An AEO audit is a structured review of whether AI answer engines, ChatGPT, Perplexity, Google's AI Overviews, Claude, are citing your site when they answer questions your buyers are asking. It checks crawlability for AI bots, passage-level structure, citation presence across a set of target queries, and the gap between where you rank on a traditional search engine results page and where you actually show up in an AI-generated summary. Those two things are no longer the same, and most businesses find that out the hard way: ranking on page one of Google while being invisible in the answer their prospect actually reads.
We built a free version of this audit because the question comes up in nearly every discovery call now. Marketing directors want to know if their content strategy still works when a growing share of queries never produce a click.
What does an AEO audit actually check?
An AEO audit checks four things: crawlability, passage structure, citation rate, and competitive gap. Crawlability means confirming that AI crawlers (GPTBot, PerplexityBot, Google-Extended, ClaudeBot) can actually reach and parse your pages, not blocked by robots.txt, not buried behind JavaScript rendering the crawler never executes. Passage structure means checking whether your content answers questions in the first 100 to 150 words, with a direct sentence an extraction model can lift cleanly, instead of three paragraphs of preamble before the actual answer. Citation rate means running your target queries against the major answer engines and logging whether your domain shows up as a source. Competitive gap means doing the same thing for the three or four competitors who do show up, so you can see what they're doing differently.
This is mechanically different from a standard SEO audit. SEO audits look at backlinks, domain authority, keyword density, and SERP position. An AEO audit looks at whether a language model, synthesizing an answer from crawled passages, chose your sentence over someone else's. Authority signals still matter, but they're not the deciding factor the way they are in classic ranking algorithms.
Why is my site ranking on Google but invisible in ChatGPT?
Because ranking algorithms and summarization models score content on different criteria, and a page built for one often fails the other. Google's classic ranking weighs domain authority, backlink profile, freshness, and click-through history. An AI summary model weighs whether it can extract a complete, factually clean answer in the first few sentences of a passage. A page can carry real authority and strong backlinks and still bury its answer under a 400-word introduction, a format that works fine for ranking and gets skipped entirely by extraction.
We cover the full mechanics of this split, including why a page one ranking stops guaranteeing visibility as AI Overviews expand, in our piece on AEO versus SEO as Google Summarize scales. The short version: you need both disciplines running at once, because the SERP and the AI summary are drawing from the same crawl but scoring it differently.
How often should I run an AEO audit?
Quarterly, at minimum, because answer engines update their citation behavior faster than traditional search ranking shifts. A page that gets cited by Perplexity in January can drop out by April if a competitor restructures their content or a model update changes extraction preferences. Businesses with active content programs, publishing weekly or faster, should check citation rate monthly on their highest-intent queries. Everyone else should treat it like a technical SEO crawl: scheduled, not reactive.
The audit itself doesn't take long to run once you have the query list built. What takes time is the fix: restructuring passages, adding direct-answer sections, fixing crawl blocks, and rebuilding pages that were written for human skimming instead of machine extraction. That's the work, not the audit.
What do I do with the results of an AEO audit?
You prioritize fixes by query volume and buyer intent, starting with the pages closest to a purchase decision, not the pages with the most traffic. A blog post that gets 5,000 visits a month but sits at the top of the funnel matters less than a comparison page or pricing explainer that an AI engine could be citing, or should be citing, right before someone becomes a lead. We rank findings into three buckets: crawl and technical blocks (fix immediately, these are usually one-line changes), passage restructuring (fix within the sprint), and content gaps where no page on your site answers the query at all (plan for the next content cycle).
We run this same discipline across the verticals we work in, car wash and multi-location franchise marketing, B2B SaaS, ecommerce, home services, because the mechanism doesn't change by industry. What changes is which queries matter. A franchise group cares about location-specific service queries. A SaaS company cares about comparison and integration questions. The audit structure stays the same; the query list doesn't.
Does fixing AEO hurt my paid search or Amazon performance?
No, AEO and paid channels are independent systems, but the audit discipline of checking what's actually getting cited, clicked, or bought instead of assuming it transfers directly. If you run Amazon as a channel, the same instinct, verify instead of assume, applies to ad spend, not just organic citations. We cover the four metrics that matter in an Amazon context in our Amazon Ads audit checklist. The common thread across both audits: don't trust the channel's own dashboard as the full picture. Check what's actually happening against a second source.
Getting started
We run a free AEO audit that checks crawlability, passage structure, and citation rate against a query set built from your actual buyer language, then hands back a prioritized list of fixes, not a vendor pitch. If you want the fuller toolkit, including the ongoing monitoring layer, our AEO Toolkit is built to get sites cited by ChatGPT, Claude, and Perplexity on a continuing basis, not just a one-time check.
An AEO audit won't tell you to abandon SEO, and it won't hand you a vanity score. It tells you, query by query, whether the answer engines your buyers are using right now can find you, cite you, and send them your way, and if they can't, exactly which sentence, page, or crawl rule is stopping them.
