AI Visibility Grader
When someone asks an AI assistant, does your business come up?
We ask Claude, with web search enabled, the actual question a customer would type — three separate times — and show you exactly what it said: who it named, whether your site was cited, and where your page falls short. It's not a ranking, and there's no way to give you one — but we print the exact question, so you can run it yourself and check us.
What We Check
Five things that decide whether an AI assistant can name you at all.
Every check is a real, measurable signal in your page's HTML and robots.txt — not a guess, and never a ranking prediction. These are the same fundamentals search engines have always rewarded; an AI assistant just reads them differently.
Business Clarity
Whether the page plainly states what you do, where you do it, and how to reach you — the things an AI has to repeat back, not infer.
Content Extractability
Whether there's enough real text an AI can quote, or whether it's locked inside images, scripts, or a JavaScript-rendered shell.
Question Coverage
Whether your page answers the questions customers actually ask — the shape of content AI tools tend to quote directly.
Structured Data
Whether Organization and FAQPage schema make your business and your answers machine-readable. Not required, but it removes the guesswork.
Crawlability
Whether anything is accidentally blocking access — a stale noindex tag or an over-broad robots.txt rule undoes everything else on this list.
See What You'll Get
A real answer, run three times — not a made-up score.
Here's an example of what your report looks like: an overall score across five fundamentals, plus what the AI assistant actually said when we asked about this business.
Sample Report
yourbusiness.com
A phone number and email are linked, but the meta description is a slogan, not a description of what the business does.
Plenty of real text for an AI to quote, and it's present in the served HTML, not rendered in by JavaScript.
No headings phrased as questions, and no FAQ section — nothing shaped the way an AI tends to quote answers.
No structured data found — nothing states, in machine-readable form, who this business is.
Nothing blocks the page and robots.txt is reachable, but no canonical URL is set — an easy fix worth making anyway.
Questions
What business owners usually ask before running this.
Claude, with web search switched on, three separate times. We name it because the report tells you to go and check us — and if we let you assume we had asked ChatGPT, a perfectly normal difference between two models would look like our tool being wrong. ChatGPT, Gemini and Perplexity are separate systems with separate indexes and will answer differently. If being named across all of them matters to you, say so and we will look at them properly rather than guessing from one.
No. There's no such thing as a ranking in AI search the way there is in Google — an AI assistant generates a fresh answer each time it's asked, it isn't pulling from a fixed ordered list. This tool asks Claude a real question and shows you exactly what it said. It's a snapshot, not a rank, and nobody — including us — can hand you one.
Because that's how these tools work — the same question can get a different answer on Monday than it did on Friday. We run the question three times and show you all three runs so you can see that variation for yourself, rather than pretend one run is the whole truth. Three runs is still a small sample; treat the pattern, not any single answer, as the signal.
The report leads with the exact question we asked, so you can run it yourself. Ask Claude and you should see something close to what we reported. Ask ChatGPT, Gemini or Perplexity and you may well see something different — they're separate systems with separate indexes, and that difference is worth knowing in itself.
Yes. Your instant preview and your full emailed report are both completely free — no credit card, no trial, nothing to cancel.
That happens, and it's genuinely useful information — it usually means an AI assistant can't find enough on your page to state plainly what you do or answer the question customers are asking. The five category scores below tell you specifically what to fix.
No, and we'll say so directly in your report rather than just congratulating you. Three runs is a small sample, and appearing today doesn't guarantee appearing next week. We'll still show you your weakest categories, because there's almost always something worth tightening up.
No — those three are rules-based checks of your site's performance, SEO fundamentals, and on-page keyword targeting. This is the only one of the four that calls a live AI model: it asks Claude a real question and reports back a real, if imperfect, answer.
We use them to send your report and, if it looks like we could genuinely help, to follow up with practical next steps. We don't sell or share your information.
Ready to Close the Gap?
Showing up today isn't the same as being built to show up.
Whatever your report turns up, we can help you fix it — from the content an AI can actually quote to the structured data that removes the guesswork.
Want to check your overall site health first? Try the Website Grader → Or see how well a page targets specific keywords with the Keyword Optimization Grader →