What is AI visibility?
AI visibility is how often AI assistants such as ChatGPT, Perplexity, Claude and Google AI Overviews name your brand when people ask about your category. It is measured per question and per engine, by asking the questions buyers ask and counting the answers that name you. Unlike a search ranking there is no fixed position to check once: the same question can name you in one answer and leave you out of the next.
What metrics best measure brand visibility in AI search results?
Mention rate is the core metric: the share of AI answers to your buyer questions that name your brand, reported per engine and overall. Everything else qualifies it.
Because one answer is a sample and not a fact, a mention rate needs an interval beside it. A Wilson interval shows the range the true rate could sit in given how many answers were read. The other measures tell you why the rate is what it is:
- Who is named instead: the other brands in the same answers, ranked by how often each appears.
- Share of voice: your share of all brand mentions in those answers. See AI share of voice.
- Position: where in the answer you appear when you are named. First in a list of five reads differently from fifth.
- Citations: the pages each engine lists as sources, and how many of them are yours. See citation share.
- Coverage: which questions and topics name you at all. See prompt coverage.
How is AI visibility calculated?
AI visibility is calculated as answers that name the brand divided by answers sampled, for a fixed set of questions on a fixed set of engines. Named in 6 of 20 answers is a mention rate of 30%, with a 95% Wilson interval of about 14.5% to 51.9%.
Three choices decide whether the number means anything. The questions must be ones buyers ask before they know your name, such as "what is the best X for Y", because a question that names you will mention you whether or not the engine knows who you are. The engines must be asked the same questions in the same words, so they can be compared. And the sample must be repeated, because answers change from one run to the next and a single check can mislead in either direction.
What does AI visibility look like in a real example?
In the public Tesla example (4 Oct 2026), 53 of 100 answers to 25 buyer questions named Tesla: a mention rate of 53%, with a 95% range of about 43% to 62%.
The engines disagreed. ChatGPT and Google AI Overviews each named Tesla in 16 of 25 answers, Claude in 11 and Perplexity in 10. The intervals overlap, so on 25 answers each that gap is a lead to watch rather than proof that one engine favours Tesla. The brand questions behaved as expected: 19 of 20 answers to questions that named Tesla mentioned it, which says almost nothing about visibility and a lot about why brand questions are kept separate.
| Engine | Named Tesla | Mention rate | 95% range |
|---|---|---|---|
| ChatGPT | 16 of 25 | 64% | 45% to 80% |
| Google AI Overviews | 16 of 25 | 64% | 45% to 80% |
| Claude | 11 of 25 | 44% | 27% to 63% |
| Perplexity | 10 of 25 | 40% | 23% to 59% |
| All four | 53 of 100 | 53% | 43% to 62% |
How do these tools measure recommendations versus simple brand mentions?
A mention is any appearance of your brand's name in an answer; a recommendation is a mention the answer endorses, usually by putting you on its shortlist or naming you as the pick for the buyer's situation.
Mention rate counts both, which is why it is read beside position and tone. A brand named last, or named as the option to avoid, is mentioned but not recommended. Reading the answer text, not just counting names, is the only way to tell the two apart, and it is why every answer should be kept in full so a figure can be traced back to the words behind it.
Where do you see AI visibility in Proofsource?
Proofsource asks your buyer questions on ChatGPT, Perplexity, Claude and Google AI Overviews and shows your mention rate per engine with its 95% interval, who is named instead, your share of voice and the pages each engine cites. The AI Visibility feature page shows the screens; the methodology explains how every figure is counted.
Common questions about AI visibility
Is AI visibility the same as SEO ranking?
No. A search ranking is a position on a results page for a keyword. AI visibility is whether an AI answer names you at all, and it varies between engines and between runs of the same question, so it has to be sampled rather than checked once. Good SEO helps, because engines that search the web read pages that rank, but it does not guarantee a mention.
How do these tools measure brand visibility and mention frequency?
They ask a set of questions on each AI engine, read every answer, and count the share that name your brand. Better tools repeat the sample, keep the full answer text, separate questions that name you from ones that do not, and show an interval so a small sample is not mistaken for a trend.
How can I check my brand's AI shortlist ranking?
Ask each engine the buyer questions that matter to you and record where your brand appears in each answer's list, if at all. Position only means something across many answers, so track it beside mention rate rather than from a single answer.
How many prompts and responses do I need for a reliable score?
Enough that the interval is narrower than the change you want to detect. Named in 6 of 20 answers is 30% with a range of about 15% to 52%; named in 60 of 200 is also 30%, but the range narrows to about 24% to 37%. Start with a focused set of buyer questions and let repeated runs build the sample.
Related terms
- AI share of voice
AI share of voice is your brand's share of all the brand mentions in AI answers to a set of questions.
- Citation share
Citation share is the share of all the citations in AI answers that point at a given website.
- Prompt coverage
Prompt coverage is how much of your set of tracked buyer questions AI engines answer with your brand named, broken down by question, topic and engine.
- Generative engine optimization (GEO)
Generative engine optimization (GEO) is the work of improving how generative AI engines describe and recommend your brand, so you are named and cited in their answers.
Worked example: Tesla, public AI answers, 4 Oct 2026: Proofsource's first scan of Tesla, 25 buyer questions and 5 brand questions, one answer per question on each of four engines. A small sample from one brand, shown as an example, not a benchmark.
See where AI leaves you out, and what to fix.
See which buyer questions name your brand, who gets named instead, and what to fix. 200 answers. 25 Prompts/Question. Top AI engines. 2 days. Free. No card.
Google AI OverviewsGoogle AI Mode
