Why Should You Perform an AI Visibility Audit?
Your site might be ranking on Google, but ChatGPT might never be citing it as a source. Perplexity might know you but present you with incorrect or incomplete information. These situations are independent of each other, and both are critical for your business.
Traditional SEO audits catch technical errors, speed issues, and content gaps. AI visibility auditing, however, goes a step further: it evaluates how AI systems read, understand, and source your content. The following 25 items allow you to perform this evaluation systematically.
Each "yes" is one point; aim for 20 and above.
Section 1: Technical Accessibility (1-7)
- 1. robots.txt check: Are you not blocking AI crawlers like GPTBot, Claude-Web, PerplexityBot, Google-Extended? (If no blocking ✓)
- 2. llms.txt file: Have you created an
/llms.txtfile in the site root? This file tells AI systems how to use your site. - 3. XML sitemap: Is it up-to-date, error-free, and does it cover all important pages? Are there no sitemap errors in Google Search Console?
- 4. Page speed: Is LCP below 2.5 seconds in Core Web Vitals? Slow pages consume crawl budget.
- 5. HTTPS: Is HTTP→HTTPS redirection active and is the certificate valid?
- 6. JavaScript dependency: Are critical contents visible without JavaScript (like a bot)? Test Google's Rich Results Test in JS-off mode.
- 7. Canonical URLs: Does the
<link rel="canonical">tag for each page point to the correct page?
Section 2: Structured Data (8-12)
- 8. Organization schema: Is there an
OrganizationorWebSiteJSON-LD schema on the homepage? Are thename,url,sameAsfields filled? - 9. Article/BlogPosting schema: Is there an
Articleschema in blog posts? Are theauthor,datePublished,dateModifiedfields included? - 10. FAQPage schema: Are you using FAQPage schema on at least 3 different pages? Does each question-answer pair truly add value?
- 11. BreadcrumbList: Have you marked the site hierarchy with BreadcrumbList schema?
- 12. Schema validation: Are there no errors in Google Rich Results Test and Schema Markup Validator?
Section 3: Content Quality (13-18)
- 13. H1 headings: Is there exactly one H1 on each page, and does this H1 clearly express the main topic of the page?
- 14. Content depth: Are blog posts at least 800 words? Are technical guides 1500+?
- 15. Concrete data and citation: Do contents include research, statistics, or external source citations?
- 16. Question-answer structure: On every important page, do you use questions a user might ask as H2/H3 headings and answer them immediately below?
- 17. Recency: Have contents been reviewed within the last 12 months? Is
dateModifiedfilled in the Article schema? - 18. No duplicate content: Are the pages on the site not duplicates of each other? Was canonical or noindex used?
Section 4: E-E-A-T Signals (19-22)
- 19. Author information: Is the author's name and a short biography visible on every blog post? Is there a Person schema for the author?
- 20. About Us page: Are company history, team, contact information, and physical address clearly stated?
- 21. External backlinks: Do industry publications, news sites, or universities link to your site?
- 22. Google Business Profile: Is the business profile complete, verified, and up-to-date?
Section 5: Specific Tests for AI Systems (23-25)
- 23. ChatGPT test: When you ask "What is / what does [your domain name] do?" with "gpt-4" or "gpt-4o", do you receive a correct and complete answer?
- 24. Perplexity test: Does your site appear as a source in a specific question related to your industry? Is the information shown correct?
- 25. Gemini test: When you ask for your company name or product on Google Gemini, are you featured in the information box or the answer?
Interpret the Results
| Score | Status | Recommendation |
|---|---|---|
| 22-25 | Excellent | Focus on content creation and backlink building |
| 17-21 | Good | Address missing structured data and E-E-A-T gaps |
| 12-16 | Average | Prioritize content depth and technical infrastructure |
| 0-11 | Critical | Start with the technical foundation and basic schemas |
Priority Order: Where Should You Start?
If your audit results show deficiencies in multiple areas, follow this order:
- Technical obstacles (1-7): If you are blocking AI crawlers, everything else is meaningless. Resolve this first.
- Organization and Article schemas (8-9): Most impactful, fastest to implement changes.
- Content depth (14): Identify and expand pages with fewer than 800 words.
- FAQPage schema (10): Critical for both Featured Snippets and AI Overviews.
- E-E-A-T (19-22): Fundamental for long-term authority building.
Automated Audit: GeoSkoru
Manually checking most of the items on this list can take hours. GeoSkoru automatically evaluates your site based on 7 GEO criteria by entering your URL; you see your scores for structured data, content depth, E-E-A-T signals, and technical readiness in minutes.