A complete AI-readiness audit checks four layers: whether AI crawlers can access your site, whether you've published the AI-specific files they look for, whether your structured data is in place, and whether your content is structured for extraction. Work through them in that order — each layer is a prerequisite for the next one to matter.
Step 1: Confirm AI crawlers can actually reach you
None of the later steps matter if AI bots are blocked from your site in the first place. Start here.
- Run your domain through the AI Crawler Access Checker — it fetches your live robots.txt and reports allow/block status for every major AI bot.
- Look specifically for unintentional blocks: a wildcard
Disallow: /rule underUser-agent: *blocks every bot, including AI crawlers, even if you never meant to target them specifically. - Decide your stance on training vs. search bots (see our guide on the trade-offs) and fix any mismatch between your stated intent and your actual rules using the robots.txt Generator for AI Bots.
Step 2: Check for llms.txt
- Visit
yoursite.com/llms.txtdirectly — does it exist? - If yes, run it through the llms.txt Validator to confirm it follows the spec (single H1, blockquote summary, H2 sections, correctly formatted links).
- If no, generate one with the llms.txt Generator — prioritize your most important 10-20 pages over an exhaustive list.
Step 3: Verify structured data
- Check whether your key pages have any JSON-LD structured data at all — view page source and search for
application/ld+json. - For content with genuine Q&A sections, confirm FAQPage schema is present and matches the visible content (see does FAQ schema still matter for context). Generate or fix it with the FAQ Schema Generator.
- For articles and instructional content, confirm Article or HowTo schema is present, using the Article & HowTo Schema Generator as needed.
Step 4: Review content structure
For each priority page, check:
- Does the first paragraph directly answer the page's core question? Or does it open with scene-setting, a definition of the general topic, or a company-history preamble?
- Is there exactly one H1, and does it match the actual query intent of the page?
- Are supporting details broken into lists/tables where the underlying information is naturally structured (steps, comparisons, specs)?
- Are subheadings specific and, where relevant, phrased as questions?
See the full structure checklist for AI Overviews for a more detailed pass on this step.
Step 5: Check meta title and description
Run your key pages through the Meta Title & Description Previewer — confirm titles and descriptions are present, appropriately sized, and accurately (not just enticingly) describe the page content, since AI systems often lean on meta descriptions when generating citations.
Step 6: Run the consolidated audit
Once the above is addressed, run the page through our AI-Readiness Audit tool, which automates checks across llms.txt presence, AI-bot access, meta tags, H1 structure, structured data presence, and a first-paragraph heuristic — giving you a single score and a prioritized checklist of what's still failing.
How often to re-audit
This category moves fast — new AI crawlers appear, the llms.txt spec may evolve, and platforms change how they weight structured data and content signals. Re-run this process on a monthly cadence for your most important pages, and immediately after any major site redesign, CMS migration, or robots.txt change (a common source of accidental AI-bot blocking).
What a passing audit does and doesn't mean
A clean audit across all six steps means the structural, technical prerequisites for AI visibility are in place — it does not guarantee that any specific AI platform will cite your pages for any specific query. Citation ultimately depends on relevance, authority, and each platform's own retrieval logic, which this audit can't see into. Think of this process as removing the structural excuses an AI system might have for skipping your content, not as a guarantee of the outcome.