All posts
Web design & development

AI Visibility Website Audit: How to Check if Answer Engines Can Use Your Site

An AI visibility audit checklist: test rendering without JavaScript, crawler access, structure, entities and AI referrals to see if answer engines can use your site.

By Mehdi Hassan7 min read

The short answer

An AI visibility audit checks whether answer engines like ChatGPT, Perplexity and Google's AI Overviews can fetch, read, understand and cite your website. It tests five layers in order: rendering without JavaScript, crawler access through robots.txt and the CDN, indexing, answer-first page structure and clear entities, then measures AI referrals and real prompt results.

Key takeaways

  • Audit in order of what blocks citations outright: rendering and access come before content.
  • Fetch your key pages without JavaScript; if the copy isn't in the HTML, most AI crawlers can't see it.
  • Check CDN and firewall logs for AI user agents getting 403s, not just robots.txt.
  • Test real buyer prompts in ChatGPT and Perplexity and record who gets cited instead of you.
  • Schema and llms.txt belong at the end of the audit, as hygiene, not the headline findings.

Part of How to Build an AI-Search-Ready Website: Structure, Schema and Crawlability

What is an AI visibility audit?

An AI visibility audit is a technical and content review that answers one question: can answer engines use this website? It checks whether AI crawlers can reach your pages, whether the content is in the HTML they read, whether pages contain quotable answers and whether engines can tell who you are. Some people call it a GEO audit or an AEO audit; the checks are the same.

It's different from tracking where you're mentioned in AI answers. Measurement tells you the outcome; the audit tells you why. This checklist covers the build side, which most AEO audits skip, and it maps to the five layers in our guide to an AI-search-ready website. For the reporting side, see how to measure AI search visibility.

layers to audit, in order: rendering, access, indexing, structure, entities
5
free intro call to walk through your results with us
30 min
Designify trial week, one week of design and development fixes
$1,000

Step 1: can AI crawlers read your pages without JavaScript?

Start here, because it's the most common silent failure. Most AI crawlers fetch the raw HTML and don't run JavaScript, so if your headings, copy and tables are added in the browser, those crawlers see a near-empty page. Googlebot renders JavaScript, which is why a site can rank in Google and still be invisible to ChatGPT.

  1. Fetch the raw HTML. Run curl on your homepage, a product page and your pricing page, and search the output for a sentence from the hero and a price.
  2. Disable JavaScript in your browser. In Chrome DevTools, turn off JavaScript and reload. What you see is roughly what a non-rendering crawler gets.
  3. Compare view-source with the inspector. If text appears in the inspector but not in view-source, it's rendered on the client.
  4. Check FAQs, tabs and accordions. Content loaded on click from an API isn't in the HTML; content that's in the HTML but visually collapsed is fine.

If key content fails these checks, the fix is server-side or static rendering. Our guide to server-side rendering for AI crawlers explains the options in Next.js and other frameworks.

Step 2: are AI crawlers allowed in?

Next, confirm that the crawlers behind AI search can reach your public pages. A crawler can be blocked in robots.txt, by your CDN or firewall, or by a platform toggle, and each one has to be checked separately.

  1. robots.txt: confirm Googlebot, Bingbot, OAI-SearchBot, ChatGPT-User and PerplexityBot are allowed on public paths. Our guide to allowing or blocking AI crawlers lists the user agents.
  2. CDN and firewall: check bot-management, AI crawler blocking and JavaScript challenge settings.
  3. Server logs: filter by AI user agents and look for 403, 429 or challenge responses on key pages.
  4. Platform settings: look for AI crawler toggles in your website builder or host and make sure they match your intent.

Step 3: are your key pages indexed where AI engines look?

Check that your important pages are indexed in Google and Bing, because AI Overviews draw on Google's index and several assistants, including Copilot, use Bing's. A page that isn't indexed rarely gets cited, however good it is.

  1. Google Search Console: inspect key URLs, check the Pages report for excluded money pages and look for stray noindex tags.
  2. Bing Webmaster Tools: verify the site, submit the sitemap and check that key pages are indexed.
  3. Snippet controls: make sure no nosnippet or tight max-snippet rules are set on pages you want quoted.
  4. Canonicals and duplicates: confirm each page's canonical points to itself, not to an old domain or a staging URL.

Step 4: does each page contain a quotable answer?

Now review the content itself. For each money page, ask whether an engine could lift a passage that answers the buyer's question on its own. If the answer is buried, split across a carousel or locked in an image, it won't be quoted.

  1. Short answer near the top: a 40–60-word passage that states what the page is about. See answer-first page structure.
  2. Question-style headings: H2s that mirror real queries, each answered in its first sentence.
  3. Real tables: prices, plan comparisons and specs as HTML tables, not images.
  4. Visible FAQ: four to six genuine buyer questions, answered on the page.
  5. Dates and authors: a visible updated date and a named author or team on guides and comparisons.

Step 5: can engines tell who you are?

Finally, check your entity signals: the facts that let an engine connect your website to your company and describe it correctly. Startups with generic names are often confused with other businesses, and engines can't cite you confidently if they're not sure which company you are.

  1. One consistent name and description across the homepage, about page, LinkedIn, Crunchbase and directories.
  2. A clear about page that says what you do, for whom, where you're based and who runs the company.
  3. Organization schema with sameAs links to official profiles. It won't lift citations, as our piece on schema markup and AI citations explains, but it reduces confusion.
  4. Descriptive titles that name the product category, not just the brand.
Where do llms.txt and schema fit in the audit?

At the end, as hygiene. 2026 studies, including SE Ranking's analysis of 300,000 domains, found no measurable citation lift from llms.txt, and Ahrefs' controlled test found no significant lift from adding schema. We note whether they exist and whether they're accurate, but we never put them at the top of a findings list.

If an audit report leads with llms.txt or AI-specific schema while your pricing page renders on the client, it has the priorities backwards. Our llms.txt explainer covers the evidence.

How do you measure the results of an AI visibility audit?

Measure with two things: AI referral traffic in analytics and a fixed set of buyer prompts you test by hand. Neither is perfect, but together they show whether fixes are reaching the engines and whether buyers are arriving from them.

  1. Set up AI referral tracking. Create a channel group for ChatGPT, Perplexity, Copilot, Gemini and Claude; our guide to tracking AI referral traffic in GA4 walks through it.
  2. Write 10 to 20 buyer prompts. Use the questions prospects ask on sales calls, such as which tool is best for a specific use case.
  3. Run them monthly in ChatGPT, Perplexity and Google, and record whether you're cited, mentioned or absent, and who is cited instead.
  4. Re-run the technical checks after every redesign, platform change or CDN change, since those are when access breaks.

Frequently asked questions

How long does an AI visibility audit take?
The technical checks for a small marketing site, such as rendering, robots.txt, logs and indexing, can be done in an afternoon by someone who knows where to look. The content review takes longer, because each money page needs a careful read. Larger sites with many templates take more time, but you can audit one page per template rather than every URL.
Can I do an AI visibility audit myself?
Yes, most of it. Fetching pages with curl, disabling JavaScript, reading robots.txt and checking Search Console need no special tools. The harder parts are reading CDN logs, diagnosing rendering problems in a JavaScript framework and rewriting pages answer-first. Many founders do the checks themselves and bring in help for the fixes.
Is an AI visibility audit different from an SEO audit?
They overlap heavily, since Google's AI features rely on normal SEO foundations. An AI visibility audit adds checks that a classic SEO audit can miss: AI crawler access, rendering for bots that don't run JavaScript, Bing indexing and passage-level quotability. A good SEO audit already covers indexing, canonicals and titles, so you can combine the two.
What tools help with an AI visibility audit?
Free tools cover most of it: curl or your browser's DevTools for rendering, Google Search Console and Bing Webmaster Tools for indexing, your CDN or host's logs for crawler access, and GA4 for referrals. Paid AI visibility trackers can automate prompt testing across engines, but run the free checks first, since they find the problems that block citations entirely.
How often should I repeat the audit?
Re-run the technical checks after any redesign, platform migration, CDN change or new bot-protection setting, because those are when crawler access and rendering break. Review prompts and referrals monthly. A full content review each quarter keeps money pages answer-first as products, prices and competitors change.

· CEO & Design Director

Copywriter, designer and developer, and founder of Designify Studio. Over five years, 60+ projects across 14+ industries, for startups that have raised $300M+.

Want us to do this for you?

We take craft seriously over at Designify Studio. We do deep research and do what we do best - pay attention to the smallest details and make sure we speak to the right audience.

Schedule Call