Every GEO claim, cross-checked against a receipt.
A living reference for Generative Engine Optimization. No hot takes, no vendor spin — each fact links to primary research so you can verify before you cite.
Emerging tactics we're monitoring
These have no industry consensus yet. Impact scores are baseline hypotheses, not verified benchmarks — treat them as experiments, watch how the community moves, then decide.
Vercel, Anthropic docs, and Stripe docs ship .md variants for agent consumption. Community anecdote is positive for agent workflows; no controlled citation-lift study yet.
Publishers (Anthropic, Cloudflare, Vercel) ship llms.txt as a courtesy index. No AI vendor has confirmed retrieval weight; independent tests show zero measurable citation lift.
Vendors floating the idea (a few AEO tool marketers) vs. established webmaster guidelines that treat bot/human divergence as cloaking. Consensus: risky until an engine explicitly permits it.
Get new receipts as they land.
Every time a claim is added, updated, or corrected, you'll hear about it — plus a Friday digest of what moved.