See your page the way an AI model reads it
A page a crawler can fetch but cannot parse is nearly as invisible as a blocked one, and it is harder to catch: from your browser it looks perfect, because your browser ran the JavaScript and the crawler did not. This fetches your page once, runs nothing, and shows you the exact words a model reads first.
No account, no card. We fetch the page once and never run its JavaScript, which is what most crawlers do.
Two ways to be invisible
The first is a blocked crawler, which the crawler check finds. The second is subtler and more common: the crawler arrives, reads the HTML your server sent, and finds a page shell with no words in it. Everything a visitor sees was assembled afterwards by JavaScript that the crawler never ran.
You cannot see this by looking. Inspect element shows the page after JavaScript; view source shows what was actually sent. The gap between those two is what this tool measures.
Why the opening matters more than the rest
Models quote passages, not pages. When one decides whether your page answers a question, it leans hard on the first stretch of text. That is why the tool shows you your own opening ninety words rather than a score: almost nobody has read their page that way, and it is rarely what they expected. Very often the first thing a model reads is a cookie notice, a navigation list, or a slogan.
The fix is not clever. Put a direct, self-contained answer to one real question in the first hundred words, in plain HTML text.
How to check it yourself
No tool needed. Fetch the page and read what comes back:
curl -s https://yoursite.com/your-page | wc -c
curl -s https://yoursite.com/your-page | grep -o "<h1[^>]*>.*</h1>"Or, in any browser, use view-source rather than the inspector. If the words you expect are not in there, neither engines nor models have them.
What to do about each result
Almost no words in the HTML
Server-render the main content or pre-render it at build time. Every modern framework can, and most now default to it. You do not have to change your stack, only make sure the words exist in the first response.
Words are there but thin
Being indexed and being quotable are different bars. A page with two hundred words of marketing prose clears the first and fails the second. Answer one question properly instead of gesturing at five.
No headings, or headings that say nothing
Headings phrased as the question a reader would type are the ones an engine matches to a question. “Pricing” matches nothing. “How much does it cost for a small team?” matches something.
No structured data
Without JSON-LD an engine has to infer who you are and what you sell from prose, and inference goes wrong. Organization and FAQPage are the two that pay, and both are about twenty minutes.
Questions
Why does JavaScript matter here?
Most crawlers read the HTML your server sends and do not run your JavaScript. If your content only appears after a framework boots, the page looks complete in your browser and empty to an engine. This tool fetches once, runs nothing, and reports what is left.
My page looks fine to me. Why does this say it is empty?
Because your browser ran the JavaScript and a crawler did not. That gap is the whole point of the check. View source rather than inspect element and you will see the same thing we do.
How do I fix a JavaScript-rendered page?
Server-render the main content, or pre-render it at build time. Every modern framework can do this and most default to it now. You do not need to abandon your stack, only to make sure the words exist in the first response.
Why show me the first ninety words?
Because models lift passages, not pages, and they weight the opening heavily. Seeing your actual opening is usually a surprise: it is a cookie banner, a menu, or a strapline rather than the answer to the question you want to be found for.
Does structured data help with AI answers?
It removes guesswork. Without JSON-LD an engine infers who you are and what you sell from prose, and inference goes wrong. Organization and FAQPage take about twenty minutes and are the two that matter most.