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.

​
What a topic represents

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

​
Topics, scenarios, and conversations

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.

​
Genezio Agent types

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.

​
Prompter Agent

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.

​
Recommender Agent

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.

​
Introspector Agent

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.

​
Comparer Agent

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.

​
Fact Checker Agent

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.

​
Single-turn and multi-turn agent types

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.

​
How Genezio creates topics

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.

​
Generate more topics

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.

​
Select answer engines for each topic

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.

​
Why topics are important

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.

​
Next steps

To know how Genezio changes topics into structured interactions with answer engines, refer to:

​
Topics in the public API

To read and change topics, refer to List topics.