Enter a URL to check it against 9 signals that matter for AI answer engines like ChatGPT, Perplexity, and Google AI Overviews — separate from classic search-engine SEO.
GEO Audit Tool checks a page the way an AI answer engine effectively "reads" it, rather than the way a classic search-engine crawler scores it. Generative Engine Optimization (GEO) is the newer, related discipline to SEO: instead of optimizing mainly to rank in a list of blue links, it looks at what makes content easy for tools like ChatGPT, Perplexity, Google AI Overviews, or Gemini to understand, extract, and confidently cite in a generated answer. Enter a URL and the tool fetches the live page and checks it against nine of the signals that matter most for that: a clear, quotable direct-answer near the top, structured data (schema.org JSON-LD), a sensible heading hierarchy and question-phrased headings, content depth, extractable lists and tables, author/date signals, outbound citations, title and meta description quality, and whether known AI crawlers are blocked in robots.txt.
Each check comes back as a pass, warning, or fail with a plain-language explanation of what was found and why it matters, plus an overall score out of 100 so you can track a page over time or compare a few pages at once. The tool also does a best-effort read of /llms.txt on the same domain - an emerging, still-informal convention some sites use to describe their content to language models directly - and reports whether one exists without penalizing pages that don't have it, since it isn't (yet) a settled standard.
Being upfront about the limits: this reads the raw HTML a normal server response returns, the same way most AI crawlers do - it does not execute JavaScript, so a page whose real content is rendered client-side after load may score lower here than it "looks" in a browser. The direct-answer, byline, and date checks are pattern-based heuristics, not a guarantee of how any specific AI system will actually treat the page; and robots.txt is read with a simplified, common-bot-name check rather than a full path-matching parser. Treat the results as a solid, fast starting checklist, not a certified score.