Is my website visible to AI? A 30-minute AI visibility audit

What visible to AI actually means

AI search tools do not see the web directly. They retrieve pages from search indexes and then decide which sources to cite. Google says its AI Overviews and AI Mode are rooted in its core Search ranking systems, and ChatGPT search relies on OpenAI's own crawler plus third-party search providers, including Bing. So visibility has layers: crawlers must be allowed in, your content must be readable, pages must be indexed, your firm must be understood, and finally it must be chosen.

Automated AI visibility checkers can speed up parts of this, but none can see inside the models. Google's guide to generative AI features warns that no third-party tool has access to its internal ranking or AI systems. The checks below test what you can actually control.

Step 1: Check robots.txt for AI crawlers (3 minutes)

Open yourdomain.com/robots.txt in a browser. Look for User-agent lines naming the crawlers below, and for a blanket rule (User-agent: * followed by Disallow: /) that blocks everyone. If a crawler is not mentioned, the * rules apply to it.

robots.txt is not the only gate. Some CDN, hosting and security products, Cloudflare among them, offer settings that block AI crawlers at the firewall regardless of robots.txt. Ask whoever manages your hosting whether such a setting is on.

Step 2: Confirm key content is in the HTML (3 minutes)

Open your homepage and your most important service page, right-click and choose View page source. Use find (Ctrl+F or Cmd+F) to search the source for your firm name, your main service, your city and your phone number.

If those words appear, crawlers can read them. If the source is mostly script tags and the text only appears after the page runs JavaScript, you have a risk: Google can render JavaScript, but not every crawler renders pages the way Google does. Also look for key information that exists only inside images, PDFs, or tabs that load content on click. The fix is usually server-side rendering or static HTML for core pages, which your web developer will recognize.

Step 3: Check indexing in Google and Bing (5-7 minutes)

1. Run a site: searchSearch site:yourdomain.com on both Google and Bing. Your homepage and main service pages should appear. If Bing shows nothing, ChatGPT search has less to work with.
2. Inspect key URLs in Search ConsoleIn Google Search Console, check the Page indexing report and use URL Inspection on your homepage and top service page. Google says a page must be indexed and eligible to show a snippet to appear in its AI features. Its Generative AI performance report shows how your content performs in those features.
3. Set up Bing Webmaster ToolsBing Webmaster Tools can import a verified site from Search Console in a few clicks. Submit your sitemap and consider IndexNow, a protocol Bing supports that lets your site notify search engines when pages change; many CMS plugins and hosts support it.

Step 4: Check entity consistency (5 minutes)

AI answers are only as accurate as the sources they find. Put these side by side: your homepage, About and Contact pages, Google Business Profile, Bing Places, LinkedIn, and the directories or association listings your clients use. Compare the firm name, address, phone, services, service area and principals. Old addresses, former names and outdated service lists are common culprits.

Then ask ChatGPT, Gemini and Perplexity to describe your firm by name: what does it do, and where is it based? Wrong or vague answers usually point to the inconsistent or missing source.

Step 5: Validate structured data (3 minutes)

Paste your homepage URL into Google's Rich Results Test and the Schema Markup Validator. For a professional-service firm, look for Organization, LocalBusiness or ProfessionalService markup with your name, address, contact details and links to official profiles, and check that it matches what the page visibly says.

Keep this in proportion. Google states that structured data is not required for its AI features and there is no special schema for them. Valid markup helps with rich results and disambiguation; errors or markup that contradicts the page are worth fixing.

Step 6: Run a fixed prompt panel (10 minutes, then monthly)

This is the closest thing to a direct answer to is my website visible to AI. Write five to eight prompts a real client would use, mixing one brand prompt with category prompts such as structural engineer for a soft-story retrofit in San Francisco. Keep the wording fixed so results are comparable over time.

Run each prompt 5-10 times in ChatGPT with search on, Gemini and Perplexity, in a logged-out or private window where possible. Answers vary between runs, so record how many runs name your firm, whether your site is cited, which competitors appear and which third-party pages are cited. Those cited pages show you where earned mentions would matter most. Our guide on how to get ChatGPT to recommend your business covers what to do with the results.

Step 7: Check whether AI agents can use your site (5 minutes)

Agents are the next layer: assistants that visit your site to compare firms or start an inquiry. Google notes that browser agents read pages through screenshots, the DOM and the accessibility tree. In Chrome DevTools, open the Elements panel and the Accessibility pane, then click your main buttons and contact form fields. Each should have a clear name such as Request a consultation, not an unlabeled icon, and every form field should have a label.

Chrome is adding an informational Agentic browsing category to Lighthouse (Chrome 150 and later) that checks accessible names, layout stability and WebMCP or declarative form signals, as described in A developer toolkit to make your website agent-ready. Treat it as guidance, not a score to chase. WebMCP itself is still behind an origin trial or flag as of 2026; trugoods.net registers a WebMCP tool, prepare_trugoods_consultation, that lets a browser agent prepare a consultation request while a human still submits it. More in our WebMCP guide.

Step 8: Cloudflare's agent-readiness scan: what matters and what is optional

Cloudflare's Is it agent ready scanner checks a site for emerging agent standards. Across Cloudflare Radar's top roughly 200,000 domains it reported about 78% with robots.txt, about 4% with Content Signals, about 3.9% supporting Markdown negotiation, and fewer than 15 sites with MCP Server Cards or API catalogs. A low score is normal, including for well-known sites.

• Matters now: robots.txt and crawler access. This overlaps with Step 1 and directly affects whether AI search can read you.
• Optional and cheap: Content Signals (a way to state in robots.txt how content may be used, for example for search versus AI training) and llms.txt. Google says Search ignores llms.txt, and its value elsewhere is unproven; see our llms.txt guide.
• Forward-looking: Markdown negotiation, MCP Server Cards, API catalogs and MCP endpoints. trugoods.net runs a read-only MCP endpoint and a JSON service catalog as live implementations, but for most small firms these come after the basics, not before.

Score your site

Give each row 2 points for pass, 1 for partial and 0 for fail. The bands are our working rule of thumb, not an industry standard: 15-18 means the fundamentals are in place and the work is earning mentions; 9-14 means fix the gaps before anything else; below 9 means start with crawler access, HTML content and indexing. For a structured version of this audit with fixes prioritized, see SEO and GEO at TruGoods.

Frequently asked questions

What is an AI visibility audit?

It is a structured check of whether AI search tools and agents can access, read, understand and cite your website, and whether they actually mention your business. A useful audit covers crawler access, rendering, indexing, entity consistency, structured data and repeated prompt tests. It should report a dated baseline, not promise rankings.

How do I make my website visible to AI?

Allow search crawlers such as Googlebot, Bingbot, OAI-SearchBot and PerplexityBot, keep key content in the HTML, get indexed in Google and Bing, and describe your firm consistently everywhere. Beyond that, AI answers lean on independent sources, so earned mentions and genuine reviews matter. There is no special AI file that replaces these basics.

How do you measure AI visibility?

Run a fixed set of prompts 5-10 times each across ChatGPT, Gemini and Perplexity and track the share of runs that name or cite you. For Google, use the Generative AI performance report in Search Console. Repeat monthly with the same wording so changes are meaningful.

Can ChatGPT do an SEO audit?

It can explain concepts, review text you paste in, and sometimes fetch a page, but it cannot see your Search Console data, server settings or firewall rules, and it can be confidently wrong. Use it as an assistant alongside the free tools in this audit, not as the audit itself.

How do I optimize my website for AI agents?

Start with accessibility: clear names on buttons and links, labeled form fields, stable layouts and content in the HTML. Then, if it fits your business, add structured service data and consider emerging options such as WebMCP. These are early-stage, so treat them as experiments rather than requirements.

Is SEO dead now with AI?

No. Google states that its AI features are built on its core Search ranking systems and that optimizing for them is still SEO. What has changed is the scoreboard: being named and cited inside answers matters more, so third-party mentions and clear entity facts deserve more attention.

Related guides

• How to get ChatGPT to recommend your business (without tricks)
• SEO and AI search for architecture firms: how to become the studio ChatGPT and Google can confidently recommend
• What is llms.txt, and is it worth adding to your website?
• WebMCP: how to make your website agent-ready, and how it differs from MCP