Every agency now sells an AI audit.
Most are guessing.
We built the instrument.

A GEO audit (generative engine optimization audit) measures whether AI engines — ChatGPT, Perplexity, Google AI Overviews, Gemini, and Claude — name and cite your business when buyers ask category-relevant questions. Where a traditional SEO audit checks whether you rank on a results page, a GEO audit checks whether you appear inside the answer itself. It baselines your visibility across every major engine, benchmarks it against your top three competitors, diagnoses why you're being left out, and hands you a prioritised list of fixes. The whole thing runs on MentionLayer, the AI-visibility SaaS we built — so it's measured data, not a subjective read of your site.

Nothing meaningful — they're the same deliverable under two names. 'GEO audit' is the term used by people who already think in generative-engine-optimization language; 'AI visibility audit' is the plainer phrasing for the same question: when a buyer asks an AI assistant about my category, do I show up? Some people also call it an 'AI SEO audit.' We use the terms interchangeably. What matters is the method underneath: real queries run through real engines, competitor benchmarking, and a diagnosis tied to the specific layer that's failing — entity, technical, authority, or content.

Four categories. AI answer visibility: your priority queries run through ChatGPT, Perplexity, Gemini, Google AI Overviews and Claude, recording for each one whether you were named, cited with a link, or absent — plus your share of model against your top three competitors. Entity and authority: whether the engines can actually identify who you are, your knowledge-graph and sameAs footprint, and your third-party citation profile. Technical AI-readiness: schema @graph coverage, FAQ and Article markup, your llms.txt and ai.txt files, and whether you're accidentally blocking GPTBot, PerplexityBot, or Google-Extended. Content extractability: whether your pages are written in the direct-answer, question-heading, citation-friendly patterns that LLMs actually pull from. Every finding is tagged with impact and effort.

An SEO audit asks 'do I rank on Google's results page.' A GEO audit asks 'does the AI answer name me at all.' The overlap is real — strong SEO foundations feed AI visibility — but the surfaces are different. AI engines choose sources through entity disambiguation, schema graphs, citation authority, and content extractability, not just links and rankings. A traditional audit won't tell you that ChatGPT names your competitor three times and never mentions you, or that PerplexityBot is blocked by your robots.txt. If you want the ranking side too, our free SEO audit covers technical, content, and authority; this one is purpose-built for the AI-answer layer.

The baseline GEO audit is complimentary — no payment required. The honest framing: it's a lead magnet. Running your queries through five engines and benchmarking three competitors gives us a chance to show how we work, and gives you a real diagnosis whether or not you ever hire us. Some recipients take the findings and fix them in-house; some book an engagement; some do nothing. We're fine with all three. There's no automated drip sequence chasing you afterward. Larger organisations with hundreds of priority queries or multi-brand footprints move into a paid scope, because the analysis is genuinely bigger — but the standard audit covers the sweet spot at no cost.

Through MentionLayer, the SaaS Joel built specifically for this. For each of your priority keywords we query ChatGPT, Perplexity, Gemini, Claude, and Google AI Overviews, then record five things per answer: were you mentioned, were you cited with a link, what was the citation context, what was the sentiment, and which competitor was named instead. We aggregate that into a share-of-model score — your percentage of AI answers in your category — and track it against your top three rivals. This is the same instrument that generated our Q1 2026 AI Visibility Index across 1,004 businesses, so the method is the one that produced published industry research, not a one-off spreadsheet.

Five to seven business days from the point we have your priority queries and competitor list. Most of that window is interpretation, not data collection — the engine queries run quickly, but the part that matters is diagnosing why you're invisible and sequencing the fixes so you work on the highest-impact items first. Joel reviews every audit personally and records a short walkthrough of the top findings. You could deliver a raw visibility scan in an afternoon; you can't deliver a prioritised plan that fast, which is the difference between a tool export and an audit.

Because we built the instrumentation instead of guessing with it. Xpand operates MentionLayer, the AI-visibility platform that runs the audit and produced our Q1 2026 AI Visibility Index — 95,392 data points across 1,004 businesses, which found 65.9% invisible in AI search. We also run PressForge, the digital-PR engine behind 300+ campaigns that earns the third-party citations LLMs learn from, and Joel House wrote AI for Revenue, published on Barnes & Noble. Most agencies now selling 'AI visibility' are reading a third-party tool's dashboard. We wrote the tool, ran the study, and published the book — so the audit is diagnosis from the people building the category, not a template with your logo dropped in.