ChatGPT visibility

Does ChatGPT recommend you, or the company you are trying to beat?

ChatGPT is where a large share of buying research now starts. Being absent from its answer for a purchase question is a commercial problem, and it is measurable.

Start Founding Agency Pilot — $499 Brand Pilot — $149

Founding pilot pricing while we validate. 14-day pilot, founder-run, cancel any time.

What one ChatGPT run records

Live capability

Surface

OpenAI API with the web search tool enabled. Logged in our data as openai_api_web_search.

Captured

The exact answer text, every source cited, whether you were named, whether a page of yours was cited, whether you were recommended, and which other brands appeared.

On failure

Marked unmeasured and excluded from the denominator — never scored as absence.

This is the surface we measure and the surface every report names. It is not a logged-in consumer session.

What we measure, precisely

We send your buyer questions to the OpenAI API with the web search tool enabled and record, for every run: the exact answer, the sources it cited, whether your brand was named, whether a page of yours was cited, whether you were recommended or merely mentioned, and which other brands appeared.

In our data this surface is recorded as openai_api_web_search. Every report says so. We use "ChatGPT" on this page because that is what people call it and what they search for — and then we tell you exactly what was measured. Both things can be true.

What this is not

It is not a logged-in ChatGPT session. It has no memory of a user, no custom instructions, no personalisation, and no access to app-only features. If someone offers to measure the consumer app at scale, ask precisely how — the honest answer is usually "through the API", the same as ours.

API with web search≠Your customer’s session

Reproducible and comparable across time, which a personalised session is not. That is the trade we made, and it is printed on every report rather than buried in a footnote.

Why one manual check is not enough

Anyone can open ChatGPT and type their category. Two things make that unreliable as a business input. First, a single run is a single sample, and answers vary — in our reliability testing, whether a brand was recommended held steady, but the exact set of competitors named did not. Second, without a frozen question set and a stored baseline, you cannot tell next quarter whether anything changed.

The same question, run twice

Illustrative example

“What is the best [category] tool for a small team?”

ChatGPT Competitor A, B, D named Run 1

ChatGPT Competitor A, C, E named Run 2

Recommendation status held; the competitor set moved. One screenshot cannot tell you which of those you are looking at.

What we are still proving. That buyer-intent prioritisation beats generic visibility tracking is our working hypothesis, not a settled finding. We think measuring the questions that precede a purchase is more useful than counting mentions across a large prompt list — but we are testing that with paying customers, not asserting it. We have no customer results yet, and we will not invent any.

Common questions

Can I check my ChatGPT visibility myself?

You can ask ChatGPT a question and see whether you appear. What you cannot do by hand is run the same question set repeatedly across engines, keep the raw answers, exclude failed calls from the denominators, and compare against a frozen baseline later. That difference is the whole product.

Does ChatGPT have a public index I can check my coverage against?

No. There is no reliable public full-index coverage figure to measure against, which is why we report citation and recommendation rates on a fixed benchmark of questions instead of an "AI indexed pages" number.

Is API measurement the same as what my customer sees?

No, and we will not claim it is. We measure the official API with web search enabled — a measurement surface, not a logged-in consumer session, so an individual customer can see a different answer. It is the closest reproducible measurement available to anyone, and we label it as exactly that on every report.

What we actually measure

Every result comes from the official OpenAI, Google Gemini and Perplexity APIs with web search or grounding enabled. In our data those surfaces are recorded as openai_api_web_search, gemini_api_google_search and perplexity_api_sonar, and reports name them explicitly.

That is not the same thing as a logged-in person using the ChatGPT app. These are official interfaces with web search on, but they are measurement surfaces rather than consumer sessions: no personalisation, no memory, no app-only features. Results can differ from what any individual user sees, and we do not claim the two are identical. We would rather put that on the pricing page than in the small print.

AI visibility · Monitoring over time · AI brand monitoring · AI search visibility · Request a free check

Start Founding Agency Pilot — $499 Brand Pilot — $149

Founding pilot pricing while we validate. 14-day pilot, founder-run, cancel any time.