Does AI recommend you by name?
Enter the questions your buyers actually ask. We put each one to the real assistants, with live web search, and show per platform whether your brand is cited, where, and who shows up instead.
AI visibility is how often assistants name your brand when someone asks a question you should be the answer to. It is measured as a rate across a fixed set of questions on a fixed set of engines, rerun on a cadence, because generative answers vary between runs and a single screenshot is one sample rather than a result.
Four engines, and they disagree
The same question put to four assistants routinely returns four different sets of sources. Which one you are absent from usually tells you why.
ChatGPT
Searches through Bing, so a site that is thin in Bing can be invisible here while ranking fine on Google. The most common surprise in a first report.
Claude
Fetches live and cites its sources explicitly, which makes it the clearest read on whether your own pages are reachable and quotable.
Gemini
Draws on Google’s index and AI Overviews, so it usually tracks your classic rankings more closely than the others do.
Perplexity
Cites heavily and over indexes on research intent, so it surfaces third party sources naming you that the others do not show.
What a report gives you
Four things, and the last one is the half that explains the other three.
Whether you are named
Per question and per engine, not as a single site wide score. A brand is usually visible for some questions and absent for others, and the average hides exactly the gap worth fixing.
Who is named instead
The competitor the assistant recommends in your place, question by question. This is the finding that gets forwarded internally, because it turns an abstraction into a name somebody recognises.
Which sources the answer used
The pages each answer was drawn from. Usually third party: directories, reviews, forums and press rather than your own site, which is what tells you where the work actually has to happen.
What is blocking you technically
The 36 check audit against your live pages: crawler access per bot, whether the page renders without JavaScript, schema, and whether anything you publish can be dated and attributed.
How this is measured, stated plainly
The unit is a prompt set, not a screenshot. Generative answers are probabilistic: the same question asked three times can return three different answers citing three different sources. So the figure is the proportion of your questions on which you were named, across the engines named above, on a stated date. Rerun the same set next month and the two numbers are comparable.
There is no invented index here. Several agencies in this category publish a proprietary "AI visibility score" with an unpublished methodology, and at least one claims a sample size with no data behind it. We are not going to do that. If a number on your report cannot be reproduced by asking the same questions yourself, it should not be on the report.
The technical half is deterministic and separate. Crawler access per bot, whether your page renders without JavaScript, whether schema parses, whether a page can be dated and attributed: these do not vary between runs, and they are where the reasons live. Run the free site audit and you get that half without an account.
What we will not promise. A position, because there is no position to hold in a generated answer. An inclusion, because no vendor has a lever that forces one. What is promised is that the reasons you are currently absent get found, and that the rate is measured the same way every month.
Find the blockers yourself, free
In most audits the reason a brand is absent is mechanical rather than editorial, and each of these finds one of the common causes in about a minute on any URL.
Reading a result rather than running one? The AI visibility checker page sets out the four outcomes a check can return and what each one means, and AI search optimization covers the five surfaces separately, since what gets you into Google AI Overviews is not what gets you into Perplexity. If the absence is on Google specifically, start with how AI Overviews work.
Prefer to interrogate it by hand? The prompt library has the same work written for ChatGPT and Claude: getting named in ChatGPT answers, ranking your business in ChatGPT and what ChatGPT search actually ranks.
Measuring it is the cheap half
Knowing you are absent is a week of relief followed by a quarter of nothing. The work is what changes the number.
What people ask about AI visibility
What is AI visibility?
How often, and in what terms, AI assistants name your brand when someone asks a question you should be the answer to. It is measured as a rate across a fixed set of questions on a fixed set of engines, not as a position, because generative answers have no ranking to hold.
How do you measure AI visibility?
We ask each engine the questions your buyers actually ask, record whether you are named, who is named instead and which sources the answer used, and rerun the same set on the same cadence. Alongside that we run 36 deterministic checks on your live pages. The prompt set gives you a rate; the audit gives you the reasons.
Why does the answer change every time I ask?
Because generative answers are probabilistic. Ask the same question three times and you can get three answers citing three different sources, which is normal and is the single most important thing to understand about this category. It is why a screenshot proves nothing, ours included, and why the unit here is a rate across a prompt set rather than a single observation.
Is this an AI visibility score you invented?
No, and the distinction matters because the category is full of invented metrics. There is no proprietary index here. The figure is the proportion of your own questions on which you were named, on engines you can name, on a date we publish. Anyone can reproduce it by asking the same questions, which is the only property that makes a measurement worth anything.
What does an AI visibility tool cost?
Dedicated AI visibility trackers start around $25 to $99 a month for prompt tracking on its own. Ours is $24 a month and includes the full technical platform beside it: 36 checks a page, site crawling, keyword tracking and exports. If you want the work done rather than measured, service plans start at $99 a month.
Can I check my AI visibility for free?
Yes. The checker on this page runs live against the real assistants, and the site audit that finds the technical reasons behind a poor result is free on any URL. There is no trial clock on either.
How is this different from tracking Google rankings?
A Google ranking is a position you hold on a page of results. An AI answer names a handful of sources or none, there is no second page, and being absent is the default rather than the failure case. Ranking well on Google helps and does not guarantee it, mostly because assistants lean far harder on third party corroboration than Google does.
How long does it take to improve?
The technical blockers, which are the most common cause, clear in days to weeks: once crawler access and rendering are fixed the next crawl sees a different site. Being named more often compounds over three to six months, because it depends on other people publishing things about you. Anyone promising citations in week two is describing a purchase.