The AEO Audit Tool Checklist: What a Real Site Audit Has to Include

What separates an AEO audit tool from a score generator

There are a lot of tools now that will give you a number and call it your "AI visibility score." Fewer of them will tell you why the number is what it is, and almost none will hand you something you can paste into your site today. If you're evaluating options — for your own site or a client's — here's what a real AEO audit tool needs to check, and how to tell a diagnostic tool from a scoreboard.

The six modules a real audit covers

Structured data. This is schema markup — LocalBusiness, FAQPage, AggregateRating — the JSON-LD blocks that tell an AI engine unambiguously what your business is, what it does, and where it operates. AI models don't have to guess at meaning when it's already labeled. A site with no structured data is asking the model to infer facts from prose; a site with complete schema is handing those facts over directly.

Content structure. This checks whether your content is shaped the way AI engines actually consume it: question-format headings (H2s phrased as questions), a dedicated FAQ section, sufficient word count to establish topical depth, and use of lists. AI answer engines tend to lift structured, question-and-answer content more readily than dense unstructured paragraphs, because it maps directly onto the question a user asked.

Entity and trust. This is about whether your business reads as a real, verifiable entity rather than an anonymous website. NAP consistency (name, address, phone matching across the site), a named owner or credentials, sameAs links connecting to other verified profiles (LinkedIn, Google Business Profile, industry directories), and an About page. AI engines weigh entity signals when deciding whether a source is trustworthy enough to cite.

AI crawler access. This one is binary and easy to get wrong by accident: does your robots.txt actually allow the bots that matter — GPTBot, ClaudeBot, PerplexityBot, OAI-SearchBot, Bingbot — and is there no stray noindex meta tag blocking indexing? A site can have perfect content and schema and still be invisible to AI engines because a crawler directive silently blocks them.

Technical health. HTTPS, a working sitemap.xml, a canonical tag, a mobile viewport meta tag, and visible dates on content (via <time> tags or article metadata). None of these are exotic. All of them are foundational, and audits that skip straight to "AI visibility" advice without checking them are skipping the floor the house is built on.

llms.txt. A newer, AI-specific standard: a plain-text file at your site root that gives AI engines a structured summary of who you are and what's on your site — an H1, a blockquote summary, and linked sections. It's the AI-era equivalent of robots.txt or sitemap.xml, and most sites don't have one yet.

Measure vs. fix — the distinction that matters

A lot of tools will tell you a number on each of the above. Far fewer will tell you what to change. That's the real dividing line to evaluate a tool by: does it just measure where you stand, or does it hand you the fix?

A tool that only measures gives you a report card. Useful for knowing you have a problem, useless for knowing what to do about it. A tool built to fix, not just score, tells you specifically: this JSON-LD block is missing this required field, this heading should be rephrased as a question, this line in robots.txt is blocking GPTBot. That's the difference between a diagnosis and a checkup that just says "something's wrong in there somewhere."

The "can I act on it today?" test

Here's the test we'd apply to any AEO audit tool, including our own: after you get the report, can you hand it to a web developer this afternoon and have them make changes without further research? Three things determine whether the answer is yes:

1. A ready-to-publish llms.txt file — pre-filled with your business name, description, services, and service area, not a blank template you have to figure out yourself.
2. A ready-to-paste identity schema (LocalBusiness, Person, Podcast, or whatever fits your site type) — complete JSON-LD with your actual business data already in it.
3. A ready-to-paste FAQPage JSON-LD schema — 3 to 5 question-and-answer pairs built from your own content or your business type, formatted and ready to drop into your site's <head>.

If a report gives you scores but none of these three things, you're getting a diagnosis with no prescription. If it gives you all three, plus specific, plain-English fix instructions for your weakest areas, you've got something a developer can act on the same day.

What this actually costs

AEO Checker runs all six modules above and produces all three deliverables, as a one-time report you can hand a client — no subscription, no account required. Quick Scan is $19.95 and covers up to 20 pages. Full Audit is $39.95 and covers up to 50 pages. Deep Scan is $69.95 and covers up to 100 pages, plus a per-page action checklist. All three tiers include the full module breakdown, the letter grade, the top 3 priority fixes, and all three paste-ready assets. Every report comes with a 30-day money-back guarantee, and delivery lands in your inbox in under 5 minutes.

If you want to see where a site stands before deciding which tier to buy, the free Grade Check below runs the core scan and gives you the grade and module breakdown at no cost.