Free tools · Website checker
Find out whether AI crawlers can reach your page, and how ready it is for ChatGPT, Claude, Gemini and Perplexity to quote. Free, with no signup.
5 free checks per hour · results cached for 24 hours.
Sample data
kestreloutdoor.com/rain-jackets/summit-shell · detected subject: “Summit Shell Rain Jacket” · 18/23 checks passed
Most pages AI engines cite score 80 or more · how the score works
Worked example · Kestrel Outdoor, a fictional brand
The free result names your top three findings, with code to paste where it can be written from your page itself. The full report adds every finding and its fix. Accounts open from the waitlist.
Top 3 findings (free)
2 more findings, in the full report
Paste-ready fixAn answer-first opening
Open with one short block: what the Summit Shell is, who the jacket is for, and the fact a buyer asks first.
NowWhen the weather turns on the ridge, you want a shell you can forget about. We spent two seasons testing fabrics to find the right one…
PasteThe Summit Shell is a packable waterproof rain jacket for hiking, 280 g in a men’s medium. It packs into its own chest pocket and is sized for layering over a fleece.
Paste-ready fixOrganization JSON-LD with sameAs
Link the brand’s official profiles from the page, and list them in Organization JSON-LD sameAs.
<script type="application/ld+json">{ "@type": "Organization", "name": "Kestrel Outdoor", "url": "https://kestreloutdoor.com", "sameAs": [ "https://www.linkedin.com/company/kestrel-outdoor", "https://www.instagram.com/kestreloutdoor" ]}</script>Paste-ready fixA meta description, from the page’s own opening
Rewrite it to 120–160 characters, answer-first.
<meta name="description" content="The Summit Shell is a packable waterproof rain jacket for hiking, 280 g in a men’s medium, that packs into its own chest pocket.">What an AI engine reads
The checker reads the HTML your server sends, the way AI crawlers fetch it. It shows the opening an engine finds first and the facts it can lift for a buyer, which depend on the page: specs for a product, plans and integrations for software, area, hours and prices for a local service. Each fact comes with the place on the page it was found, or why it couldn’t be read.
kestreloutdoor.com/rain-jackets/summit-shell
The opening it finds first
“Built for those who refuse to let the weather decide. Meet the shell that goes where you go.”
Marketing line, no product facts: nothing here answers what the Summit Shell is.
Facts it can lift 2 of 6
kestreloutdoor.com/rain-jackets/summit-shell
The opening it finds first
“The Summit Shell is a 280 g, fully waterproof rain jacket for backpacking and day hikes, priced at $159.”
Answer-first: the sentence an engine lifts when a buyer asks what the jacket is.
Facts it can lift 3 of 6
clauseminds.com/pricing
The opening it finds first
“Contracts, simplified. Join the teams who never miss a clause.”
A slogan: nothing tells an engine what ClauseMinds is or who it suits.
Facts it can lift 2 of 6
clauseminds.com/pricing
The opening it finds first
“ClauseMinds is contract review software for law firms of 2 to 20 lawyers, from $49 per seat a month, with Word and Salesforce integrations.”
Answer-first: what the product is, who it suits and what it costs, in the sentence an engine quotes.
Facts it can lift 4 of 6
hallaplumbing.fi
The opening it finds first
“Your trusted partner for all things water since 1998.”
No service, no place, no hours: an engine can’t match the page to “emergency plumber in Espoo”.
Facts it can lift 1 of 6
hallaplumbing.fi
The opening it finds first
“Halla Plumbing is a 24-hour emergency plumber covering Espoo, Helsinki and Vantaa, with call-outs from €89.”
Answer-first: the service, the area and the price in one sentence an engine can match to a local search.
Facts it can lift 4 of 6
Choose your first fix
Weigh each finding three ways: how much of the score its factor carries, who can make the change, and what the change gives an engine to quote. Fix the one that wins on all three, re-check, then move down the list.
kestreloutdoor.com/rain-jackets/summit-shell · 75 of 100 · the three free findings · worked example
Fix firstLead block is 164 words (aim 50–120)
Answer-first content · 30%
A marketer, in the page editor
A short opening to quote
NextMeta description sub-optimal length
Metadata · 15%
A marketer, in the page settings
A preview that ends on the point
With the next releaseNo entity-grounding links
Entity grounding · 15%
A developer, once for the whole site
Proof of which Kestrel Outdoor this is
Why the opening goes first. Answer-first content carries 30% of the score, the change is one paragraph a marketer can write without waiting for a release, and the new opening is the sentence an engine lifts when a buyer asks what the product is.
Then. Re-check the page once the new opening is live. The opening check passes, and the next finding joins the top three. The sameAs links go into the next developer release.
The checker fetches your page’s HTML once, the way many AI crawlers read a page, and scores the five factors that decide whether an engine can quote the page. Each factor lists the checks your page passed and failed.
30% of the score
The JSON-LD that tells an engine what this page and its subject are.
62 on this page · 4 of 5 checks passed
30% of the score
Whether the page opens by defining its subject rather than building up to it.
44 on this page · 6 of 7 checks passed
15% of the score
The names and links that let an engine confirm you are who you say you are.
40 on this page · 1 of 2 checks passed
15% of the score
Title, description and canonical, read as the page’s own summary of itself.
72 on this page · 3 of 5 checks passed
10% of the score
Whether the page can be fetched and indexed at all.
92 on this page · 4 of 4 checks passed
Fix the page in front of you first. Then work through every check, and find out whether AI recommends you.
Compare the page with the pages engines cite for the same question, then write what the page lacks in your own voice.
Explore content & site optimizationThe page checklist explains each check behind the score and how to pass each one, grouped by factor and weight.
Open the page checklistPages change, and the engines reading them change too. We’ll re-run this exact check in 30 days and email you only if something moved.
Check a page first
A ready page is the start. Track whether ChatGPT, Claude, Gemini and Perplexity name your brand for the questions your buyers ask, and who they name instead.
Explore AI visibility & competitorsSomething else on your mind? Contact us.
No. The score measures readiness: whether an engine can fetch, read and quote the page. Whether ChatGPT or Gemini recommends you also depends on the question, your competitors and the sources the engine trusts. The platform tracks that part.
Five factors: structured data 30%, answer-first content 30%, entity grounding 15%, metadata 15% and crawlability 10%. Within each factor, a critical check counts three times a minor one and an important check twice.
We ran the checker on 266 pages that ChatGPT and Perplexity cited in our own runs between July and September 2026, from 209 sites. Their median is 87, and three in four score 79 or more. Your result shows where your page sits among them. A higher score is not a promise of a citation: it means the page is as easy to read and quote as the pages engines already cite.
Yes. Beside the score, it fetches your site as each answer engine’s crawlers, from OpenAI, Anthropic, Perplexity and others, and reports which ones got through. robots.txt can allow a crawler that a firewall or CDN rule still refuses, and only a real fetch shows that.
This check reads the HTML the server sends, as many AI crawlers do. A page that renders its content with JavaScript can’t be read that way. The checker then holds back the score instead of showing a misleading one, and the rendering problem becomes the top finding.
Yes. The five factors apply to any public page, and the facts the checker looks for follow the page: specs and price for a product, plans, integrations and who it suits for software, services, area, hours and prices for a local business. The structured data check looks for Product, SoftwareApplication or LocalBusiness markup to match.
Yes. Results are cached for 24 hours, then the checker fetches the page fresh. Run the check again once your change is live, or ask us to re-check the page in 30 days.
It checks a public page’s readiness to be found, understood and cited. Ongoing tracking and work belong in the platform.
The score reads your page. The platform asks ChatGPT, Claude, Gemini and Perplexity the questions your buyers ask, and finds out whether your brand is mentioned, recommended and cited.