How to Check Your Website's AI Visibility: A 5-Step Self-Audit

Research shows that companies with strong traditional SEO rankings appear in only 18% of AI-generated answers. Checking your website's AI search visibility takes five concrete steps: verify Bing indexing, audit your robots.txt and llms.txt, inspect your schema, test content extractability, and run a live citation spot-check across ChatGPT, Perplexity, and Gemini.

This guide walks through each step and explains what to look for. The final section covers what a professional audit checks that a self-check cannot.

Why strong Google rankings don't guarantee AI visibility

Ahrefs' analysis of 1.4 million ChatGPT prompts found that 88.46% of sources cited by ChatGPT came from pages already in Google's search index. That sounds like good news for anyone who ranks well. In practice, Google rankings are a necessary condition for AI citations, not a sufficient one.

AI systems add their own filters on top of indexing. Your content might be indexed and ranking while still being blocked by a CDN rule, too dense for an AI to extract a clean answer from, or lacking the third-party signals that give AI models confidence to cite it.

The gap comes down to three things:

• → Access — Can AI crawlers reach your content?
• → Understanding — Can they extract and interpret your key claims?
• → Authority — Do they trust your content enough to cite it?

The five steps below check each of these in order. Start with access before worrying about understanding or authority.

Step 1: Check your Bing indexing status

For AI systems outside Google's own ecosystem, Bing's search index is the primary source. Perplexity, Microsoft Copilot, and many other AI tools pull citation sources directly from Bing. If your pages aren't indexed there, those systems can't cite you regardless of how well you rank on Google.

How to check Bing indexing

Go to Bing Webmaster Tools at webmaster.bing.com. Under Reports, check your indexed page count against your expected site size. For a quick manual check, search site:yourdomain.com on Bing.com and compare the result count to the same query on Google. A significant gap points to a Bing indexing problem.

What a Bing indexing gap looks like

Common symptoms: recently published content takes weeks to appear in Bing, the Bing site: count is much lower than Google's, or key pages rank on Google but return nothing on Bing. The fastest fix is submitting URLs directly through Bing Webmaster Tools or using IndexNow for real-time submission.

Step 2: Check your robots.txt and llms.txt

Your robots.txt is the first file AI crawlers check when they visit your site. If they're blocked there, nothing else in this guide matters.

AI bots to check for in your robots.txt

Open yourdomain.com/robots.txt and check that none of these user agents are blocked:

A blanket User-agent: * Disallow: / catches all of these. More common, and harder to catch, is a block introduced by a CDN or WAF configuration that doesn't show up in robots.txt at all.

The Cloudflare silent-blocking problem

Cloudflare WAF configurations silently block AI crawlers on a significant portion of sites, even when robots.txt looks fine. One documented case: a brand was completely invisible to Claude AI despite ranking well on Google and Bing. The cause was a WAF rule categorising AI crawlers as bot traffic at the network layer — nothing in robots.txt indicated a problem.

To check, review your Cloudflare Bot Management settings and WAF rules for any policies that block bots or scrapers broadly.

Does your site need an llms.txt file?

llms.txt is a newer file standard that gives AI systems a structured map of your site's AI-readable content. Check whether one exists at yourdomain.com/llms.txt.

For Google AI Search, llms.txt is not required and has no documented effect. For ChatGPT, Perplexity, and Claude, it can help those systems navigate your content more efficiently. If you're targeting non-Google AI citations and don't have one yet, it's a straightforward addition.

Step 3: Check your schema and structured data

Google removed FAQ rich results in May 2026. But FAQ schema didn't lose its value. It shifted from being a rich snippet driver to being one of the primary mechanisms AI systems use to extract structured answers from content pages.

The minimum schema baseline for AI visibility

Every site should have JSON-LD Organization and WebSite schema on the homepage. These give AI agents a machine-readable identity signal: who you are, what your URL is, what you do. For content pages, FAQPage and HowTo schema are the most useful types for AI extraction.

One nuance worth knowing: an Ahrefs correlation study found that schema markup alone had no direct citation uplift in Google AI Overviews. Schema works as a parsing and identity layer, not a standalone ranking signal. Getting the baseline right matters; stacking additional schema types beyond it has diminishing returns.

How to check your current schema

Run your homepage and key content pages through Google's Rich Results Test. Look for:

• → Organization on the homepage
• → Article or BlogPosting on blog content
• → FAQPage on pages with question-and-answer content
• → BreadcrumbList for site hierarchy signals

Missing Organization schema on the homepage is the most common gap.

Step 4: Test your content extractability

Access and schema are about whether AI systems can reach and identify your content. Extractability is about whether they can accurately quote it.

The quick-answer block test

Look at the opening of each key page. Is there a direct, self-contained answer to the primary question within the first 40 to 60 words?

AI systems prefer content that front-loads its answer. If your introduction spends two paragraphs setting context before reaching the actual claim, AI systems may extract from a different part of the page or skip to a competitor that answered immediately.

A clear quick-answer block states the answer in the first two sentences. An opening that begins "The world of AI search has changed dramatically..." is not independently extractable.

The live extraction test

Take a key section from one of your important pages. Paste it into ChatGPT or Perplexity and ask: "Based only on this text, what is the main answer?" If the AI returns a clear, accurate summary that reflects your framing, the content is extractable. If it returns something generic or misses your main point, that section probably needs restructuring: a clearer heading, shorter sentences, or a direct-answer opening.

Step 5: Run a live AI citation spot-check

The most direct way to check your AI search visibility is to ask AI systems about your brand and document what comes back.

What queries to run

Run these three query types across ChatGPT (with web search enabled), Perplexity, and Gemini:

• → Branded: "Tell me about [Brand Name]" or "What does [Brand] do?"
• → Topic-adjacent: "What should I know about [your main category]?" — note whether your brand appears in the cited sources
• → Problem-specific: "How do I [the primary problem your product or service solves]?" — check whether your content is referenced

What to look for

Document the result for each query-platform combination:

This spot-check creates your AI visibility baseline. Without a before-state, you have no way to measure whether changes are working.

Free tools that automate citation tracking across platforms: Ahrefs AI Visibility Checker and Semrush AI Brand Visibility Tool.

What a professional AI visibility audit checks that this guide does not

A self-check surfaces the most common and visible gaps. A professional AI visibility audit goes further on signals a manual check can't reach.

• →
  Server-level crawler simulation — Testing whether AI bot user agents actually receive your full page content, not a blocked or degraded response, even when robots.txt looks clean.
• →
  Entity disambiguation — Checking whether AI systems correctly identify your brand, or conflate it with similar entities or outdated training data.
• →
  Source ecosystem audit — Reviewing third-party signals outside your website: directory listings, review platforms, Reddit mentions, media references, and partner pages that shape what AI models have learned about your brand.
• →
  Cross-platform citation consistency — Mapping how your brand is cited across ChatGPT, Perplexity, Gemini, and Google AI Overviews, and identifying where facts are inconsistent or missing.

Frequently asked questions

If you've worked through all five steps, you have a baseline picture of where your AI visibility stands. The most common gaps — Bing indexing misses, Cloudflare-blocked crawlers, missing Organization schema — are fixable without specialist help. Deeper issues, including entity confusion and off-site source gaps, are what a professional AI Visibility Audit is designed to surface.