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.

​
The three core KPIs of AI visibility

​
KPI 1: AI Recommendations %

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.

​
KPI 2: AI Visibility %

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.

​
KPI 3: Share of Voice %

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.

​
How Genezio measures the KPIs

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.

​
What does not increase your score

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: 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.

​
Comparer and Introspector conversations

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.

​
How to read your scores

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:

​
Example: the KPIs of a CRM brand

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.

​
Why Genezio measures the KPIs separately

A brand can have high visibility and low recommendations, or the opposite. When you know your pattern, you know what to repair:

​
How to report the KPIs to leadership

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.

​
Next steps

Read How the 5 agents work.

​
In the API

The public API gives these numbers in the visibility report and the recommendation report.