The short version: there is no dashboard that tells you whether AI systems can see your website, which is why most business owners have no idea. The only way to find out is to look for the symptoms — and unlike most website problems, these ones don't show up in your analytics, your rankings, or anything your web developer monitors. This is a diagnostic guide: what the signs are, what each one indicates, and how to read what you find. Not a repair manual — a way to know whether you have the problem.
Start with the question your buyers ask
The most direct diagnostic is also the simplest: think of the question a good prospect would ask an AI assistant when looking for what you sell — not your brand name, but the need. "Who's the best [what you do] for [who you serve]?" Then ask it, in a fresh session where the AI doesn't have your history influencing the answer, and read what comes back carefully.
What you're looking for isn't just whether your name appears. Read for three things: are you named at all; who is named instead; and, if the engine shows its sources, what it cited to justify the recommendation. That third one is the most informative and the most overlooked. The sources an engine leans on tell you where the authority in your category actually lives — and it's frequently not where businesses assume. Do this for a handful of your real buyer questions and a pattern emerges quickly.
The symptoms, and what each one means
You're not named, and competitors are. The baseline symptom. It means the engine had enough confidence to recommend specific businesses and you weren't among them — so this isn't a case of the AI refusing to name anyone. Something about your competitors made them safer to name than you, and it's usually corroboration rather than quality of work.
Nobody is named — the AI answers generically or points to a directory. Different signal, and a more hopeful one. It means the engine doesn't feel confident enough about anyone in your category to make a specific recommendation, and is deferring to a list instead. That's an opening: categories where no one has established themselves are the easiest to enter, and the directory being cited tells you exactly where to be.
The AI describes your business incorrectly. If you ask about your company by name and the answer is wrong, outdated, or confuses you with another business, you have an entity problem — the machine can't cleanly resolve who you are. This is worth taking seriously, because an engine uncertain about your identity will not confidently recommend you to someone else, and misinformation about you may be actively circulating.
Zero referral traffic from AI surfaces. If your analytics show essentially no visits arriving from AI assistants while your category is one people research, that's corroborating evidence for the other symptoms rather than a standalone diagnosis. It's the downstream shadow of not being cited.
Your pages are missing from the indexes AI reads. This one requires looking rather than inferring, and it's the most mechanically fixable of the set. Different AI systems retrieve from different underlying indexes, and a site that's healthy in Google can still be absent from another index that a major assistant relies on. When this is the cause, everything looks fine from the outside — which is why it goes undetected for so long.
How to interpret what you find
A single query is a snapshot and can mislead — engines are probabilistic, and one answer isn't a verdict. What matters is the pattern across several of your real buyer questions, ideally checked more than once. If you're consistently absent across most of them, that's a genuine visibility problem rather than bad luck on one phrasing. If you appear for some questions and not others, that's diagnostic in a different way: it usually means you're legible to the machine for the topics you've addressed clearly and invisible for the ones you haven't.
Also resist the temptation to read a bad result as a verdict on your business. Being absent from AI answers correlates poorly with being good at what you do. Some of the best-run companies in a category are invisible while a mediocre competitor gets named because they happen to appear in a listicle the engine trusts. That's frustrating, and it's also the encouraging part: the gap is about machine legibility and corroboration, not about deserving it, which means it's addressable by work rather than by being a fundamentally better business.
What the diagnosis doesn't tell you
Knowing you're invisible is genuinely useful — it converts a vague worry into a specific, confirmed problem. What it doesn't tell you is which of the underlying causes is responsible in your case, and that matters enormously, because the causes require completely different work. An entity-resolution problem, a lack of citable content, an absence of third-party corroboration, and a missing index presence look identical from the outside — you're just not there — but they're different projects with different timelines.
Sorting that out is what a proper visibility audit does: it goes past the symptom to establish which layers are actually failing, in what order they'd need to be addressed, and what the realistic timeline looks like for each. That distinction between symptom and cause is the whole reason a diagnostic exists. You can absolutely establish that you have the problem yourself in an afternoon — and if the answer is that you're being consistently omitted from the questions that matter to your business, the honest next step is finding out exactly why before spending effort on the wrong layer.
FAQ
How can I check if AI knows about my business?
Ask the AI assistants the questions your buyers would ask — the need, not your brand name — in a fresh session, and read whether you're named, who's named instead, and which sources the engine cited. Repeat across several of your real buyer questions. That pattern, not any single answer, is your diagnosis. No dashboard reports this, so looking directly is the only reliable method.
What does it mean if the AI names my competitor but not me?
It means the engine was confident enough to recommend specific businesses and yours wasn't among the ones it felt safe naming. In our experience this usually reflects corroboration — third-party sources, reviews, and mentions that make a business look established — rather than the actual quality of the work. It's a legibility and evidence gap, not a verdict on your business.
The AI described my business wrong. Is that bad?
It's worth addressing promptly. It indicates the machine can't cleanly resolve your identity, and an engine that's uncertain about who you are won't confidently recommend you to someone else. It also means inaccurate information about your business may be circulating in answers to other people's questions, which is a reputational issue as much as a visibility one.
Is one bad result enough to conclude I have a problem?
No — these systems are probabilistic and a single query can mislead. Check several of your genuine buyer questions, more than once. Consistent absence across most of them is a real visibility problem; sporadic appearance usually indicates you're legible for some topics and not others, which is a different and more targeted issue.
If I find out I'm invisible, what's the next step?
Establish which underlying cause applies, because the four main ones — unclear entity, nothing quotable, no corroboration, missing from the index — look identical from outside but require entirely different work and have different timelines. Acting before you know which layer is failing usually means effort spent in the wrong place. That's what a proper audit is for.
Should I check on ChatGPT, Perplexity, or Google's AI Overviews?
All of them, because they behave differently and draw on different underlying sources — being named by one is no guarantee of the others. A business can be well-represented in one engine and entirely absent from another, and knowing which is which matters, since the causes differ. Checking a single engine gives you a partial and potentially misleading picture of where you stand.
How often should I re-check?
Periodically rather than obsessively — monthly is a sensible rhythm. These systems change frequently enough that a snapshot ages, and a regular check turns a one-time observation into a trend you can actually read. Daily checking mostly produces noise, since answers vary somewhat between sessions even when nothing about your business has changed.
What if I'm named for some questions but not others?
That's a more informative result than blanket absence, and it's common. It usually means you're legible to the machine for the topics you've addressed clearly and specifically somewhere, and invisible for the ones you haven't — so the engine can place you confidently in one context but not another. The useful exercise is to look at which questions you win and what those have in common, then compare against the ones you lose. The pattern almost always points to where your business has been concrete in public versus where it has been vague, and that contrast is the most actionable thing a self-check can surface.
Want the diagnosis done properly? Our AI Visibility Audit runs your real buyer questions across every major engine and identifies which layers are failing — with the transcripts. Related: why websites go invisible and what AI actually reads.