How to run an AI-readiness audit
To audit AI readiness, work in three steps: confirm AI crawlers are allowed and your facts are in raw HTML, check your key pages are extractable (answer-first, sourced, schema), and baseline whether AI already names you.
Step 1: can they reach you?
Check robots.txt allows GPTBot, ClaudeBot and PerplexityBot, and that your important facts appear in server-rendered HTML rather than only after JavaScript. If an engine can't crawl or read the page, nothing else matters.
Step 2: can they extract you?
Open your top commercial pages and ask three questions. Does each section answer its heading in the first sentence? Are claims backed by a number, a source and a year? Is there FAQ and Organization schema? Gaps here are why a ranking page goes uncited.
Step 3: do they already name you?
Ask ChatGPT, Perplexity and Google your category questions and record whether your brand appears, where, and against which competitors. That baseline tells you the size of the gap and gives you a number to improve.
Frequently asked
Baseline now, then re-check quarterly or after major site changes. AI answers shift day to day, so track the trend across months rather than reacting to single results.
Fix crawlability first: allow the AI bots in robots.txt and move key facts into raw HTML. Extraction and citation work is wasted until the engine can actually read the page.
You can start free with URL Inspection, a raw page fetch, and manual questions to the AI engines. A citation tracker automates the third step across engines at scale.
Grade your site against our 250-point AEO/GEO checklist and see exactly what to fix, free, no signup.