How to Run an AEO Audit: Step−by−Step Guide for 2026

What an AEO audit checks (and how it differs from an SEO audit)

An AEO audit evaluates your brand's visibility in AI−generated responses. An SEO audit focuses on Google rankings and organic traffic. An AEO audit focuses on whether ChatGPT, Perplexity, Gemini, and Claude mention, cite, or recommend your brand when users ask questions in your category.

The audit covers five areas:

1. AI visibility across three layers. Does the AI know your brand from training data (parametric)? Does it find your content when searching with conversation context? Does it find you in a fresh session with no context?
2. Technical access. Can AI crawlers reach your pages? Are they blocked by robots.txt, slow load times, or JavaScript rendering?
3. Content structure. Is your content formatted for AI extraction? Answer−first paragraphs, standalone sections, structured data?
4. External signals. Are you mentioned on review platforms, Reddit, industry publications, and comparison articles that AI retrieves?
5. Competitive position. How does your AI visibility compare to direct competitors on the same target queries?

For a full overview of AEO and the three−layer model, see: What Is AEO: Complete Guide.

Gather your audit tools and inputs

Before starting, prepare these:

AI platforms to test (free accounts work)

• ChatGPT (with web search toggled on and off)
• Perplexity
• Google Gemini
• Claude
• Microsoft Copilot

Technical tools

• Google Search Console (indexing and crawl data)
• Google PageSpeed Insights or Lighthouse (load time and Core Web Vitals)
• Schema.org Validator or Google Rich Results Test (structured data)
• Your site's robots.txt file

Inputs you need

• 10–20 target queries (the questions your potential customers ask AI)
• 3–5 direct competitors (brands that serve the same audience)
• A spreadsheet for tracking results

For a step−by−step guide on choosing target queries and running manual checks: How to Check If Your Brand Is Recommended by ChatGPT.

Step 1: Test parametric knowledge

Parametric knowledge is what AI knows about your brand from training data, without searching the web. This is Layer 1 of the three−layer AI visibility model.

How to test

1. Open ChatGPT. Turn web search OFF.
2. Ask: "What do you know about [Your Brand]?"
3. Ask: "What companies do you recommend for [your category]?"
4. Ask your 10 target queries without mentioning your brand.
5. Record: Does the AI mention your brand? Is the information accurate? What competitors does it name instead?

What to look for

• Mentioned and accurate = strong parametric knowledge. Your brand is in training data with correct positioning.
• Mentioned but inaccurate = the AI has outdated or wrong information. You need to fix the sources it trained on.
• Not mentioned = no parametric knowledge. Common for brands under ~5 years old or those without Wikipedia, press coverage, or Reddit presence.

Score this step: Count how many of your 10 target queries return a mention. Fewer than 3 out of 10 means parametric knowledge is a gap.

Step 2: Test contextual web search

Layer 2 tests whether AI finds your content when it has conversation context. The user has already mentioned your category or a related topic, and the AI searches the web with that context.

How to test

1. Open ChatGPT with web search ON.
2. Start a conversation about your category: "I'm researching [your category] for [use case]."
3. Then ask: "What are the best options?"
4. Ask follow−up questions that match your target queries.
5. Repeat on Perplexity, Gemini, and Claude.

What to look for

• Does the AI cite your website or content?
• Does it cite your competitors instead? Which ones?
• When it cites you, does it pull accurate information?
• Which specific pages does it link to?

Score this step: Record your citation rate across platforms. Below 30% of contextual queries means your content structure or topical authority needs work.

Step 3: Test fresh session retrieval

Layer 3 is the hardest test. No conversation history, no context. The AI must find you purely from real−time web search.

How to test

1. Open a fresh session (new conversation) on each platform.
2. Ask each of your 10 target queries cold, without prior context.
3. Record whether your brand appears, what position it holds (first mentioned, second, third, or absent), and what source the AI cites.

What to look for

• Perplexity shows citation sources inline — check if your site appears.
• ChatGPT names brands in text — check if yours is named.
• Gemini may link to Google results — check if your pages appear.

Score this step: Below 20% appearance rate in fresh sessions means your external signals and content freshness need attention.

Step 4: Audit AI crawler access

If AI crawlers cannot reach your content, nothing else in this audit matters.

Check robots.txt

Open yourdomain.com/robots.txt and verify these bots are not blocked:

Many CMS platforms blocked AI bots by default during 2023–2024. If you find Disallow rules for these bots, remove them.

Check page load speed

Run your top 5 content pages through Google PageSpeed Insights. AI crawlers have timeout limits — pages taking more than 3 seconds may not be fully crawled.

Check rendering

View your pages with JavaScript disabled (Chrome DevTools → Settings → Disable JavaScript). If the main content disappears, AI crawlers may not see it either. Server−side rendered or static content is accessible to all crawlers.

Score this step: Any blocked AI bot = critical issue. Any page above 3 seconds = high−priority fix.

Step 5: Audit content structure for AI extraction

AI systems extract passages from your pages. The structure of those pages determines whether AI cites you or skips to a competitor.

Check each of your top 10 content pages for:

Score this step: Count how many of the 6 elements each page passes. Pages below 4/6 need restructuring.

Step 6: Audit schema markup

Schema markup helps AI systems understand your brand's identity and content structure.

Check for these schema types

How to test

1. Open the Google Rich Results Test.
2. Enter each key page URL.
3. Check which schema types are detected.
4. Validate that all required fields are populated.

Score this step: Homepage without Organization schema, or blog posts without Article schema = high−priority gaps.

For a detailed implementation guide: Schema Markup for AEO.

Step 7: Audit external mention density

AI systems cross−reference multiple sources before recommending a brand. External mention density is often a stronger signal than on−site content quality (Far & Wide research, 2025).

Check each of these sources

An analysis of AI citation patterns found that 46.7% of Perplexity's cited sources come from Reddit. Zero Reddit presence means invisibility on a major AI platform.

Score this step: Fewer than 10 independent sources mentioning your brand = weak external signals. Fewer than 5 = critical gap.

Step 8: Audit competitor AI visibility

Run the same tests from Steps 1–3 for your top 3–5 competitors. This shows your relative position and reveals which competitors are already optimized.

For each competitor, record

• Parametric knowledge: Does AI mention them with web search off?
• Citation rate: How often do they appear in your target queries?
• Source quality: What pages/platforms does AI cite for them?
• Positioning: How does AI describe them vs how it describes you?

Build a comparison table

This comparison often explains why competitors appear and you do not. If a competitor has 50 G2 reviews and you have 3, the fix is obvious.

For a detailed breakdown of why competitors outperform you: Why Your Competitors Show Up in ChatGPT and You Don't.

Step 9: Score your findings

Use this rubric to prioritize fixes.

Total your score out of 16. This gives you a baseline for measuring progress.

• 0–5: Your brand is invisible to AI. Start with AI crawler access and external signals.
• 6–10: Some visibility, but competitors outperform you. Focus on content structure and Layers 2/3.
• 11–16: Good foundation. Optimize specific gaps and monitor monthly.

Step 10: Build your fix roadmap

Based on your scores, prioritize by impact and timeline.

Fix within 1 week (critical/quick wins)

• Unblock AI crawlers in robots.txt
• Update outdated information on key pages
• Fix inaccurate brand information AI is citing

Fix within 2–4 weeks

• Restructure top 5 content pages for answer−first extraction
• Add missing schema markup (Organization, Article, Product)
• Create or complete profiles on G2, Capterra, or Product Hunt

Fix within 1–3 months

• Build content clusters covering your full topic area
• Create comparison pages ([Your Brand] vs [Competitor])
• Launch review collection campaign
• Build Reddit presence in relevant subreddits

Fix within 3–6 months (Layer 1)

• Evaluate Wikipedia notability and create page if eligible
• Pitch guest articles to industry publications
• Publish original research

For a complete optimization playbook organized by layer: How to Get Your Brand Recommended by AI Assistants.

AEO audit vs SEO audit: what's different

An AEO audit overlaps with SEO in some areas but diverges in others.

The two audits complement each other. Strong SEO feeds AI visibility — Google's index powers Gemini and AI Overviews. External signals that help AEO (review platforms, press coverage) also strengthen SEO. If you must prioritize: an SEO audit covers Gemini and Google AI Overviews; an AEO audit covers ChatGPT, Perplexity, and Claude.

Common AEO audit mistakes

1. Testing only one AI platform. ChatGPT, Perplexity, and Gemini use different retrieval methods and different sources. Testing only ChatGPT gives an incomplete picture. Test at least three platforms.
2. Not separating parametric from retrieval visibility. Testing only with web search on means you cannot tell whether AI knows your brand from training data or found it through search. Test with web search off first (Layer 1), then on (Layers 2 and 3).
3. Auditing your pages but ignoring external sources. Your website might be structured well, but without third−party mentions, AI has nothing to cross−reference. External mention density is often the bigger gap.
4. Running the audit once and never repeating it. AI models update, competitors optimize, content ages. Re−run your top 10 queries monthly.
5. Checking only whether you are mentioned, not how. Being mentioned with inaccurate positioning or outdated information can hurt more than being absent. Check context and accuracy of every mention.
6. Skipping the competitor comparison. Without competitor data, you cannot set priorities. A competitor with 50 G2 reviews sets the bar you need to clear.
7. Treating the audit as a checklist instead of a diagnostic. Every finding should connect to a specific fix with a timeline, not just a green or red mark.

When to DIY vs when to hire an expert

A DIY audit works when:

• You have 10 or fewer target queries
• Your team has 4–6 hours for the process
• You need a baseline understanding
• Budget is under €500

Hire an expert when:

• 20+ target queries or multiple product lines
• You need platform−specific optimization recommendations
• You want a scored audit with prioritized fixes and timelines
• You need before/after benchmarking with quantified targets
• Your team lacks dedicated AEO expertise

The DIY approach reveals the problem. An expert audit provides the roadmap. Many brands start with a manual check using our brand visibility guide, then bring in an expert for the full strategic audit.

Case study: online school goes from invisible to visible in 30 days

An online school came to us with zero AI visibility across all platforms. Our audit identified three critical gaps: AI crawlers were blocked in robots.txt, content was structured for human reading but not AI extraction, and external mention density was near zero.

After implementing the audit recommendations — unblocking crawlers, restructuring content for answer−first extraction, building review platform profiles — the school started appearing in AI recommendations within 30 days. Full details in the online school AEO case study.

Next steps

Start with a manual check: How to Check If Your Brand Is Recommended by ChatGPT.

For monitoring tools after your audit: Best AEO Tools to Monitor AI Visibility.

Ready for a professional audit? Get a Full AEO Audit & Strategy from Far & Wide (from €750). We check 200+ signals across 5 AI platforms, score your visibility on all three layers, and build a prioritized fix roadmap with before/after benchmarking. Includes a strategy session to walk through every finding.

Want to start smaller? Order a Brand Visibility Report (€80) — a snapshot of how ChatGPT sees your brand on your target queries, with 10 specific recommendations.