Reverse Engineering ChatGPT's Referral Logic

#The Experiment

We asked ChatGPT, Claude, and Perplexity the same 1,000 questions across 50 B2B verticals. Then we tracked which brands were mentioned, how often, and in what context.

The results surprised us.

#What We Found

#Pattern #1: Content Density Trumps Domain Authority

High-DA sites with thin content were consistently outperformed by specialized publications with comprehensive coverage. A niche blog with 50 detailed posts about compliance software beat a Forbes listicle every time.

#Pattern #2: Structured Data Matters (A Lot)

Brands with robust schema markup appeared 3.2x more often than those without. AI systems use structured data to understand entity relationships and build confidence in their recommendations.

#Pattern #3: Recent Updates Signal Relevance

Content updated within the past 6 months was cited 4x more frequently than evergreen content left untouched. AI systems interpret recent updates as a signal of ongoing relevance and accuracy.

#The Citation Framework

Based on our analysis, AI systems appear to weight these factors:

#What This Means for Your Strategy

If you want to be cited by AI systems, stop thinking about keywords and start thinking about becoming the definitive resource for your topic.

The brands winning in AI search aren't gaming the system. They're building genuine authority through depth, accuracy, and consistency.

#Action Items

1. Audit your core topic areas for content depth
2. Implement schema markup on key landing pages
3. Create an update schedule for cornerstone content
4. Build entity presence across knowledge bases

Want to see how your brand performs in AI citation tests? Contact us for a complimentary AI visibility audit.