Skip to content
ceeme

What does an AI crawler see of my page?

This inspector fetches your page the way a crawler does — one request, without running scripts — and shows what is left: the title, the heading structure, the readable text, the structured data and seven other signals. Anyone getting back an empty page while their browser shows text has the explanation at a glance for why an assistant knows them and can say nothing about them.

By Njusja Orban, founder of ceeme3 min read

Measure your own company → Rather talk it through first

In short

What you get back
The skeleton of your page: headings, text, structure. Exactly what a machine is left with and nothing more.
The ten signals
Each with a verdict in words. There is a half score too, because reality usually sits between good and weak.
The most expensive outcome
Zero words of readable text. Your page then does not exist for part of the crawlers, with no error reported anywhere.

What does it mean when zero words come back?

That your page only assembles its content after scripts have run. For a visitor that takes a fraction of a second; for a program fetching only the first response there is nothing. This is the most common explanation for an assistant that knows your address and can give no detail about you.

The repair is usually not a rebuild: the server can prepare the page once and send the text along in the first response. With most modern builders that is a setting.

Why does the heading skeleton matter so much?

Because a model infers from headings where a passage starts and ends. Without headings there is no passage to cut and the smallest citable unit is your whole page — which is nearly always too long to serve as an answer.

What can this inspector not see?

What happens after the page has loaded. A crawler that does wait for scripts sees more than this inspector — this is the floor, not the average. That is deliberate: if it reads well here, it reads well everywhere.

What this tool doesWhat it does not do
Fetches your page in one request, without running scriptsIs therefore stricter than some crawlers, which do wait — it is the floor, not the average
Judges ten signals with a verdict in wordsGives no figure comparable with another site — this is about this page
Works on any public addressSays nothing about whether you are named — readability is a condition, not an outcome

Frequently asked questions

My page scores low but ranks fine in Google. How?

The major search engines run scripts before judging a page; several fetchers feeding AI answers do not, or only sometimes. So you can be well indexed and still have nothing to offer in an AI answer — and that is exactly the gap this inspector makes visible.

Can I also inspect someone else's page?

Yes, because nothing is done that an ordinary visitor may not do: request one public address. What does not happen is that we store that page or build a file on it — the result is on your screen and nowhere else.

Read on

Free, no account and no card.