Aliceby TranX
FREE · NO SIGN-UPGEO / AEO

Does AI name you when buyers ask?

One brand-name question and one buyer question, checked live on Google AI Overviews, ChatGPT and Claude, plus the technical conditions that decide whether an answer engine can read you at all. 30 seconds, no sign-up.

https://
Covers title, meta description, H1, canonical, Open Graph, structured data, robots.txt, sitemap, HTTPS, search-index presence, AI-crawler access, server-rendered content, /llms.txt, and tracking-tag detection. Takes about 30 seconds.
Free check is rate-limited per IP. Sign up to scan unlimited URLs and connect data sources for deeper insights.
SAMPLE RESULT

AI knows this business by name, but does not suggest it to new buyers yet.

When someone asks about you by name
Yes, all three named you
GoogleChatGPTClaude
Google, ChatGPT and Claude all answered and described the business correctly.
When a buyer asks for what you do
No, none named you
GoogleChatGPTClaude
They suggested competitor-one.com and another-vendor.com instead.

A sample of the free result. Run your site above to see your own, with the exact questions asked and who AI named instead.

The scan grades AI visibility alongside your classic SEO, crawlability and index presence, so you get the whole picture, not just the AI half.
THE BLIND SPOT

A perfect SEO score can hide a total AI blackout.

Googlebot and the AI crawlers are different bots obeying different lines in the same file. When a site blocks GPTBot, nothing breaks in a way you would notice: Google keeps crawling, rankings hold, Search Console stays green, GA4 shows nothing unusual. There is no alert, because from every dashboard you own, nothing happened.

The most common cause is not a decision — it is inheritance. A robots.txt written before 2023 that allowlists the crawlers of its day still parses perfectly today, and quietly excludes every AI crawler invented since. The second most common cause is a client-rendered front end: the page looks full in a browser, and arrives empty at anything that does not run JavaScript.

Result first, then the conditions: the scan asks Google AI Overviews, ChatGPT and Claude one brand-name question and one buyer question in your category, live, and shows per engine whether you were named or who was named instead — a snapshot, not a guarantee. Then it tests whether the engines can reach you at all — access, readability, extractability. Re-checking the buyer questions week over week is Alice’s job once you sign up. Recognition is not recommendation, and an engine that cannot read you will never name you.

THE FOUR CHECKS

What we test, and why each one decides a citation.

What happens when an answer engine meets your site
01
Fetch
The AI’s crawler knocks on your door and checks whether it is welcome.
Fails when
your site’s rules never let it in
This scan tests: Which AI crawlers you allow in, one by one.
02
Render
It reads what your server sends, not what a browser builds afterwards.
Fails when
your page is blank until JavaScript runs
This scan tests: Whether your words arrive without JavaScript.
03
Extract
It picks out facts it can credit to you by name.
Fails when
your facts are buried in paragraphs
This scan tests: Whether your key facts are labelled for machines.
04
Select
It chooses your page to name inside its answer.
Fails when
someone else said it better
Not something a scan can grade. Alice checks this for you daily.
Every stage is a gate, not a score.
Fail the first one and the rest never happen. Most sites that are invisible to AI fall at stage one or two, not at four.
The four stages between your HTML and a cited answer. The full walkthrough is in How AI reads your website in 2026.
01
AI crawler access
Stage one, Fetch: whether each engine is actually allowed in.
We resolve your robots.txt the way a bot does and report each AI crawler separately — GPTBot, OAI-SearchBot and ChatGPT-User (ChatGPT), ClaudeBot, Claude-SearchBot and Claude-User (Claude), PerplexityBot and Perplexity-User (Perplexity), Google-Extended (Gemini) and CCBot — as allowed or blocked. robots.txt groups are exclusive, so a file that names Googlebot in its own group silently blocks every crawler invented since.
02
Server-rendered content
Stage two, Render: whether your words exist before JavaScript runs.
We read the raw HTML your server returns, with no browser and no JavaScript execution, and report whether the title, meta description, H1 and body copy are actually in it — plus the word count that survived and the framework we detected. A React or Vue app that renders entirely client-side can look empty to a crawler that does not execute scripts.
03
Schema.org markup
Stage three, Extract: whether your facts are extractable, not just readable.
We list the structured-data types found on the page and flag whether you have entity markup (Organization, Product), answer markup (FAQPage, HowTo) and article markup. Schema is how a model lifts a specific claim out of a wall of prose and attributes it to you rather than paraphrasing it away.
04
/llms.txt
Also stage three. Present or absent, graded as optional.
We check for the file and report its size. We also tell you plainly that the convention is experimental and unproven for citations, because a free tool that grades you red for missing an unratified standard is selling anxiety, not information.

Every finding comes back graded healthy / warning / error with the specific issue and the fix — not a number with no explanation.

THE OBVIOUS SHORTCUT

Why not just ask ChatGPT or Claude if they can see your site?

Most people try this first, and the answer always sounds convincing. It is also the wrong test. Chatting with an AI shows you what one bot saw once. Being cited depends on a different bot, on a different visit, following rules the chat never looks at.

The chat window is not the crawler
When you paste a link into ChatGPT, one bot goes and fetches it. A different bot builds the answers ChatGPT gives everyone else. Your site can let the first one in and still block the second. A friendly "yes, I can see your site" only proves the first.
It will answer even when it did not look
Ask an AI to describe your site and it will, from memory, from a guess, or from an old copy. The answer reads the same in every case, and it will not tell you which one you got.
Each AI only knows about itself
ChatGPT cannot tell you what Claude or Perplexity see, and Claude cannot tell you about ChatGPT. You would have to ask all of them, and each one still only speaks for its own chat bot.
It cannot report what it never received
If your page is blank until JavaScript runs, the AI fills the gap and sounds fine. It will not say "your page arrived empty", which is the one thing you need to know.
Ask twice, get two answers
Chat answers change from one try to the next. A check has to give the same result every time, so you can fix something, run it again, and know the fix worked.

What the scan does instead: it visits your site the way the crawlers do, with no guessing, and gives you the same graded result every time. Fix something, run it again, and watch the grade change.

SAME SCAN, OTHER HALF

Looking for the classic SEO side?

This page leads with AI visibility because that is where most sites are quietly broken. The same scan also grades on-page SEO, crawlability, search-index presence and your measurement tags — if that is the angle you came for, the free SEO + AI visibility check frames the same report around search.

Want the full picture — which queries you rank for, which AI answers cite you, and where paid spend is leaking? Connect Search Console, GA4 and Google Ads with one click after the scan, and Alice starts from the results you just got.

QUESTIONS

AI visibility, answered.

How do I check if ChatGPT can see my website?+

Three things have to be true. First, your robots.txt must allow the crawlers that actually feed citations, which are the search agents rather than the training ones: OAI-SearchBot and ChatGPT-User for ChatGPT, Claude-SearchBot and Claude-User for Claude, PerplexityBot and Perplexity-User for Perplexity, and Googlebot for Google AI Overviews. The training-only crawlers (GPTBot, ClaudeBot, CCBot, Google-Extended) are a separate, optional choice: blocking them opts your content out of model training, but it does not stop the search side from reaching or citing you. Because each vendor runs distinct agents for training and for search, and robots.txt groups are exclusive, allowing one says nothing about the other. Second, your page content must exist in the raw HTML rather than being rendered by JavaScript in the browser. Third, your key facts should be marked up with schema.org so they can be extracted cleanly. The scan on this page tests all three against your live URL in about 30 seconds.

Is an AI visibility check different from an SEO check?+

Yes, though they overlap. An SEO check asks whether Googlebot can crawl, index and rank a URL so a human clicks it. An AI visibility check asks whether an answer engine can fetch, parse and quote your content inside a generated answer where there may be no click at all. The most common failure mode is a site that scores well on SEO and badly on AI visibility, because Googlebot is explicitly allowed in robots.txt while the AI crawlers — all invented after 2023 — are not.

What is GEO (Generative Engine Optimization)?+

GEO is the practice of getting your brand, URL or content cited inside an AI-generated answer — Google AI Overviews, ChatGPT, Perplexity, Claude, Gemini — rather than ranking as a blue link. AEO (Answer Engine Optimization) is the closely related discipline of structuring content so a machine can lift a clean, complete answer out of it. Neither replaces SEO; they are what you do on top of it when the user may never click through.

Does blocking GPTBot in robots.txt hurt my traffic?+

It does not hurt your Google rankings, because GPTBot has nothing to do with Googlebot, which is exactly why blocking it is so easy to miss: nothing in Search Console or GA4 shows a symptom. But GPTBot is the OpenAI training crawler, not a citation crawler. Blocking it opts your content out of model training and does not stop ChatGPT Search from reading or citing you. The crawlers that gate ChatGPT citation are OAI-SearchBot (the search index) and ChatGPT-User (the live fetch when someone asks about your page), so those are the ones to check. Blocking GPTBot is a legitimate choice if you do not want your content used for training, but it should be a decision you made on purpose, not one a robots.txt file inherited from 2019 made for you.

Is /llms.txt worth adding?+

Probably harmless, currently unproven. /llms.txt is a proposed convention for giving language models a curated map of your site, and no major AI engine has publicly committed to honouring it as a ranking or citation input. We report whether you have one because it costs nothing to check, but we grade it as optional and would not put it ahead of fixing crawler access or server-rendered content.

Does this scan tell me whether ChatGPT actually cites my site?+

Yes, for two questions. The free scan asks Google AI Overviews, ChatGPT and Claude one brand-name question (does AI know you when asked about you by name — recognition) and one buyer question about your category (does AI name you when a buyer who has never heard of you asks — selection), live, and shows per engine whether you were named, who was named instead, or whether that engine gave no AI answer at all. Treat it as a point-in-time snapshot, not a verdict: answers vary by query and by week, which is why Alice re-checks the buyer questions weekly once you sign up. The ChatGPT reading is an OpenAI model with web search, not the ChatGPT app; the Claude reading is the Anthropic API with web search, not claude.ai. And capability comes first — a site the crawlers cannot read will never be cited, no matter how the snapshot looks.

GO DEEPER

The reading behind the checks.

A founder blocked ChatGPT’s crawler for two years — and nothing looked wrong
The exclusive-groups mechanism behind most silent AI-crawler blocks, and the 30-second fix.
SEO vs. GEO vs. AEO: what actually changed (and what didn’t)
A row-by-row comparison of the three disciplines, and which metrics replace rank.
Why Google AI Overviews aren’t optional anymore
AI Overviews eat the click before users scroll — and cite sites that don’t rank top 10.