Evidence-led analysis of how AI answer engines retrieve, select and cite their sources.
A new preprint finds limited overlap between sources shown by chatbot interfaces and their APIs. AEO reports need to state precisely what they observe.
Read the full report: A September AI shopping audit exposes the gap between API tracking and consumer answersWhat published research says about how AI answer engines choose sources, where AI referral traffic is heading, how the engines differ, and what still cannot be measured.
Read more: The State of AI Visibility in 2026Retrieval, query fan-out, freshness and authority: what the platforms document, what independent studies have measured, and what is still hypothesis.
Read more: How AI answer engines choose which sources to cite: what the evidence saysGoogle says optimizing for its AI features is still SEO. The evidence largely agrees on the foundations, and diverges on how visibility is won, measured and turned into traffic.
Read more: AEO vs SEO: what's different, and what still matters
