The AEO Audit Playbook: Grade Your Site in 30 Minutes

Most site owners who complain they're invisible in ChatGPT have never actually checked why. They assume it's a content problem, or a mystery, or that AI search just doesn't like them. Usually it's something far more boring and fixable, and you can find it yourself in half an hour.

This is a full AEO audit you can run with nothing but a browser. No tools, no login, no consultant. You'll grade your site across the six things that actually decide whether AI engines can crawl, understand, and cite you, and you'll walk away with a readiness score and a prioritized fix list. Founders keep asking on Reddit how to check their AEO and GEO visibility for free. Here's the free way.

Why bother? Because AI answers increasingly resolve without a click, and if you're not the source being quoted, you're simply not in the conversation. The frustrating truth from teams who audit sites for a living is that the blockers are usually mundane and self-inflicted: a crawler quietly disallowed, a page with no schema, content trapped behind JavaScript. You can find and fix those in an afternoon once you know where to look.

How the Scorecard Works

Six categories, 25 points total, roughly 30 minutes. Run each checklist below, give yourself a point for every item you pass, and add it up. Grade honestly, because the only person you fool by inflating the score is yourself.

Here's how to read the final number:

One rule before you start: the categories aren't equal. A zero in the first one caps everything else, so start at the top and don't skip.

1. Crawler Access (0-3 points)

This is the gate. If the AI bots can't reach your pages, nothing else on this list matters. As one practitioner bluntly put it after auditing dozens of sites, if the model can't extract it, you don't exist.

• Open robots.txt. Visit yoursite.com/robots.txt and confirm you're not disallowing GPTBot, OAI-SearchBot, ClaudeBot, PerplexityBot, or Google-Extended, and that there's no blanket Disallow: /. (1 point)
• Check for silent blocks. Even with an open robots.txt, a CDN or firewall can quietly reject bots. Skim your server logs for actual hits from those AI crawlers. If they've never visited, something upstream is blocking them. (1 point)
• Confirm llms.txt exists. Check yoursite.com/llms.txt for a real file with your description and key links, not a placeholder. (1 point)

If you scored zero here, stop and fix it before anything else. Our guide on why AI search ignores your content walks through the exact checks.

RANKCONTROL

15+ content types. Published on your domain. Matched to your brand.

Guides, comparisons, listicles, case studies, and more. RankControl generates content that gets cited by ChatGPT, Claude, Perplexity, and more.

Start free→

2. Content Structure and Extractability (0-5 points)

A model has to be able to lift a clean, quotable chunk from your page. Dense walls of text and content buried in JavaScript are where citations go to die. On Reddit, one auditor summed up the pattern perfectly: accurate structure beats fluff, and JS-hidden content silently kills you.

• Answer-first sections. Each major H2 opens with a direct 40 to 60 word answer before the detail. (1 point)
• Logical, descriptive headings. One H1, clean hierarchy, and headings that state a topic or question instead of "Section 3." (1 point)
• FAQ blocks on key landing pages, one clear question per heading. (1 point)
• Tables and lists for data and steps, not paragraphs pretending to be tables. (1 point)
• No hidden core content. Your main answers are visible in the HTML, not locked inside "click to expand" accordions or rendered only by JavaScript. (1 point)

3. Structured Data (0-5 points)

Schema is how you hand a machine your facts in a format it can't misread. It's close to table stakes now, and pages with it get cited noticeably more often.

• Organization schema site-wide with name, logo, and sameAs links. (1 point)
• Article or WebPage schema on content, including author and datePublished and dateModified. (1 point)
• FAQ, HowTo, or Product schema wherever that content type appears. (1 point)
• Breadcrumb schema where you show breadcrumbs. (1 point)
• Validation. Run two or three key pages through Google's Rich Results Test with no critical errors. (1 point)

Quick way to check without any tools: open a page, view source, and search the HTML for application/ld+json. If you find nothing, you have no schema, and that's an easy point to lose. Our structured data blueprint has the exact markup if you're starting from scratch.

4. Entity Consistency (0-4 points)

AI engines have to be sure all your scattered mentions describe one real business. Small mismatches make a model hedge and reach for a cleaner-looking competitor instead.

• Consistent name and NAP across your header, footer, contact page, and schema. (1 point)
• sameAs populated with your real profiles: LinkedIn, Crunchbase, G2, and the like. (1 point)
• Cross-profile match. Your site, LinkedIn, and one major directory tell the same story with the same name and positioning. (1 point)
• Accuracy check. Ask ChatGPT and Perplexity to describe your company. If the basics are wrong, your entity is muddy. (1 point)

Your competitors are getting cited by AI. You're not.

Every day without citation tracking is a day your competitors pull ahead in ChatGPT, Perplexity, and Claude.

See what you're missing→

5. Authority and E-E-A-T (0-5 points)

A model won't confidently recommend a source it can't vouch for. Trust signals are what separate a page that gets cited from one that gets skipped, and a lot of this lives off your own site. As a commenter on one GEO audit thread noted, even a technically perfect site loses if nobody off-site is talking about it.

• Author bylines and bio pages with credentials, ideally with Person schema. (1 point)
• About, contact, and legal pages, plus HTTPS across the whole site. (1 point)
• External citations. Key pages link out to credible sources. (1 point)
• First-hand experience. Real examples, specific numbers, actual case studies, not generic advice. (1 point)
• Third-party presence. Three to five independent sites (directories, reviews, press, communities) mention you. (1 point)

If this is your weak spot, Reddit and community mentions are one of the highest-impact places to start.

6. Freshness (0-3 points)

AI answers lean toward recent content, and a visible date is a trust cue. Stale pages quietly fall out of the rotation.

• Visible dates. Your key pages show a published or last-updated date. (1 point)
• Recent updates. Strategic pages have had a substantive refresh in the last 6 to 12 months, with real new material rather than a bumped date. (1 point)
• Dates in schema. Your datePublished and dateModified fields are present and accurate. (1 point)

The Three Failures That Sink Most Sites

If you're short on time, these are the three that show up again and again in real audits, and any one of them can drag an otherwise solid site to the bottom.

The first is a blocked crawler. A robots.txt rule or a firewall setting, often added years ago during a scraping scare, quietly shuts out the exact bots that build AI answers. The site owner has no idea, because nothing looks broken. The second is missing schema. The content is fine, but with no structured data the models have to guess at your facts, and they'd rather cite a page that spells them out. The third is anonymous, stale content: no author, no dates, no signal that a real, current expert stands behind it. Each of these is a fast fix, and each is worth more than a month of writing new posts on top of a broken foundation.

Read Your Score, Then Fix in Order

Add up your six categories. Wherever you landed, the fix order is the same, because the categories build on each other. Crawler access first, always, since a failure there makes everything below invisible. Then structure and schema, so the pages you've unblocked can actually be parsed. Then authority and freshness, which decide whether you get chosen once you're in the running.

Here's the honest limitation of a manual audit: it's a photograph, and AI search is a moving target. You can score a 23 today and slip to 17 by spring when a competitor cleans up their schema, an engine changes how it weights sources, or a page you forgot about goes stale. Running this by hand every quarter across every engine adds up to real hours. That's the gap we built RankControl to close: it runs this audit continuously, tracks your score as it moves, and tells you which fix will actually shift your citations. You can absolutely grade yourself with the checklist above. The tool just keeps grading you after you've stopped looking.

Run the audit today, fix your lowest category first, and re-check in 90 days. Most sites that felt invisible discover they were never actually in the game, one blocked crawler or one missing schema block away from answers they could have been winning all along.

RANKCONTROL

200+ SaaS teams already track their AI citations.

They know exactly when ChatGPT mentions their brand, and when it stops. Do you?

See plans→