AI visibility audit: can AI engines read you, and do they cite you?
An AI visibility audit answers two questions. Can answer engines such as ChatGPT and Google AI Overviews read your pages and lift a clean answer from them? And when people ask about your field, do those engines actually name you? BeacX scores the first on your own pages and measures the second by asking the engines.
The readiness part comes with every BeacX audit. The measured part is in the AI visibility audit, because each question asked costs us money with the engines, and we would rather show that than hide it.
Start with the free readiness check
Why keep readiness and citations apart?
Because they need different work, and mixing them sells the wrong fix. A site can have pages that are easy to quote and still never be cited. A site can be cited today on weak pages and lose it after one engine update.
| Not cited yet | Cited | |
|---|---|---|
| Pages ready | Authority gap. The pages are fine; engines quote other sites. The work happens off your site: reviews, directories, mentions in the press and on trusted lists. Rewriting the pages will not fix it. | Working. Keep it that way, and let monthly checks catch the day a page stops being readable. |
| Pages not ready | Content gap. Fix the pages first, or nothing else counts. | Visible on weak pages. Strengthen them before an engine update takes the advantage away. |
The quadrant is what we tell you, more than the score. A rewriting programme sold against an authority gap cannot work, and we would rather say so before you pay for one.
What the readiness score measures
The score is out of 100 and averages four layers with equal weight. Every one of them is something you can change on your own site.
| Layer | The question it asks | What we look at |
|---|---|---|
| Crawler access | Can AI bots fetch the pages? | robots.txt rules for GPTBot, OAI-SearchBot, ClaudeBot, PerplexityBot and Google-Extended; noindex; pages that only render with JavaScript |
| Entity clarity | Can engines tell what the site is? | structured data, the business type you declare, a consistent name, address and phone |
| Answer extractability | Can an answer be lifted cleanly? | a direct answer at the top of each page, passages short enough to quote, question headings, llms.txt |
| Authorship and trust | Can they see who wrote it, and when? | named authors, update dates, credentials, and on health sites a separate medical reviewer |
A layer we could not measure is left out of the average and named in the report. It is never counted as zero, because a zero would report your site as broken when the gap was ours.
Our own site, joycheck.io, scored 87 on 17 October 2026: 100 on the first three layers and 48 on authorship and trust. The fix that would move it most is a visible update date on 37 of 85 pages, worth 3.9 points. See that report in full
How the citations are measured
We do not guess citations from the crawl. We ask the engines a fixed set of questions about your field and read the answers.
The question set is agreed with you before the first run and then pinned. Every later run asks the same questions, word for word, so this quarter can be compared with the last. When the set changes, the report starts a new line instead of drawing a false trend.
Today we measure two places where answers carry links: Google AI Overviews and ChatGPT with web search. Engines that answer from memory alone cannot cite anyone, so we never score them on citations.
Answers move between runs. The same question asked twice can get different sources, and an engine update can reshuffle them overnight. That is why we show how many questions were asked and how many answers came back, and why we never promise that any engine will cite you.
What you receive
- The one-page verdict: the readiness score, the quadrant, and the fixes worth the most points.
- Page-level findings for each layer, checked by a person against the live page.
- Paste-ready structured data for the pages that need it.
- With the measured add-on: the questions asked, which sites each engine cited instead of you, and the pages of yours that should have answered.
Prices
| Package | Price | Includes |
|---|---|---|
| Free readiness check | $0 | The one-page verdict for your site |
| AI visibility audit | $499 | Readiness on every page, the quadrant, checked findings, schema files, and one measured run: up to 50 agreed questions on Google AI Overviews and ChatGPT search, compared with three competitors |
| Full SEO and AI audit | $749 | Everything above except the measured run, plus the full SEO workbook, Search Console opportunities and a 45-minute call |
| Monitor and Fix | $449 a month | Monthly re-checks, fixes made by us, and a measured citation run every quarter |
Comparable AI audits we found cost between $399 and £1,500 [3][4]. Ours is about half the average, with the page-level fixes included.
Questions about the AI visibility audit
Is this the same as GEO or AEO?
They overlap. Answer engine optimisation (AEO) is about being the answer an engine gives. Generative engine optimisation (GEO) is about being cited as a source. The readiness score covers the on-page side of both, and the measured run checks the citations.
Can you get us cited by ChatGPT?
No one can promise that honestly. We can make sure nothing on your site stands in the way, and show you who is being cited instead.
Does blocking AI crawlers hurt us?
It depends on which crawler. OpenAI runs GPTBot for model training and OAI-SearchBot for ChatGPT search, and says each setting is independent [1]; block the search one and you drop out of ChatGPT’s search answers. Google says Google-Extended does not affect inclusion in Google Search [2]. The audit lists each crawler your robots.txt allows or blocks, so you can decide on purpose.
Do you check Arabic and Turkish pages?
Yes. Pages in Arabic, Turkish, Urdu, Azerbaijani and German are scored in their own language. English-only rules mark good non-English pages down; we found that in our own engine and fixed it.
Start with the free check
The free check is the readiness half for your site. If it shows an authority gap, the measured run is the next step. If it shows a content gap, fix the pages first.
Sources
- OpenAI, Overview of OpenAI crawlers: https://developers.openai.com/api/docs/bots (read 20 Oct 2026)
- Google Search Central, Google's common crawlers: https://developers.google.com/search/docs/crawling-indexing/google-common-crawlers (read 20 Oct 2026)
- Travel Tractions, AI visibility audit, $399: https://traveltractions.com/services/ai-visibility-audit/ ; Rank GEO Pro, GEO Blueprint Audit, $750: https://rankgeo.pro/pricing/ (read 20 Oct 2026)
- Novi Digital, AI search visibility audit, GBP 1,500: https://novi.digital/ai-search-visibility-audit/ (read 20 Oct 2026)