What Is a GEO Audit? What It Checks and How to Run One

A GEO audit is a structured check of whether AI answer engines — ChatGPT, Perplexity, Gemini, and Google's AI features — name and cite your brand when buyers ask questions in your category, plus a diagnosis of what's blocking you when they don't. It has two halves: measurement (run real buying questions through the engines and record who gets named) and inspection (examine your site and your evidence trail for the signals engines use to decide whom to trust). The deliverable that matters is not a score. It's a short, ranked list of fixes, ordered by what's costing you the most answers.

TL;DR

• A GEO audit measures whether you appear in AI answers and diagnoses why you don't — the AI-era counterpart to an SEO audit.
• It checks both sides of the problem: what the engines say about you today (presence, share of the answer, accuracy) and the machine-facing signals behind it (crawler access, entity clarity, answer-shaped content, third-party evidence).
• A good report is a ranked fix list tied to specific gaps, not a single vanity score.
• You can run a rough version by hand in an afternoon — the five-minute manual check is the fastest starting point.
• A GEO audit and an AEO audit are the same deliverable under two names; don't pay for both.

What is a GEO audit?

A GEO audit is the AI-answer equivalent of an SEO audit. An SEO audit tells you why you're not ranking in search results; a GEO audit tells you why you're not showing up in AI answers. GEO stands for Generative Engine Optimization — the practice of getting your content retrieved, trusted, and cited when an AI engine composes a response — and if the discipline itself is new to you, start with What is GEO? before auditing your position in it.

The audit exists because answer-side losses are invisible by default. A buyer can ask an assistant "which tool should I buy for X," get a confident answer naming three competitors, and never visit a results page you could have ranked on. There's no click to lose, so nothing shows up in your analytics — conversion from search just quietly erodes. A GEO audit is the act of going and looking: asking the questions your buyers ask, reading the answers honestly, and tracing each absence back to a fixable cause.

You'll also see the same deliverable sold as an "AEO audit" or an "AI visibility audit." The names differ; the checklist doesn't. Treat them as one thing throughout this post.

What does a GEO audit check?

A thorough GEO audit checks what the engines say about you, and then why they say it. On the measurement side:

• AI presence. For a set of real buying questions — category questions, not your brand name — do the engines mention you at all, and on which platforms? Being strong in ChatGPT and absent from Perplexity is a common and diagnosable pattern.
• Share of the answer. When engines do name brands, how often is it you versus a competitor? Answers name a handful of options and everything else is absent — there's no long tail to hide in.
• Accuracy and sentiment. When you are mentioned, is the description current and correct, and is the framing working for you or against you? A stale price or a misstatement of what you sell is a live gap even when you're technically "visible."

On the inspection side, the audit looks for the blockers behind those results:

• Crawler access. Are AI crawlers allowed to read your site, or is a robots rule quietly shutting them out? A blocked crawler eliminates you before any judgment of quality happens.
• Entity clarity and structured data. Is your name, category, and description consistent everywhere you appear, and do your pages carry schema that tells a machine unambiguously what you are? An engine that isn't sure who you are doesn't hedge — it omits you.
• Answer-ready content. Do you actually answer the questions buyers ask, in a direct, liftable form — or would a model have to excavate your point from a brand story?
• Third-party evidence. Do independent reviews, comparisons, and community mentions exist that an engine can treat as corroboration? Engines discount what you say about yourself, so this is often the binding constraint — and it's the check most "site audit" tools skip entirely.

What should a GEO audit report contain?

A useful GEO audit report contains five things, and the last one is the whole point:

1. The question set. The actual prompts tested, so you can re-run them and so nobody is grading against softball queries.
2. A per-engine results matrix. For each question and each engine: named with a reason, named without one, misdescribed, or absent. Dated — answers drift as models update and re-crawl.
3. A competitor comparison. Who is winning the answers you're losing, and what evidence the engines appear to be leaning on when they pick them.
4. A diagnosis per gap. Not "your AI visibility is weak" but "Perplexity has no recent third-party source to cite you from" or "your robots file blocks the crawler this engine uses." Specific enough to act on.
5. A ranked fix list. The handful of changes that unblock the most answers, in order, with the reasoning shown.

What a good report does not lead with: a single composite score with no diagnosis behind it, undated screenshots, or a guarantee of future placement. Nobody can promise what a generative engine will say next month — an audit's honest job is to show you where you stand and which lever to pull first.

How do you run a GEO audit yourself?

You can run a credible manual GEO audit in an afternoon with nothing but browser tabs and a spreadsheet:

1. Build the question list. Ten to twenty questions a buyer would ask right before spending money — "best X for Y," "X vs Z," "is X worth it." Category questions, never your own brand name.
2. Run them across engines. ChatGPT, Perplexity, Gemini, and Google's AI features at minimum. For each answer, record the outcome: named with a reason, named without one, misdescribed, or absent — and note which sources the engines that show citations are actually citing.
3. Inspect the blockers. Check your robots rules for AI crawlers, check that your key pages carry structured data and answer the listed questions directly, and search for the third-party sources an engine could cite you from. Absences here usually explain absences above.
4. Rank, fix one thing, re-run. Pick the gap that blocks the most answers, fix it, and re-test on a schedule — a single snapshot ages fast, and the re-run is what tells you whether the engines responded.

If you only have five minutes today, do the compressed version first: our AI visibility audit walks through the three-tab check and how to read what comes back. The full audit above is that check, made systematic and pointed at causes.

The honest caveat about manual audits: they're tedious, they capture one moment in time, and answers vary between sessions. That's tolerable for a first diagnosis and painful as a monthly practice — which is why the measurement half of this work tends to end up automated.

How is a GEO audit different from an SEO audit?

They overlap at the foundation and diverge at the scoreboard. Both care about crawlability, site structure, and content quality — several engines retrieve through search indexes, so SEO health is partly the substrate AI visibility sits on. But an SEO audit grades you on positions in a list, while a GEO audit grades you on inclusion in an answer; an SEO audit lives almost entirely on your own site, while a GEO audit spends much of its time on the third-party evidence engines use to corroborate you; and SEO tooling can watch rankings continuously, while answer inclusion has to be sampled by asking. A clean SEO audit tells you nothing about whether ChatGPT names you — that's precisely the blind spot a GEO audit exists to cover.

Should you pay for a GEO audit?

Sometimes — but know what you're buying. The measurement half of a GEO audit is software-shaped work: running questions across engines and recording outcomes is exactly what tools automate well, and paying agency rates for it is paying humans to do a robot's job. The judgment half — deciding which fixes matter for your business, and doing the reputation work of earning third-party evidence — is where human help can genuinely earn its fee. Our breakdown of answer engine optimization services covers what's really in those retainers and the red flags to catch before signing one.

One red flag worth repeating here: a vendor selling you a "GEO audit" and an "AEO audit" as separate line items is selling the same checklist twice.

The cheapest honest starting point is a free automated scan. AnswerX runs your category's buying questions across the major engines, records who gets named, checks the machine-facing blockers on your site, and returns the ranked fix list — the same deliverable this post describes, without the spreadsheet afternoon.