How to Run a GEO Audit

A GEO audit is the diagnostic foundation of any AI search visibility strategy. It tells you exactly where your brand appears in AI-generated answers, where it does not, and why. Unlike an SEO audit — which measures rankings and traffic — a GEO audit measures citations and mentions across ChatGPT, Perplexity, Google AI Overviews, and Gemini. This guide walks through a complete audit process you can run yourself.

Step 1: Define your target query set

Your audit is only as good as the queries you test. Build a list of 20-50 queries that represent the moments where your brand should be visible in AI answers. Break them into four categories:

• Brand queries — "what is [your brand]", "is [your brand] legit", "[your brand] reviews"
• Category queries — "best [your service]", "top [your service] tools", "[your service] alternatives"
• Comparison queries — "[your brand] vs [competitor]", "[competitor A] vs [competitor B]"
• Informational queries — "how does [your service] work", "what is [your industry term]"

Source these from:

• Google Search Console (queries that already drive traffic)
• People Also Ask boxes for your target keywords
• Customer support tickets and sales calls (what people ask before buying)
• Competitor pages and their target keywords

Record each query in a spreadsheet. This is your benchmark — every audit step references this list.

Step 2: Test each query across four AI engines

For each query, run it through four AI engines and record the result:

Record results in four categories:

• Cited with link — your brand is mentioned and linked to
• Mentioned without link — your brand is named but no source link
• Absent — your brand does not appear
• Competitor cited — a competitor is cited instead

This is your citation baseline. Here is a sample tracking sheet:

Query                          | ChatGPT | Perplexity | AI Overviews | Gemini
"best SEO consultant London"   | Absent  | Competitor  | Competitor    | Absent
"what is GEO"                  | Absent  | Absent      | Absent        | Absent
"Izajul Haque reviews"         | Cited  | Cited       | Mentioned     | Mentioned

The baseline tells you where you stand. Most brands discover they are absent from 60-80% of queries where they should be visible. See how to track AI citations for ongoing monitoring.

Step 3: Audit entity clarity and schema

AI engines cite entities they can identify and verify. If your brand is not a clearly defined entity, AI models will not reference you. Check:

• Organization schema on homepage — Does your homepage have Organization JSON-LD with name, url, logo, and sameAs links? See how to add Organization schema.
• Person schema on author pages — Do your author bio pages have Person schema with name, jobTitle, and sameAs links?
• Article schema on blog posts — Do your blog posts have BlogPosting schema with headline, author, datePublished, and dateModified? See how to implement Article schema.
• Knowledge Panel — Search your brand name on Google. Is there a Knowledge Panel? If not, your entity is not well-established in Google's Knowledge Graph.
• sameAs links — Are your LinkedIn, Twitter, YouTube, and other authoritative profiles linked in your Organization schema? Broken or missing sameAs links weaken entity identity.

Use the Meta Title & Description Checker and HTML Headings Checker to verify on-page elements that support entity clarity.

Step 4: Audit content extractability

AI engines extract content that is structured, concise, and factual. Review your top 10-20 pages for:

• Question-answer patterns — Does the page have clear questions as headings with concise answers below? AI engines extract these directly.
• Concise definitions — Are key terms defined in 40-60 word blocks? This is the sweet spot for extraction.
• Structured headings — Are H2 and H3 headings descriptive and hierarchical? AI models use headings to understand content structure.
• FAQ schema — Do Q&A pages have FAQPage schema? See how to write FAQ schema.
• HowTo schema — Do tutorial pages have HowTo schema? See how to set up HowTo schema.
• Tables and lists — Does the page use structured formats (tables, numbered lists, bullet lists) that AI engines can extract?

Content that is purely narrative — long paragraphs without headings, no question-answer patterns, no structured data — is hard for AI engines to extract and cite. See how to structure a page for Google and AI answers for the full framework.

Step 5: Audit technical crawlability for AI bots

AI engines need to crawl your content to cite it. Check:

• robots.txt — Are you blocking AI crawlers? Check for User-agent: GPTBot, User-agent: PerplexityBot, and User-agent: Google-Extended with Disallow: /. If you are blocking these bots, AI engines cannot access your content. Use the robots.txt checker to verify.
• Server-side rendering — View your page source (Ctrl+U). Is the content in the initial HTML, or is it loaded by JavaScript? AI crawlers are less likely to execute JavaScript than Googlebot. If your content is JS-rendered, AI engines may see a blank page.
• Page speed — AI crawlers have crawl budgets too. Slow pages may be abandoned. Check Core Web Vitals (LCP under 2.5s, INP under 200ms, CLS under 0.1). See the technical SEO checklist.
• Canonical tags — Are canonical tags correctly set? Duplicate or conflicting canonicals can prevent AI engines from identifying your canonical content. Use the Canonical Tag Checker.

Step 6: Run competitor share-of-voice analysis

For each query where your brand is absent, record which competitors are cited. Look for patterns:

• Schema advantage — Do cited competitors have stronger Organization, Article, or FAQ schema?
• Content advantage — Do cited competitors have more concise, structured, factual content?
• Entity advantage — Do cited competitors have stronger Knowledge Graph presence (Wikipedia pages, more sameAs links, Knowledge Panels)?
• Authority advantage — Do cited competitors have more backlinks, domain authority, or media mentions?

This analysis reveals your competitive gaps. If competitors are consistently cited for category queries and you are not, the gap is likely in entity clarity or content extractability — not in traditional SEO metrics. See SEO vs AEO vs GEO for how the disciplines differ.

Step 7: Produce a prioritized action list

Compile all findings into a prioritized action list:

1. Entity gaps (fix first) — Missing Organization schema, broken sameAs links, no Knowledge Panel, no Person schema for authors
2. Content gaps (fix second) — Queries where you have no extractable answer, pages missing FAQ/HowTo schema, narrative-heavy content that needs restructuring
3. Technical gaps (fix third) — AI bot blocks in robots.txt, JS-rendered content, slow pages, missing canonical tags
4. Tracking gaps (set up ongoing) — No citation monitoring, no competitor tracking, no quarterly audit cadence

Prioritize by query volume and business impact. A brand query where you are absent is more urgent than a long-tail informational query. See GEO consulting for help executing the action list, and the glossary for terminology.

Common mistakes

• Testing too few queries — Testing 5 queries gives you a snapshot, not an audit. You need 20-50 queries across four categories to see the full picture. Too few queries leads to false confidence.
• Only testing one AI engine — ChatGPT, Perplexity, Google AI Overviews, and Gemini use different retrieval methods. Being visible in one does not mean you are visible in all. Test all four. See ChatGPT Search vs Perplexity for why they differ.
• Skipping the entity audit — Many brands jump to content fixes without checking whether their entity is clearly defined. If AI models cannot identify your brand as an entity, no amount of content optimization will get you cited. Start with entity clarity.
• Not recording competitors — Recording only your own presence/absence misses the competitive context. Always note which competitor is cited when you are not — this reveals where to focus. See how to track AI citations for ongoing monitoring.
• Running the audit once and never again — AI engine behavior changes with every model update. A one-time audit is a snapshot. Run quarterly and track changes over time. Explore GEO consulting and AEO consulting for systematic, ongoing AI visibility management.