reference","crumbs":["API reference","Products","Product Skus"],"url":"/docs/api-reference/products/product-skus","method":"GET","path":"/customer/v1/brands/{brand_id}/products/{product_id}/skus","text":"The visibility of each variant (SKU) of the product, such as a color.","keywords":"/customer/v1/brands/{brand_id}/products/{product_id}/skus product_id A product of the brand. brand_id start_date YYYY-MM-DD (a day in UTC) or an ISO 8601 datetime. Both ends of the range are inclusive: an end date covers its whole day. end_date YYYY-MM-DD (a day in UTC) or an ISO 8601 datetime. Both ends of the range are inclusive: an end date covers its whole day. models The answer engines. With no list, the engines of the brand. topic_ids The topics of the brand. With no list, each topic."},{"kind":"route","title":"Product Sources","heading":"","section":"API reference","crumbs":["API reference","Products","Product Sources"],"url":"/docs/api-reference/products/product-sources","method":"GET","path":"/customer/v1/brands/{brand_id}/products/{product_id}/sources","text":"The pages that the answers that show the product read and cite.","keywords":"/customer/v1/brands/{brand_id}/products/{product_id}/sources product_id A product of the brand. brand_id start_date YYYY-MM-DD (a day in UTC) or an ISO 8601 datetime. Both ends of the range are inclusive: an end date covers its whole day. end_date YYYY-MM-DD (a day in UTC) or an ISO 8601 datetime. Both ends of the range are inclusive: an end date covers its whole day. models The answer engines. With no list, the engines of the brand. topic_ids The topics of the brand. With no list, each topic. sentiment sort_by order page page_size"},{"kind":"route","title":"Citation Report","heading":"","section":"API reference","crumbs":["API reference","Citations","Citation Report"],"url":"/docs/api-reference/citations/citation-report","method":"GET","path":"/customer/v1/brands/{brand_id}/reports/citations","text":"How often the answer engines cite a source, period by period. One row for each period, with change against the previous period. With domains, each domain also gets its own series. info.summary gives each occurrence of the range and the share of each kind of owner of the pages.","keywords":"/customer/v1/brands/{brand_id}/reports/citations brand_id start_date YYYY-MM-DD (a day in UTC) or an ISO 8601 datetime. Both ends of the range are inclusive: an end date covers its whole day. end_date YYYY-MM-DD (a day in UTC) or an ISO 8601 datetime. Both ends of the range are inclusive: an end date covers its whole day. models The answer engines. With no list, each engine. topic_ids The topics of the brand. With no list, each topic. ownership Who owns the page, for example competitor_owned. content_types What the page is, for example review_editorial. sentiment How the page speaks of the brand. first_party true: only the pages of the brand. false: only the pages of others. No value: each page. granularity The length of a period: a day, a week (from Monday), or a month. domains Root domains, for example g2.com. Each one also gets its own series, beside the series of each domain together."},{"kind":"route","title":"Fact Check Report","heading":"","section":"API reference","crumbs":["API reference","Fact check","Fact Check Report"],"url":"/docs/api-reference/fact-check/fact-check-report","method":"GET","path":"/customer/v1/brands/{brand_id}/reports/fact-check","text":"The share of the answers that give the expected facts, period by period. Each period gives its interval of confidence, change and significant. info.summary gives the whole range and the last period.","keywords":"/customer/v1/brands/{brand_id}/reports/fact-check brand_id start_date YYYY-MM-DD (a day in UTC) or an ISO 8601 datetime. Both ends of the range are inclusive: an end date covers its whole day. end_date YYYY-MM-DD (a day in UTC) or an ISO 8601 datetime. Both ends of the range are inclusive: an end date covers its whole day. models The answer engines. With no list, each engine. topic_ids The fact_checker topics. With no list, each one. granularity The length of a period: a day, a week (from Monday), or a month."},{"kind":"route","title":"Fact Check By Model","heading":"","section":"API reference","crumbs":["API reference","Fact check","Fact Check By Model"],"url":"/docs/api-reference/fact-check/fact-check-by-model","method":"GET","path":"/customer/v1/brands/{brand_id}/reports/fact-check/models","text":"The share of the answers that pass, for each answer engine. The engine with the lowest share comes first.","keywords":"/customer/v1/brands/{brand_id}/reports/fact-check/models brand_id start_date YYYY-MM-DD (a day in UTC) or an ISO 8601 datetime. Both ends of the range are inclusive: an end date covers its whole day. end_date YYYY-MM-DD (a day in UTC) or an ISO 8601 datetime. Both ends of the range are inclusive: an end date covers its whole day. models The answer engines. With no list, each engine. topic_ids The fact_checker topics. With no list, each one."},{"kind":"route","title":"Fact Check By Topic","heading":"","section":"API reference","crumbs":["API reference","Fact check","Fact Check By Topic"],"url":"/docs/api-reference/fact-check/fact-check-by-topic","method":"GET","path":"/customer/v1/brands/{brand_id}/reports/fact-check/topics","text":"The share of the answers that pass, for each topic and each of its scenarios. The scenarios with the lowest share come first: fix those facts first.","keywords":"/customer/v1/brands/{brand_id}/reports/fact-check/topics brand_id start_date YYYY-MM-DD (a day in UTC) or an ISO 8601 datetime. Both ends of the range are inclusive: an end date covers its whole day. end_date YYYY-MM-DD (a day in UTC) or an ISO 8601 datetime. Both ends of the range are inclusive: an end date covers its whole day. models The answer engines. With no list, each engine. topic_ids The fact_checker topics. With no list, each one."},{"kind":"route","title":"Products Report","heading":"","section":"API reference","crumbs":["API reference","Products","Products Report"],"url":"/docs/api-reference/products/products-report","method":"GET","path":"/customer/v1/brands/{brand_id}/reports/products","text":"The visibility or the recommendation of the products, period by period. One row for each period and each product: the top products that appear in the most answers.","keywords":"/customer/v1/brands/{brand_id}/reports/products brand_id start_date YYYY-MM-DD (a day in UTC) or an ISO 8601 datetime. Both ends of the range are inclusive: an end date covers its whole day. end_date YYYY-MM-DD (a day in UTC) or an ISO 8601 datetime. Both ends of the range are inclusive: an end date covers its whole day. models The answer engines. With no list, the engines of the brand. topic_ids The topics of the brand. With no list, each topic. metric granularity label With no label, the products of the brand and of its competitors. category top The products that appear in the most answers."},{"kind":"route","title":"Recommendation Report","heading":"","section":"API reference","crumbs":["API reference","Reports","Recommendation Report"],"url":"/docs/api-reference/reports/recommendation-report","method":"GET","path":"/customer/v1/brands/{brand_id}/reports/recommendation","text":"How often the answer engines recommend the brand, period by period. The same shape as the visibility report, with recommendation in place of visibility. The value of a period reads the same sliding window as the visibility report.","keywords":"/customer/v1/brands/{brand_id}/reports/recommendation brand_id start_date YYYY-MM-DD (a day in UTC) or an ISO 8601 datetime. Both ends of the range are inclusive: an end date covers its whole day. end_date YYYY-MM-DD (a day in UTC) or an ISO 8601 datetime. Both ends of the range are inclusive: an end date covers its whole day. granularity The length of a period: a day, a week (from Monday), or a month. A week is the average of its days. A month is the average of the days of its last week. models The answer engines. With no list, the engines of the brand. topic_ids The topics of the brand. With no list, each topic. group_by model: one series for each answer engine, in place of one series for each engine together. competitors Also give the series of each competitor."},{"kind":"route","title":"Recommendation By Topic","heading":"","section":"API reference","crumbs":["API reference","Reports","Recommendation By Topic"],"url":"/docs/api-reference/reports/recommendation-by-topic","method":"GET","path":"/customer/v1/brands/{brand_id}/reports/recommendation/topics","text":"How often the answer engines recommend the brand in each topic. The same shape as the visibility by topic, with recommendation in place of visibility.","keywords":"/customer/v1/brands/{brand_id}/reports/recommendation/topics brand_id start_date YYYY-MM-DD (a day in UTC) or an ISO 8601 datetime. Both ends of the range are inclusive: an end date covers its whole day. end_date YYYY-MM-DD (a day in UTC) or an ISO 8601 datetime. Both ends of the range are inclusive: an end date covers its whole day. models The answer engines. With no list, the engines of the brand. topic_ids The topics of the brand. With no list, each topic. scenarios Also give the value of each scenario of the topic. Without them, the report is faster."},{"kind":"route","title":"Sentiment Report","heading":"","section":"API reference","crumbs":["API reference","Sentiment","Sentiment Report"],"url":"/docs/api-reference/sentiment/sentiment-report","method":"GET","path":"/customer/v1/brands/{brand_id}/reports/sentiment","text":"The statements about the brand by sentiment, period by period. Each row gives the positive, neutral and negative statements as counts and as shares, and the change of the positive share. info.summary gives the shares of the whole range and the last period.","keywords":"/customer/v1/brands/{brand_id}/reports/sentiment brand_id start_date YYYY-MM-DD (a day in UTC) or an ISO 8601 datetime. Both ends of the range are inclusive: an end date covers its whole day. end_date YYYY-MM-DD (a day in UTC) or an ISO 8601 datetime. Both ends of the range are inclusive: an end date covers its whole day. models The answer engines. With no list, each engine. topic_ids The topics of the brand. With no list, each topic. granularity The length of a period: a day, a week (from Monday), or a month."},{"kind":"route","title":"Sentiment By Topic","heading":"","section":"API reference","crumbs":["API reference","Sentiment","Sentiment By Topic"],"url":"/docs/api-reference/sentiment/sentiment-by-topic","method":"GET","path":"/customer/v1/brands/{brand_id}/reports/sentiment/topics","text":"The sentiment of each topic and of each scenario of the topic. Each topic gives the sentiment of the statements of the answers, and the sentiment of the pages that the answers cite.","keywords":"/customer/v1/brands/{brand_id}/reports/sentiment/topics brand_id start_date YYYY-MM-DD (a day in UTC) or an ISO 8601 datetime. Both ends of the range are inclusive: an end date covers its whole day. end_date YYYY-MM-DD (a day in UTC) or an ISO 8601 datetime. Both ends of the range are inclusive: an end date covers its whole day. models The answer engines. With no list, each engine. topic_ids The topics of the brand. With no list, each topic."},{"kind":"route","title":"Visibility Report","heading":"","section":"API reference","crumbs":["API reference","Reports","Visibility Report"],"url":"/docs/api-reference/reports/visibility-report","method":"GET","path":"/customer/v1/brands/{brand_id}/reports/visibility","text":"How often the answer engines mention the brand, period by period. One row for each period and each brand: the brand of the path and, with competitors, each competitor. Each row has its interval of confidence, and change and significant against the previous period. info.summary gives the last period of the brand and its rank. The value of a day is not the share of the answers of that day only. It reads a sliding window: the newest answers up to the end of that day, until the confidence interval is narrow enough. A week is the average of its days, and a month is the average of the days of its last week. The window does not read an answer before start_date. The guide of the visibility reports gives an example.","keywords":"/customer/v1/brands/{brand_id}/reports/visibility brand_id start_date YYYY-MM-DD (a day in UTC) or an ISO 8601 datetime. Both ends of the range are inclusive: an end date covers its whole day. end_date YYYY-MM-DD (a day in UTC) or an ISO 8601 datetime. Both ends of the range are inclusive: an end date covers its whole day. granularity The length of a period: a day, a week (from Monday), or a month. A week is the average of its days. A month is the average of the days of its last week. models The answer engines. With no list, the engines of the brand. topic_ids The topics of the brand. With no list, each topic. group_by model: one series for each answer engine, in place of one series for each engine together. competitors Also give the series of each competitor."},{"kind":"route","title":"Visibility By Topic","heading":"","section":"API reference","crumbs":["API reference","Reports","Visibility By Topic"],"url":"/docs/api-reference/reports/visibility-by-topic","method":"GET","path":"/customer/v1/brands/{brand_id}/reports/visibility/topics","text":"How often the answer engines mention the brand in each topic. One row for each topic that has answers in the range, with the value of the brand, the five competitors with the highest value, and the same values for each scenario of the topic. These are the numbers of the Topics page of the dashboard.","keywords":"/customer/v1/brands/{brand_id}/reports/visibility/topics brand_id start_date YYYY-MM-DD (a day in UTC) or an ISO 8601 datetime. Both ends of the range are inclusive: an end date covers its whole day. end_date YYYY-MM-DD (a day in UTC) or an ISO 8601 datetime. Both ends of the range are inclusive: an end date covers its whole day. models The answer engines. With no list, the engines of the brand. topic_ids The topics of the brand. With no list, each topic. scenarios Also give the value of each scenario of the topic. Without them, the report is faster."},{"kind":"route","title":"List Scenarios","heading":"","section":"API reference","crumbs":["API reference","Scenarios","List Scenarios"],"url":"/docs/api-reference/scenarios/list-scenarios","method":"GET","path":"/customer/v1/brands/{brand_id}/scenarios","text":"List the scenarios of the brand, the newest first. Each scenario gives its topic in topic_id. Use topic_ids to get the scenarios of some topics, and status to get only the active or only the paused scenarios.","keywords":"/customer/v1/brands/{brand_id}/scenarios brand_id topic_ids The topics of the brand. With no list, each topic. status Only the scenarios with this status. A scenario of a paused topic is paused. With no status, each scenario."},{"kind":"route","title":"Set Scenarios Status","heading":"","section":"API reference","crumbs":["API reference","Scenarios","Set Scenarios Status"],"url":"/docs/api-reference/scenarios/set-scenarios-status","method":"PATCH","path":"/customer/v1/brands/{brand_id}/scenarios","text":"Pause or activate up to 500 scenarios of the brand in one request. A scenario of another brand is in rejected, and the other scenarios change.","keywords":"/customer/v1/brands/{brand_id}/scenarios brand_id Idempotency-Key Any unique string, such as a UUID. Send the same key again to get the first answer and not a second write. The API keeps an answer for 24 hours."},{"kind":"route","title":"Create Scenarios","heading":"","section":"API reference","crumbs":["API reference","Scenarios","Create Scenarios"],"url":"/docs/api-reference/scenarios/create-scenarios","method":"POST","path":"/customer/v1/brands/{brand_id}/scenarios","text":"Create up to 100 scenarios, in one or more topics of the brand. Each item is written or rejected on its own: a scenario of a topic of another brand, or a scenario that its topic already has, is in rejected with its index, and the other items are written. When the plan of the account cannot hold the new scenarios, nothing is written (403 plan_limit_reached). A new scenario runs with the next run of its topic.","keywords":"/customer/v1/brands/{brand_id}/scenarios brand_id Idempotency-Key Any unique string, such as a UUID. Send the same key again to get the first answer and not a second write. The API keeps an answer for 24 hours."},{"kind":"route","title":"Delete Scenario","heading":"","section":"API reference","crumbs":["API reference","Scenarios","Delete Scenario"],"url":"/docs/api-reference/scenarios/delete-scenario","method":"DELETE","path":"/customer/v1/brands/{brand_id}/scenarios/{scenario_id}","text":"Delete one scenario. It stops running, and restore brings it back. A delete is idempotent in itself, thus it takes no Idempotency-Key: a second delete of the same scenario gets 404.","keywords":"/customer/v1/brands/{brand_id}/scenarios/{scenario_id} scenario_id A scenario of the brand. brand_id"},{"kind":"route","title":"Get Scenario","heading":"","section":"API reference","crumbs":["API reference","Scenarios","Get Scenario"],"url":"/docs/api-reference/scenarios/get-scenario","method":"GET","path":"/customer/v1/brands/{brand_id}/scenarios/{scenario_id}","text":"One scenario of the brand.","keywords":"/customer/v1/brands/{brand_id}/scenarios/{scenario_id} scenario_id A scenario of the brand. brand_id"},{"kind":"route","title":"Update Scenario","heading":"","section":"API reference","crumbs":["API reference","Scenarios","Update Scenario"],"url":"/docs/api-reference/scenarios/update-scenario","method":"PATCH","path":"/customer/v1/brands/{brand_id}/scenarios/{scenario_id}","text":"Change the text or the status of one scenario. Only the fields of the body change. A new text starts a new series of answers: the answers of the old text stay in the reports.","keywords":"/customer/v1/brands/{brand_id}/scenarios/{scenario_id} scenario_id A scenario of the brand. brand_id Idempotency-Key Any unique string, such as a UUID. Send the same key again to get the first answer and not a second write. The API keeps an answer for 24 hours."},{"kind":"route","title":"Scenario Recommendation Report","heading":"","section":"API reference","crumbs":["API reference","Scenarios","Scenario Recommendation Report"],"url":"/docs/api-reference/scenarios/scenario-recommendation-report","method":"GET","path":"/customer/v1/brands/{brand_id}/scenarios/{scenario_id}/reports/recommendation","text":"How often the answer engines recommend the brand in one scenario, day by day. The same shape as the visibility report of the scenario, with recommendation in place of visibility.","keywords":"/customer/v1/brands/{brand_id}/scenarios/{scenario_id}/reports/recommendation scenario_id A scenario of the brand. brand_id start_date YYYY-MM-DD (a day in UTC) or an ISO 8601 datetime. Both ends of the range are inclusive: an end date covers its whole day. end_date YYYY-MM-DD (a day in UTC) or an ISO 8601 datetime. Both ends of the range are inclusive: an end date covers its whole day. models The answer engines. With no list, the engines of the brand. competitors Also give the series of each competitor."},{"kind":"route","title":"Scenario Visibility Report","heading":"","section":"API reference","crumbs":["API reference","Scenarios","Scenario Visibility Report"],"url":"/docs/api-reference/scenarios/scenario-visibility-report","method":"GET","path":"/customer/v1/brands/{brand_id}/scenarios/{scenario_id}/reports/visibility","text":"How often the answer engines mention the brand in one scenario, day by day. The shape of the visibility report of the brand: one row for each day and each brand, with change and significant against the previous day with answers. info.summary gives the last day of the brand and its rank.","keywords":"/customer/v1/brands/{brand_id}/scenarios/{scenario_id}/reports/visibility scenario_id A scenario of the brand. brand_id start_date YYYY-MM-DD (a day in UTC) or an ISO 8601 datetime. Both ends of the range are inclusive: an end date covers its whole day. end_date YYYY-MM-DD (a day in UTC) or an ISO 8601 datetime. Both ends of the range are inclusive: an end date covers its whole day. models The answer engines. With no list, the engines of the brand. competitors Also give the series of each competitor."},{"kind":"route","title":"Restore Scenario","heading":"","section":"API reference","crumbs":["API reference","Scenarios","Restore Scenario"],"url":"/docs/api-reference/scenarios/restore-scenario","method":"POST","path":"/customer/v1/brands/{brand_id}/scenarios/{scenario_id}/restore","text":"Bring back a deleted scenario. Its answers come back to the reports. A scenario that is not deleted, or a scenario of a deleted topic, gets 409 conflict.","keywords":"/customer/v1/brands/{brand_id}/scenarios/{scenario_id}/restore scenario_id A deleted scenario of the brand. brand_id"},{"kind":"route","title":"List Goals","heading":"","section":"API reference","crumbs":["API reference","Scorecard","List Goals"],"url":"/docs/api-reference/scorecard/list-goals","method":"GET","path":"/customer/v1/brands/{brand_id}/scorecard/goals","text":"The goals of the board, in the order of the board, each with the datapoints of its last 90 days. status says if the goal is reached.","keywords":"/customer/v1/brands/{brand_id}/scorecard/goals brand_id"},{"kind":"route","title":"List Archived Goals","heading":"","section":"API reference","crumbs":["API reference","Scorecard","List Archived Goals"],"url":"/docs/api-reference/scorecard/list-archived-goals","method":"GET","path":"/customer/v1/brands/{brand_id}/scorecard/goals/archived","text":"The goals that left the board. restore brings one back.","keywords":"/customer/v1/brands/{brand_id}/scorecard/goals/archived brand_id"},{"kind":"route","title":"Archive Goal","heading":"","section":"API reference","crumbs":["API reference","Scorecard","Archive Goal"],"url":"/docs/api-reference/scorecard/archive-goal","method":"DELETE","path":"/customer/v1/brands/{brand_id}/scorecard/goals/{goal_id}","text":"Take a goal off the board. It stops running, and restore brings it back.","keywords":"/customer/v1/brands/{brand_id}/scorecard/goals/{goal_id} goal_id A goal of the Business Scorecard of the brand. brand_id"},{"kind":"route","title":"Get Goal","heading":"","section":"API reference","crumbs":["API reference","Scorecard","Get Goal"],"url":"/docs/api-reference/scorecard/get-goal","method":"GET","path":"/customer/v1/brands/{brand_id}/scorecard/goals/{goal_id}","text":"One goal, its datapoints, and its ten newest runs. A failed run tells why a card stopped moving.","keywords":"/customer/v1/brands/{brand_id}/scorecard/goals/{goal_id} goal_id A goal of the Business Scorecard of the brand. brand_id"},{"kind":"route","title":"Update Goal","heading":"","section":"API reference","crumbs":["API reference","Scorecard","Update Goal"],"url":"/docs/api-reference/scorecard/update-goal","method":"PATCH","path":"/customer/v1/brands/{brand_id}/scorecard/goals/{goal_id}","text":"Change the title, the category or the dimensions of a goal. It does not change what the goal computes, thus it runs nothing.","keywords":"/customer/v1/brands/{brand_id}/scorecard/goals/{goal_id} goal_id A goal of the Business Scorecard of the brand. brand_id Idempotency-Key Any unique string, such as a UUID. Send the same key again to get the first answer and not a second write. The API keeps an answer for 24 hours."},{"kind":"route","title":"Restore Goal","heading":"","section":"API reference","crumbs":["API reference","Scorecard","Restore Goal"],"url":"/docs/api-reference/scorecard/restore-goal","method":"POST","path":"/customer/v1/brands/{brand_id}/scorecard/goals/{goal_id}/restore","text":"Bring an archived goal back to the board. A goal on the board gets 404.","keywords":"/customer/v1/brands/{brand_id}/scorecard/goals/{goal_id}/restore goal_id A goal of the Business Scorecard of the brand. brand_id"},{"kind":"route","title":"Swot Report","heading":"","section":"API reference","crumbs":["API reference","SWOT","Swot Report"],"url":"/docs/api-reference/swot/swot-report","method":"GET","path":"/customer/v1/brands/{brand_id}/swot","text":"What the answers say is strong and weak about the brand and about each competitor, as strengths, weaknesses, opportunities and threats. Each statement is a group of statements that say the same thing.","keywords":"/customer/v1/brands/{brand_id}/swot brand_id start_date YYYY-MM-DD (a day in UTC) or an ISO 8601 datetime. Both ends of the range are inclusive: an end date covers its whole day. end_date YYYY-MM-DD (a day in UTC) or an ISO 8601 datetime. Both ends of the range are inclusive: an end date covers its whole day. models The answer engines. With no list, each engine. topic_ids competitor_ids Only these competitors, and the brand. types"},{"kind":"route","title":"Swot Statement","heading":"","section":"API reference","crumbs":["API reference","SWOT","Swot Statement"],"url":"/docs/api-reference/swot/swot-statement","method":"GET","path":"/customer/v1/brands/{brand_id}/swot/{statement_id}","text":"The statements of the answers that a SWOT statement groups, the newest first, each with its answer.","keywords":"/customer/v1/brands/{brand_id}/swot/{statement_id} statement_id A statement of the SWOT of the brand. brand_id start_date YYYY-MM-DD (a day in UTC) or an ISO 8601 datetime. Both ends of the range are inclusive: an end date covers its whole day. end_date YYYY-MM-DD (a day in UTC) or an ISO 8601 datetime. Both ends of the range are inclusive: an end date covers its whole day. models The answer engines. With no list, each engine. topic_ids competitor_ids Only these competitors, and the brand. types page page_size"},{"kind":"route","title":"List Tags","heading":"","section":"API reference","crumbs":["API reference","Tags","List Tags"],"url":"/docs/api-reference/tags/list-tags","method":"GET","path":"/customer/v1/brands/{brand_id}/tags","text":"List the tags of the topics of the brand.","keywords":"/customer/v1/brands/{brand_id}/tags brand_id"},{"kind":"route","title":"Create Tag","heading":"","section":"API reference","crumbs":["API reference","Tags","Create Tag"],"url":"/docs/api-reference/tags/create-tag","method":"POST","path":"/customer/v1/brands/{brand_id}/tags","text":"Create a tag. A name that the brand already has gets 409 conflict.","keywords":"/customer/v1/brands/{brand_id}/tags brand_id Idempotency-Key Any unique string, such as a UUID. Send the same key again to get the first answer and not a second write. The API keeps an answer for 24 hours."},{"kind":"route","title":"Delete Tag","heading":"","section":"API reference","crumbs":["API reference","Tags","Delete Tag"],"url":"/docs/api-reference/tags/delete-tag","method":"DELETE","path":"/customer/v1/brands/{brand_id}/tags/{tag_id}","text":"Delete a tag. It goes away from each topic.","keywords":"/customer/v1/brands/{brand_id}/tags/{tag_id} tag_id A tag of the brand. brand_id"},{"kind":"route","title":"Update Tag","heading":"","section":"API reference","crumbs":["API reference","Tags","Update Tag"],"url":"/docs/api-reference/tags/update-tag","method":"PATCH","path":"/customer/v1/brands/{brand_id}/tags/{tag_id}","text":"Change the name or the color of a tag.","keywords":"/customer/v1/brands/{brand_id}/tags/{tag_id} tag_id A tag of the brand. brand_id Idempotency-Key Any unique string, such as a UUID. Send the same key again to get the first answer and not a second write. The API keeps an answer for 24 hours."},{"kind":"route","title":"List Templates","heading":"","section":"API reference","crumbs":["API reference","Templates","List Templates"],"url":"/docs/api-reference/templates/list-templates","method":"GET","path":"/customer/v1/brands/{brand_id}/templates","text":"The templates of the brand, the newest first. A brand with no template gets a copy of the templates of the platform.","keywords":"/customer/v1/brands/{brand_id}/templates brand_id type search A part of the title. page page_size"},{"kind":"route","title":"Get Template","heading":"","section":"API reference","crumbs":["API reference","Templates","Get Template"],"url":"/docs/api-reference/templates/get-template","method":"GET","path":"/customer/v1/brands/{brand_id}/templates/{template_id}","text":"One template, with its text.","keywords":"/customer/v1/brands/{brand_id}/templates/{template_id} template_id A template of the brand. brand_id"},{"kind":"route","title":"List Topics","heading":"","section":"API reference","crumbs":["API reference","Topics","List Topics"],"url":"/docs/api-reference/topics/list-topics","method":"GET","path":"/customer/v1/brands/{brand_id}/topics","text":"List the topics of a brand, each with its scenarios.","keywords":"/customer/v1/brands/{brand_id}/topics brand_id"},{"kind":"route","title":"Set Topics Status","heading":"","section":"API reference","crumbs":["API reference","Topics","Set Topics Status"],"url":"/docs/api-reference/topics/set-topics-status","method":"PATCH","path":"/customer/v1/brands/{brand_id}/topics","text":"Pause or activate up to 200 topics of the brand in one request. A paused topic does not run, and each of its scenarios counts as paused.","keywords":"/customer/v1/brands/{brand_id}/topics brand_id Idempotency-Key Any unique string, such as a UUID. Send the same key again to get the first answer and not a second write. The API keeps an answer for 24 hours."},{"kind":"route","title":"Create Topics","heading":"","section":"API reference","crumbs":["API reference","Topics","Create Topics"],"url":"/docs/api-reference/topics/create-topics","method":"POST","path":"/customer/v1/brands/{brand_id}/topics","text":"Create up to 20 topics, each with up to 100 scenarios, in one request. Each topic is written or rejected on its own. A topic with the name, the type and the language of another topic of the brand is rejected with conflict. A topic that the plan of the account cannot hold is rejected with plan_limit_reached. A new topic runs with the next run of the brand.","keywords":"/customer/v1/brands/{brand_id}/topics brand_id Idempotency-Key Any unique string, such as a UUID. Send the same key again to get the first answer and not a second write. The API keeps an answer for 24 hours."},{"kind":"route","title":"Delete Topic","heading":"","section":"API reference","crumbs":["API reference","Topics","Delete Topic"],"url":"/docs/api-reference/topics/delete-topic","method":"DELETE","path":"/customer/v1/brands/{brand_id}/topics/{topic_id}","text":"Delete one topic. It stops running, and restore brings it back.","keywords":"/customer/v1/brands/{brand_id}/topics/{topic_id} topic_id A topic of the brand. brand_id"},{"kind":"route","title":"Get Topic","heading":"","section":"API reference","crumbs":["API reference","Topics","Get Topic"],"url":"/docs/api-reference/topics/get-topic","method":"GET","path":"/customer/v1/brands/{brand_id}/topics/{topic_id}","text":"One topic of the brand, with its scenarios.","keywords":"/customer/v1/brands/{brand_id}/topics/{topic_id} topic_id A topic of the brand. brand_id"},{"kind":"route","title":"Update Topic","heading":"","section":"API reference","crumbs":["API reference","Topics","Update Topic"],"url":"/docs/api-reference/topics/update-topic","method":"PATCH","path":"/customer/v1/brands/{brand_id}/topics/{topic_id}","text":"Change the name, the status, the language, the persona or the engines. Only the fields of the body change. answer_engines: [] goes back to the engines of the brand. An engine that the plan does not hold gets 403 plan_limit_reached.","keywords":"/customer/v1/brands/{brand_id}/topics/{topic_id} topic_id A topic of the brand. brand_id Idempotency-Key Any unique string, such as a UUID. Send the same key again to get the first answer and not a second write. The API keeps an answer for 24 hours."},{"kind":"route","title":"Restore Topic","heading":"","section":"API reference","crumbs":["API reference","Topics","Restore Topic"],"url":"/docs/api-reference/topics/restore-topic","method":"POST","path":"/customer/v1/brands/{brand_id}/topics/{topic_id}/restore","text":"Bring back a deleted topic, with its scenarios. A topic that is not deleted gets 409 conflict.","keywords":"/customer/v1/brands/{brand_id}/topics/{topic_id}/restore topic_id A deleted topic of the brand. brand_id"},{"kind":"route","title":"Set Topic Tags","heading":"","section":"API reference","crumbs":["API reference","Tags","Set Topic Tags"],"url":"/docs/api-reference/tags/set-topic-tags","method":"PUT","path":"/customer/v1/brands/{brand_id}/topics/{topic_id}/tags","text":"Give a topic exactly these tags. [] removes each tag from the topic. A PUT gives the same result each time, thus it takes no Idempotency-Key. A tag of another brand gets 400.","keywords":"/customer/v1/brands/{brand_id}/topics/{topic_id}/tags topic_id A topic of the brand. brand_id"},{"kind":"route","title":"List Tracked Url Groups","heading":"","section":"API reference","crumbs":["API reference","Tracked URLs","List Tracked Url Groups"],"url":"/docs/api-reference/tracked-urls/list-tracked-url-groups","method":"GET","path":"/customer/v1/brands/{brand_id}/tracked-url-groups","text":"The tracked URL groups of the brand, each with its number of pages. A tracked URL group is a folder for the tracked URLs, such as \"Blog\" or \"Q4 campaign\". Filter the tracked URLs of one group with group_id. A tracked URL group is not a brand group: a brand group joins competitors into one brand.","keywords":"/customer/v1/brands/{brand_id}/tracked-url-groups brand_id"},{"kind":"route","title":"Create Tracked Url Group","heading":"","section":"API reference","crumbs":["API reference","Tracked URLs","Create Tracked Url Group"],"url":"/docs/api-reference/tracked-urls/create-tracked-url-group","method":"POST","path":"/customer/v1/brands/{brand_id}/tracked-url-groups","text":"Create a tracked URL group: a folder for the tracked URLs of the brand. Put a page in the group with group_id when you track the page or when you change it. A name that the brand has gets 409 conflict. A tracked URL group is not a brand group: a brand group joins competitors into one brand.","keywords":"/customer/v1/brands/{brand_id}/tracked-url-groups brand_id Idempotency-Key Any unique string, such as a UUID. Send the same key again to get the first answer and not a second write. The API keeps an answer for 24 hours."},{"kind":"route","title":"Delete Tracked Url Group","heading":"","section":"API reference","crumbs":["API reference","Tracked URLs","Delete Tracked Url Group"],"url":"/docs/api-reference/tracked-urls/delete-tracked-url-group","method":"DELETE","path":"/customer/v1/brands/{brand_id}/tracked-url-groups/{group_id}","text":"Delete a tracked URL group. Its pages stay tracked, in no group.","keywords":"/customer/v1/brands/{brand_id}/tracked-url-groups/{group_id} group_id A tracked URL group of the brand. brand_id"},{"kind":"route","title":"Rename Tracked Url Group","heading":"","section":"API reference","crumbs":["API reference","Tracked URLs","Rename Tracked Url Group"],"url":"/docs/api-reference/tracked-urls/rename-tracked-url-group","method":"PATCH","path":"/customer/v1/brands/{brand_id}/tracked-url-groups/{group_id}","text":"Give a tracked URL group a new name. Its pages stay in the group.","keywords":"/customer/v1/brands/{brand_id}/tracked-url-groups/{group_id} group_id A tracked URL group of the brand. brand_id Idempotency-Key Any unique string, such as a UUID. Send the same key again to get the first answer and not a second write. The API keeps an answer for 24 hours."},{"kind":"route","title":"List Tracked Urls","heading":"","section":"API reference","crumbs":["API reference","Tracked URLs","List Tracked Urls"],"url":"/docs/api-reference/tracked-urls/list-tracked-urls","method":"GET","path":"/customer/v1/brands/{brand_id}/tracked-urls","text":"The pages that the brand tracks, with the answers of the range that cite each one. info gives the citations of each page together.","keywords":"/customer/v1/brands/{brand_id}/tracked-urls brand_id start_date YYYY-MM-DD (a day in UTC) or an ISO 8601 datetime. Both ends of the range are inclusive: an end date covers its whole day. end_date YYYY-MM-DD (a day in UTC) or an ISO 8601 datetime. Both ends of the range are inclusive: an end date covers its whole day. models The answer engines. With no list, each engine. topic_ids The topics of the brand. With no list, each topic. group_id Only the pages of this tracked URL group. ungrouped Only the pages of no tracked URL group. search A part of the address or the title. sort_by order page page_size"},{"kind":"route","title":"Create Tracked Urls","heading":"","section":"API reference","crumbs":["API reference","Tracked URLs","Create Tracked Urls"],"url":"/docs/api-reference/tracked-urls/create-tracked-urls","method":"POST","path":"/customer/v1/brands/{brand_id}/tracked-urls","text":"Track up to 200 pages. A page with no title gets the title of the page. An address that the brand already tracks is rejected with conflict. The API compares the normal form of the addresses, thus https://acme.com/a/ and https://acme.com/a are one page.","keywords":"/customer/v1/brands/{brand_id}/tracked-urls brand_id Idempotency-Key Any unique string, such as a UUID. Send the same key again to get the first answer and not a second write. The API keeps an answer for 24 hours."},{"kind":"route","title":"Delete Tracked Url","heading":"","section":"API reference","crumbs":["API reference","Tracked URLs","Delete Tracked Url"],"url":"/docs/api-reference/tracked-urls/delete-tracked-url","method":"DELETE","path":"/customer/v1/brands/{brand_id}/tracked-urls/{tracked_url_id}","text":"Stop tracking a page.","keywords":"/customer/v1/brands/{brand_id}/tracked-urls/{tracked_url_id} tracked_url_id A tracked URL of the brand. brand_id"},{"kind":"route","title":"Update Tracked Url","heading":"","section":"API reference","crumbs":["API reference","Tracked URLs","Update Tracked Url"],"url":"/docs/api-reference/tracked-urls/update-tracked-url","method":"PATCH","path":"/customer/v1/brands/{brand_id}/tracked-urls/{tracked_url_id}","text":"Change the address, the title or the group of a page. group_id: null takes the page out of its group.","keywords":"/customer/v1/brands/{brand_id}/tracked-urls/{tracked_url_id} tracked_url_id A tracked URL of the brand. brand_id Idempotency-Key Any unique string, such as a UUID. Send the same key again to get the first answer and not a second write. The API keeps an answer for 24 hours."}]