How this works
AI assistants like ChatGPT, Claude, Perplexity, and Google's AI Overviews answer questions by fetching and quoting from real websites. To do that, your site needs to be reachable, parseable, and trustworthy by the standards these assistants apply.
This audit checks the technical and structural signals that determine whether your site shows up in those answers. We don't run the AI tools and check whether they cite you — that's a different kind of test. Instead, we check whether the foundations are in place so that AI assistants can cite you.
We check your robots.txt rules against the 11 major AI crawlers (GPTBot, ClaudeBot, PerplexityBot, OAI-SearchBot, Google-Extended, CCBot, Bytespider, Applebot-Extended, and others). We also fetch the homepage as each bot, look for CDN-level blocks, and check for an llms.txt and sitemap.
We check page structure (one main heading, sensible paragraph length, lists), structured data (JSON-LD, Open Graph, Twitter Cards), and how much of your content needs JavaScript to render. Pages that look fine to a human but appear empty to a bot get flagged.
We check trust signals (HTTPS, HSTS, author markup with verifiable links) and content-shape signals (TL;DR-style intros, question-style headings, fresh dates, descriptive internal links) that nudge AI assistants toward citing you.
Some sites — typically large brands behind Akamai, Cloudflare Bot Management, or AWS WAF — block our request at the front door because we come from a generic cloud IP. When that happens, we route the content fetches through a managed proxy so the rest of the audit can still complete.
The one thing we can't tell from outside is whether real AI crawlers (GPTBot, ClaudeBot, PerplexityBot) are explicitly allowlisted on your CDN. That's an internal question for your IT/security team. The audit calls this out clearly when it applies.
The audit is anonymous — we don't collect your email or sign you up for anything. Each report lives at a UUID-based URL that anyone with the link can view. There's no public directory. Audits are kept for analytics purposes.
For technical legibility we record the IP that requested the audit (used for rate limiting), and we store the rendered HTML of pages we sampled. We don't share audit data with third parties.