Which technical SEO factors actually matter for AI visibility? Our site ranks well on Google but gets zero AI citations

Great Google rankings do NOT guarantee AI visibility. Here’s why:

How AI crawlers differ from Googlebot:

What Google masks that AI exposes:

1. Authority compensation - Google weighs your DA and backlinks heavily. AI doesn’t care about links - only content quality and accessibility.

2. JavaScript rendering - Googlebot renders JS after initial crawl. AI crawlers see only raw HTML.

3. Mobile-first - Both care, but AI crawlers may fail faster on poor mobile.

4. Speed tolerance - Google factors speed but compensates with authority. AI systems just skip slow sites.

Your likely culprits:

Given good Google rankings but no AI citations, check:

1. JavaScript rendering of critical content
2. Robots.txt blocking AI user-agents
3. CDN/Cloudflare blocking AI bots
4. Content structure (machine-readable vs. human-readable)

Page speed perspective - this matters more for AI than Google:

Why speed hits AI harder:

AI platforms crawl billions of pages consuming massive computational resources. OpenAI’s expansion needs 10 gigawatts of power. Every slow page wastes resources.

The math:

• Slow site = more crawl resources
• More resources = higher cost
• Higher cost = deprioritization
• Result = fewer AI citations

Speed benchmarks for AI:

Your “all green” Core Web Vitals:

Google’s thresholds are lenient. For AI:

• Google “good” = 2.5s LCP
• AI preference = Under 1.5s LCP

You might pass Google’s bar but still be slow for AI.

Speed optimization priority:

1. Server response time (TTFB)
2. Image optimization (WebP/AVIF, lazy loading)
3. JavaScript reduction (fewer/smaller bundles)
4. CDN caching (serve from edge)
5. Eliminate render-blocking resources

Schema markup and structured data - often missing on high-ranking sites:

Why schema matters more for AI:

Google uses signals beyond schema (links, authority, engagement). AI systems rely heavily on structured data to:

• Understand content type
• Extract information confidently
• Verify entity information
• Reduce ambiguity

Schema that impacts AI (~10% of Perplexity ranking):

1. Article/TechArticle - Content type identification
2. FAQPage - Question-answer extraction
3. HowTo - Step-by-step processes
4. Organization - Entity recognition
5. Product/Service - Commercial intent clarity
6. BreadcrumbList - Site hierarchy understanding

Implementation checklist:

{
  "@context": "https://schema.org",
  "@type": "Article",
  "headline": "Your Page Title",
  "datePublished": "2026-01-06",
  "dateModified": "2026-01-06",
  "author": {
    "@type": "Person",
    "name": "Author Name",
    "url": "https://yoursite.com/author"
  },
  "publisher": {
    "@type": "Organization",
    "name": "Your Company"
  }
}

Common mistakes:

• Schema that doesn’t match visible content
• Outdated dateModified timestamps
• Missing author/publisher info (E-E-A-T signals)
• No FAQPage schema on FAQ sections

Validate with Google’s Rich Results Test AND Schema Markup Validator.

Content structure perspective - what AI needs vs. what humans see:

The human vs. machine reading gap:

Humans:

• Scan visually
• Interpret context
• Fill in gaps
• Navigate intuitively

AI crawlers:

• Parse HTML sequentially
• Need explicit context
• Can’t infer meaning
• Follow structure rigidly

Structural elements that matter:

1. Heading hierarchy

H1 (one per page)
  H2 (major sections)
    H3 (subsections)

Never skip levels. Each heading = content boundary.

2. URL structure Good: /features/sso-configuration Bad: /page?id=12345

Descriptive URLs signal content before parsing.

3. Internal linking

• Bidirectional links show relationships
• Descriptive anchor text aids understanding
• Topic clusters signal authority

4. Content chunking

• Short paragraphs (2-3 sentences)
• Self-contained sections
• Lists for scannable info
• Tables for comparisons

The visibility test:

If you removed all styling from your page, would the structure still make sense? That’s what AI crawlers see.

I just checked Cloudflare - AI bots were blocked by default. This explains everything.

My audit findings:

1. Cloudflare blocking - AI bots blocked (FIXED NOW)
2. JavaScript content - Some critical content JS-rendered
3. Schema gaps - No FAQPage schema, incomplete Article schema
4. Speed - 2.3s LCP (passes Google, but not ideal)

My technical action plan:

Immediate (Today):

• Enable AI crawler access in Cloudflare (DONE)
• Test with curl to verify access

Week 1:

• Audit JavaScript rendering on top 50 pages
• Implement SSR for critical content
• Add FAQPage schema to all FAQ sections

Week 2-4:

• Complete Article schema with author info
• Speed optimization (target 1.5s LCP)
• Heading hierarchy audit

Ongoing:

• Monitor AI citations via Am I Cited
• Track correlation between fixes and visibility
• Regular infrastructure audits

Key takeaways:

1. Google rankings mask technical debt - AI exposes issues Google compensates for
2. Third-party infrastructure matters - Cloudflare was blocking us without our knowledge
3. Different crawlers, different requirements - Can’t assume Googlebot success = AI success
4. Schema matters more for AI - Not optional anymore

The humbling realization:

We thought our technical SEO was solid because Google said so. AI crawlers revealed a completely different story.

Thanks everyone for helping diagnose this!