Share of Answer

Share of Answer is the percentage of a market’s priority buying questions in which AI answer engines cite your brand. It tells you how often a buyer who asks AI about what you sell hears your name. It is our measure of AI search visibility, the result of generative engine optimization (GEO) and answer engine optimization (AEO).

Why the number matters

A buyer who asks AI for a shortlist reads the answer, not a page of links. The brands named there are the ones considered, and a first-page ranking on Google doesn’t guarantee you are one of them.

In November 2026 we put 20 buyer questions to four AI engines with web search on. 73 of the 80 answers cited their sources, with an average of 2.9 to 7.8 domains each, depending on the engine. Share of Answer counts how often you are among them.

How it is calculated

Share of Answer =

prompts where your brand is cited

priority prompts monitored

× 100

  • Priority prompts monitored are the buying questions that matter to your sales, written the way buyers ask them. A portfolio usually holds 100 to 300.
  • Prompts where your brand is cited are the ones in which an AI engine names your brand or links to your site as a source.

A brand named in 23 of 100 priority buyer questions has a Share of Answer of 23%.

A question counts for an engine when your brand is named or linked in at least two of its three runs. We report Share of Answer for each engine and as the average across engines, with mentions and links shown separately.

How we measure it

  1. Build the portfolio. We write 100 to 300 buyer questions with you, from your funnel, your sales team and the searches that bring people to your site.
  2. Ask the engines. Each question goes to ChatGPT, Gemini, Perplexity and Claude with web search on, three times, because answers vary from one run to the next. We also check whether Google’s AI Overviews cite you on the matching searches.
  3. Record who is named. For every answer we keep the brands mentioned and the sources linked, yours and your competitors’.
  4. Repeat. A Baseline includes a second measurement at 90 days. The AI Search Visibility program measures every month. Same questions, same engines, so a change in the number reflects a change in what AI says.

What the number doesn’t tell you

  • Answers vary. Two runs of the same question can name different brands. Three runs per question reduce the noise without removing it.
  • It is close to what buyers see, not identical. We query the models with web search on, which approximates the consumer apps. A buyer’s history and location can still change what they are shown.
  • Not every engine is covered. Microsoft Copilot is left out, because no data source we trust reports on it yet.

The free comparison is a different measure

The comparison on our Start page counts how often ChatGPT mentions a domain and how often Google AI Overviews cite it, in an index of answers already collected in the United States, in English. It shows who the engines draw on in your market, in a minute and at no cost.

It is not your Share of Answer. That is measured on your own buyer questions: four engines are asked each question directly, and Google’s AI Overviews are checked on the matching searches.

Share of Answer, share of voice and rankings

MetricWhat it measuresWhere
RankingsYour position in a list of linksSearch results
Share of voiceYour share of visibility across a market’s searches or advertisingSearch results and ads
Share of AnswerHow often AI names or cites you on your buyers’ questionsAI answer engines

Questions

How is it different from AI visibility?

AI visibility usually describes whether a brand shows up in AI answers at all. Share of Answer puts a number on it: the share of a defined set of buyer questions in which you appear, measured the same way every month, so it can be tracked over time and compared with competitors.

What is a good Share of Answer?

It depends on the market. The useful comparison is with your competitors on the same questions. If the brands buyers shortlist appear in half of the answers and you appear in a tenth, that gap is the work to be done.

How often should it be measured?

Monthly. Engines update and new sources appear, so the answers keep changing. A monthly reading on the same questions shows the trend without chasing day-to-day noise.

Can we measure it ourselves?

Yes. The method on this page is the one we use. What takes time is building a portfolio that reflects real buying questions and reading thousands of answers consistently. A Baseline does that work in four weeks.

Find out your Share of Answer.

A Baseline measures it on your own buyer questions in four weeks, and shows who AI names in your place.