Definition
A GEO audit is a structured review of how a brand appears in AI-generated answers and why. It establishes a baseline of mentions and citations across engines such as ChatGPT, Google AI Overviews and AI Mode, Perplexity, Gemini, Claude and Copilot, finds the technical, content and reputation causes of gaps, and ends in a ranked list of fixes. It is the GEO counterpart of a technical SEO audit, and it usually includes some of the same checks.
A complete audit covers five areas:
- Visibility: a fixed panel of realistic prompts, run several times on each engine, scored for mention rate, position, AI share of voice against competitors, sentiment and citation share.
- Sources: which domains and URLs each engine cites for those prompts, split into your own pages, competitors' pages and third-party sites such as reviews, forums and media.
- Access: robots.txt rules and CDN or firewall responses for search crawlers such as OAI-SearchBot, Claude-SearchBot, Googlebot and Bingbot, and the separate decision on training crawlers.
- Content: whether pages answer the query fan-out sub-queries behind priority prompts, render key content as HTML, and state facts clearly with dates and sources.
- Entity: whether the brand's name, category, products and key facts are consistent across its own site and third-party profiles, and whether AI answers get them right.
First-party data has improved. Google Search Console added a generative AI performance report in June 2026 and rolled it out to all sites on August 31, 2026; it shows impressions in AI Overviews and AI Mode by page, country and device, without clicks or queries. Google's guide to generative AI features also notes that a site must be included in Search generative AI features in Search Console to be eligible, so an audit should confirm that setting has not been switched off. Bing Webmaster Tools' AI Performance report, in public preview since February 2026, adds citation counts and sample grounding queries for Copilot and Bing's AI summaries. Server-side, AI crawler logs show which bots fetch which pages and which requests fail. None of these sources covers competitors or engines like ChatGPT and Perplexity, which is why audits also rely on prompt tracking.
The output should be a short, ranked plan rather than a long list of findings. Typical high-priority items are crawler blocks, important pages that are not retrieved for their fan-out queries, outdated facts repeated in AI answers, and topics where third-party sources cite competitors but not you. Each item should name an owner and the metric that will show it worked.
Re-run the audit quarterly and after major changes on the engine side, such as the switch to Gemini 3.5 Flash as AI Mode's default model in May 2026 or the GPT-6 rollout in ChatGPT in October 2026, because answers can shift without any change on your site. The free AI SEO audit and the AI visibility report are quick starting points before a full audit.
Examples of GEO Audit
- An audit finds a B2B brand mentioned in 41% of tracked ChatGPT answers but cited in only 9%, with most citations going to two review sites. The plan pairs new comparison pages on the brand's own site with a review collection campaign.
- A publisher's audit shows no data at all in Search Console's generative AI report. The cause turns out to be a previous agency having excluded the property from Search generative AI features.
- Crawler logs reviewed during an audit show Claude-SearchBot receiving 429 rate-limit responses on the docs subdomain, which explains why Claude never cites the documentation.
- After GPT-6 reaches all ChatGPT users in October 2026, a team re-runs its audit prompt set to set a new baseline before planning next quarter's content.
