Global GEO Whitepaper (Part 2): Empirical Insights from 10.04M AI Responses on Citation Ecology, Funnel Intent & Drift
Global GEO Whitepaper (Part 2): Empirical Insights from 10.04M AI Responses on Citation Ecology & Drift
Quantitative findings from 10,045,768 AI responses and 1.35M unique domains across ChatGPT, Gemini, Perplexity, Google AI Overviews, and AI Mode.
1. The Zero-Consensus Law: Cross-Platform Domain Jaccard Is Only 0.056–0.130
Across 10.04 million AI responses, the pairwise Jaccard similarity of cited domains between any two AI platforms ranges from just 0.056 to 0.130. Winning citations in Google AI Overviews does not guarantee visibility in ChatGPT or Perplexity.
Citation concentration also varies dramatically: ChatGPT exhibits a power-law concentration toward mega-platforms (Wikipedia, Reddit, Forbes, Amazon), whereas Google AI Mode distributes citations across a much broader long tail of specialized vertical domains.
- Source Role Shift Across the Funnel: From ToFu (Informational) to BoFu (Transactional/Comparison), UGC and encyclopedic share drops sharply while Official Brand Sites and B2B Review Platforms (G2, Trustpilot) surge.
- Freshness Lifeline: Over 65% of cited URLs in dynamic commercial categories were published or substantively updated within the trailing 12 months.
- Monthly Citation Drift: 40%–59% of cited URLs for identical prompts rotate month-over-month, proving that GEO requires continuous monitoring and quarterly freshness updates.
Map our brand's GEO content assets across the 3 funnel stages based on the 10M+ AI Response Study: - [ToFu / Awareness]: Create definitional guides, industry benchmark reports, and YouTube/Reddit educational threads. - [MoFu / Consideration]: Build structured "Brand A vs Brand B" comparison matrices, G2/Trustpilot review velocity programs, and use-case selection tables. - [BoFu / Decision]: Publish SSR pricing tables, API/security compliance sheets (SOC2/GDPR), and TCO calculators on our official domain.
Engineering Comparison Matrix: Anti-Patterns vs. Production Best Practices
| Dimension | Common Anti-Pattern | 2026 Production Standard |
|---|---|---|
| Cross-Platform Overlap | Tracking only Google SERP rank as a proxy for AI visibility | Independent Share-of-Model tracking across ChatGPT, Perplexity, Gemini, and Google AIO |
| Citation Stability | One-off "publish and forget" blog posts | Quarterly data refreshes with explicit dateModified schema to combat 40–59% monthly drift |
Pre-Flight Production Verification Checklist
- 01Audit domain overlap across ChatGPT, Perplexity, and Google AIO separately.
- 02Align content format to funnel intent (Comparison tables for MoFu; Official specs & pricing for BoFu).
- 03Click "Switch to Original Chinese Full-Text & 56 Charts" above to explore all 16 empirical charts in Part 2.
Found this actionable? Copy standard citation & summary
Formatted with canonical attribution link and executive summary—ready for Slack, Notion, or LLM context.