How ChatGPT's Retrieval Stack Actually Works

RESONEO's analysis of 1,200 ChatGPT answers reveals three distinct retrieval layers—OpenAI's own index (labrador), cached Google results, and live-fetched pages—each with different staleness and citation rules.

9 Months of AI Overview Data Reveals Traffic Misattribution

A nine-month analysis of 51,000+ AI Overview events reveals that 22.4% of that traffic is misattributed to Direct in GA4, AI Overview prominence is volatile, and Google favors structured, specific content.

Fractl's analysis of AI model brand recall found that traditional SEO authority doesn't guarantee visibility in LLM answers—third-party coverage and category alignment matter more.

ChatGPT Fan-Out Queries Favor Trusted Sites, Study Shows

Recent research suggests ChatGPT prioritizes fan-out searches to already-trusted domains as a spam-reduction strategy.