Geo grader: how to audit your website for AI search (a GEO audit)
Updated June 11, 2026 · 9 min read
A geo grader is a tool or repeatable scorecard that audits how ready your website is to be found, understood, and cited by AI search engines like ChatGPT, Perplexity, Google AI Overviews, and Gemini. To run a GEO audit (generative engine optimization audit), inspect five things per page: whether your key content is in crawlable server-rendered HTML, whether each page leads with an extractable answer, whether you have valid structured data, whether your facts and entity are consistent, and whether AI crawlers are allowed to access you. Score each dimension, fix the lowest-scoring and highest-traffic pages first, then re-crawl to confirm the changes are visible to machines.
What is a geo grader, and what does a GEO audit check?
A geo grader scores how ready a page or site is to be retrieved, understood, and cited by generative AI search systems — ChatGPT, Perplexity, Google AI Overviews, Gemini, and Copilot. Where a classic SEO checker asks 'will this rank in the blue links?', a geo grader (or generative engine optimization audit) asks a different question: 'when an AI engine assembles an answer, can it find this page, extract a clean claim from it, and trust it enough to cite the source?'
A GEO audit is the structured process behind that score. It is a diagnostic, not a definition exercise: you crawl your real pages, inspect the machine-readable signals AI systems rely on, grade each one, and produce a prioritized fix list. This guide is the how-to for running that audit. If you need the underlying concepts first, see our explainer on what answer engine optimization is and the playbook on how to get cited by ChatGPT — this page focuses on the audit procedure itself.
What are the five dimensions a GEO audit scores?
A defensible AI search audit grades five independent dimensions per page. Each is binary enough to score, and each maps to a real reason AI engines include or skip a source:
- •Crawlability and access — can AI crawlers reach the page? Check robots.txt, meta robots, and whether you block known AI user agents (e.g. GPTBot, PerplexityBot, Google-Extended). A page you've blocked cannot be cited at all.
- •Extractable HTML — does your key content exist in the server-rendered HTML, or does it only appear after JavaScript runs? Use 'View Source' (not DevTools' rendered DOM): if your main answer isn't in the raw HTML, many crawlers and AI fetchers won't see it.
- •Answer-readiness — does the page lead with a concise, self-contained answer to the question it targets, plus question-style headings? AI engines lift short, context-independent passages; buried or hedged answers rarely get quoted.
- •Structured data and semantics — is there valid schema (FAQPage, Article, Organization, Product/Service where relevant), one H1, and clean heading order? Structured data gives engines explicit, machine-readable context.
- •Entity trust and consistency — are your facts, name, and key claims consistent across the site, and are claims specific (dates, numbers, named sources) rather than vague? Inconsistent or unverifiable claims undermine the trust signals that drive citation.
How do you run a GEO audit step by step?
Run the audit as a repeatable loop rather than a one-off scan. The goal is a scored inventory of your important pages and a ranked list of fixes.
- •1. Pick the pages that matter — your homepage, money pages, and the content pages that target real buyer questions. Don't audit the whole site at once; audit what AI is most likely to be asked about.
- •2. Test access — fetch each URL as a crawler would (curl or view-source) and confirm it returns content, isn't noindexed by accident, and isn't blocking AI user agents in robots.txt unless that's a deliberate choice.
- •3. Check extractability — view the raw HTML source and confirm your primary answer, headings, and key facts are present without JavaScript.
- •4. Grade each dimension — score the five dimensions above (a simple 0–2 or pass/partial/fail per dimension is enough) so weaknesses are comparable across pages.
- •5. Validate structured data — run your schema through a structured-data validator and confirm FAQPage/Article/Organization markup is present and error-free.
- •6. Prioritize fixes — multiply impact (page traffic and commercial value) by the size of the gap, and fix the worst-scoring high-value pages first.
- •7. Re-crawl and confirm — after changes ship, fetch the raw HTML again to verify machines actually see the new answer, headings, and schema. A fix that isn't in the server HTML didn't happen.
Which AI-search signals matter most — and how do you weight them?
Not every signal is worth equal effort. In a generative engine optimization audit, weight your fixes by how directly each signal gates citation. The table below is a practical prioritization, framed as general industry guidance rather than guaranteed outcomes.
How is a geo grader different from a classic SEO checker?
They overlap on fundamentals — crawlable content, clean structure, valid schema, and credibility help you in both classic search and AI answers — but they optimize for different end states. A classic SEO checker scores ranking factors for a list of links: titles, meta descriptions, internal links, page speed, keyword targeting. A geo grader scores citation-readiness for a synthesized answer: can an engine extract a clean, quotable claim and trust the source behind it?
The practical difference shows up in what each flags. An SEO tool may pass a page that ranks fine; a GEO audit can still flag it because its answer is buried below 600 words of preamble, its key content only loads via JavaScript, or it blocks AI crawlers. Run both — they're complementary. Use a page SEO checker for ranking hygiene and an AI-search grader for extractability and citation-readiness.
How do you fix the most common GEO audit failures?
Most pages fail on the same handful of issues. Map each finding to a concrete fix:
- •JavaScript-only content → move the primary answer and key facts into server-rendered HTML so they appear in 'View Source'.
- •No leading answer → add a 2–4 sentence, self-contained answer near the top that makes sense lifted out of context.
- •Wall-of-text structure → break content into question-style H2s with short, declarative answers under each.
- •Missing or broken schema → add valid FAQPage, Article, and Organization markup and validate it.
- •Blocked AI crawlers → review robots.txt and decide deliberately which AI user agents you allow; accidental blocks are common.
- •Vague or inconsistent claims → replace hedged statements with specific, consistent facts (dates, numbers, named sources) so engines can trust your entity.
How often should you re-run an AI search audit?
Treat the GEO audit as a recurring check, not a launch task. AI engines, their crawlers, and the schema and access conventions they respect all change, and your own content drifts as pages get edited. A practical cadence: re-audit a page whenever you substantially edit it or its schema, and run a site-level sweep of your most important pages roughly quarterly — or sooner if you notice your pages stop appearing in AI answers.
The fastest way to keep score is to grade AI-search readiness automatically. Our AI SEO page grader checks the same extractability, structured-data, and answer-readiness signals this audit covers, so you can re-run it as often as you ship changes.
AI SEO Page Grader (AEO / GEO)
Run a free AI-search audit — get your Revenue Grade and the specific fixes in seconds.
Frequently asked questions
- What is a geo grader?
- A geo grader is a tool or scorecard that audits how ready your website is to be found, understood, and cited by AI search engines like ChatGPT, Perplexity, and Google AI Overviews. It scores generative engine optimization (GEO) signals — crawler access, extractable HTML, answer-readiness, structured data, and entity consistency — rather than classic ranking factors alone.
- What's the difference between a GEO audit and an AEO audit?
- In practice the terms overlap heavily and many tools use them interchangeably. Generative engine optimization (GEO) emphasizes being cited inside AI-generated answers, while answer engine optimization (AEO) emphasizes directly answering questions in an extractable way. A combined AI search audit covers both: extractability, structured data, answer-readiness, crawl access, and trust.
- Can a geo grader guarantee my page appears in AI answers?
- No. No tool or audit can guarantee inclusion or citation in AI systems — outcomes depend on the engines, your broader authority, and competition. A geo grader scores readiness: the on-page signals that make citation more likely, and the gaps to fix first.
- Do I still need a regular SEO checker if I run a GEO audit?
- Yes — they're complementary. A page SEO checker covers ranking hygiene like titles, meta data, internal links, and speed; a geo grader covers extractability and citation-readiness for AI answers. The fundamentals overlap, but each flags issues the other can miss, so run both on your important pages.
- How do I check whether AI engines can actually read my page?
- View the page's raw HTML source (not the rendered DOM) and confirm your main answer, headings, and key facts are present without JavaScript. Then check robots.txt for blocks on AI user agents, and validate your structured data. A geo grader automates these checks across pages.
- How often should I run an AI search audit?
- Re-audit a page whenever you substantially change its content or schema, and run a site-level sweep of your most important pages roughly quarterly. AI engines and their crawling conventions evolve, so periodic re-grading keeps your pages citation-ready.