Someone in your market is asking ChatGPT, Claude or Perplexity for what you do, right now. They do not get a page of links any more. They get three names.
If yours is not one of them you never got the chance to compete — and nothing in your analytics will ever show it. This tells you where you stand, in sixty seconds.
Two fields. No signup, no sales sequence. Your score and exactly what is broken appear on this page the moment you submit. Or skip the form entirely and download the cheat sheet — no email needed.
Your customers have started asking AI instead of Google. When they ask for the best firm in your category, an assistant reads a handful of websites and names a few companies. If it cannot read yours, you are not in the running — and this is invisible in Google Analytics, which is why almost nobody knows it is happening to them. Your own results are further down this report. First, here is what was actually tested, and why it matters.
AI assistants do not use Google’s index. They run their own crawlers, and each one has a name. Your website decides, in a single text file, whether they are allowed in.
| Crawler | What it feeds | Why it matters |
|---|---|---|
| GPTBot | ChatGPT | Biggest |
| OAI-SearchBot | ChatGPT’s live search results | Citations |
| ClaudeBot | Claude | Growing fast |
| PerplexityBot | Perplexity, an answer engine that always cites | Sends traffic |
| Google-Extended | Gemini and Google AI Overviews | Above your rank |
| CCBot | Common Crawl, the open dataset nearly every model trains on | Long game |
In your browser, go to yoursite.com/robots.txt. Every website has one. Read what it says about the names above. Most owners have never opened this file, and a great many site builders and security plugins block AI crawlers by default.
User-agent: GPTBot
Disallow: /
User-agent: GPTBot
Allow: /
Almost nobody blocks AI crawlers on purpose. It is nearly always inherited, which is why it goes unnoticed for years.
Being allowed in is worth 25 points of 100 — the single biggest check, and the only one you can verify from your own browser. But a crawler that reads a page it cannot understand, quote or attribute will still never name you. The rest of the score decides that. Here is how it is put together.
That is the whole shortlist. Someone asks, AI answers, three companies get mentioned, and most people never click anything at all.
Of what AI says about a company comes from pages that company does not own. Your own website is the small part. That is why most people are invisible and cannot work out why.
Try it yourself. Open ChatGPT and ask "who are the best [what you do] in [your city]?" Whatever it says back is what your next customer sees.
Either AI cannot read your website. Or it can read it, and still has no reason to say your name. Those are different problems and they need different fixes, which is why you get two scores instead of one.
Sometimes your site turns AI away at the door, because your site builder or security plugin blocked it by default and nobody ever looked. Sometimes the page is simply blank to it: a person sees a full website, AI sees an empty page. Google hides this from you completely, because Google can read what AI cannot.
This is the one nobody checks. Your site is fine. AI just has no reason to repeat your name: nothing says who you are FOR, you describe yourself three different ways, you have not written the pages it likes to quote, and nobody else on the internet is talking about you. So it names a competitor it understands better.
One for whether AI can read your site at all. One for whether it has any reason to say your name. Nothing here is guessed and nothing is written by an AI — run it twice and you get the same answer, and you can check every finding yourself.
Each check below is worth a share of the 100. The bigger the number, the more it matters — getting shut out at the door costs you 25, a broken sitemap costs you 5. Your report shows what you scored on each one, like 18 / 25.
Our robots.txt shipped with a placeholder the site builder never filled in, so we pointed every crawler at a URL that does not exist. Our sitemap listed all 17 pages under a domain we no longer own, and four of them had slugs deleted months ago. And the site itself is a React app, so an AI crawler read exactly zero words of it. We scored 40 out of 100.
Sitemap: https://[YOUR_WEBSITE_URL]/sitemap.xmlSitemap: https://hiremina.com/sitemap.xmlWe fixed it before we offered to fix yours. If it can happen on a web designer's own site, it can happen on yours — and you would have no way of knowing.
It is the reasonable fear, and the honest answer is uncomfortable: blocking mostly costs you the credit, not the copying. Models trained on the open Common Crawl years before you had a say. What a block reliably prevents is ChatGPT naming you and linking to you when it answers. You keep the exposure and lose the attribution. If something is genuinely confidential, keep it off a public website entirely, not behind a robots.txt line that is a request rather than a lock.
No. These are different crawlers. Googlebot handles search and is untouched by any of this. Google-Extended only governs Gemini and AI Overviews, and Google states it is not used as a ranking signal. Fixing a broken sitemap helps ordinary search too. There is no ranking downside.
Two scores out of 100 — whether a machine can read you, and whether it names you — every check broken down, and the specific things that are wrong on your site with what each one is costing you. It is free, it takes about a minute, and nobody calls you.
What it does not hand over is the fix. We know exactly what each one takes and how long — that is what the call is for, and it is thirty minutes, not a sales sequence. You decide after you have read the report, and you are welcome to take the findings to your own developer instead.
No. An SEO audit tells you how you rank in a list of blue links. This tells you whether ChatGPT and Claude can read, understand and quote you when someone asks a question instead of searching. The two overlap a little and diverge a lot — a site can rank on page one of Google and be completely invisible to ChatGPT. Ours was.