Why a site can rank in Google and still be invisible to ChatGPT
Googlebot has been crawling the web for twenty-five years and every hosting provider knows to let it through. The crawlers behind AI answers are new, and a good deal of the internet’s security tooling now blocks them by default on domains set up in the last year or so. That block happens at the network edge, before your robots.txt is ever read, which is why a perfectly welcoming robots.txt proves nothing on its own.
The failure is total and silent. You do not slip down a ranking, you are simply not in the data the assistant is summarising, and no dashboard anywhere reports it. The only way to find out is to ask for your own page as one of those crawlers and look at what comes back, which is what the third of the checks above does.
The three questions the checks are grouped under
- Search results (SEO)
- The classic one: does your page have a title and a description worth showing in a list of results, and does the page work without JavaScript running first.
- AI answers (AEO)
- Whether the systems that answer the question themselves can lift a clean answer out of your page: your structured data, your visible contact details, whether your questions are marked up as questions, and whether you have told them not to use snippets.
- AI assistants (GEO)
- Whether an assistant can get in at all: robots.txt, the edge block, and whether you have published an llms.txt telling one what matters on your site.
What this cannot tell you
- Whether an assistant actually recommends you. That depends on what the rest of the internet says about your business, not on your own page, and anybody selling you a number for it is guessing.
- Anything about pages other than the one address you gave us. We read that page, your robots.txt and your llms.txt, and nothing else.
- Whether your Google Business Profile is in order. That needs a paid lookup, so the signed-in version of this check does it and this free one leaves the question out rather than answering it badly.