Everything we publish, in one place

Original data from our own sampling — volatility, engine divergence, Chinese citation patterns, recrawl lag. Method and sample size published every time.

Field notes on measuring AI answers: what breaks, what we changed our minds about, and what agencies should actually report.

AEO, GEO, citation share, query fan-out, grounding — defined properly, with the distinctions that matter and the ones that don't.

Every AI crawler you'll see in your logs, with its user-agent string, robots.txt token, verification method and what blocking it costs.

robots.txt and llms.txt generators, a prompt-set builder and a sampling calculator. No signup, nothing leaves your browser.

The API itself: runs, sampling, webhooks, evidence exports, canaries and response schemas.

Direct answers to the questions buyers actually ask about methodology, coverage and how we make money.

Per-engine pages covering access model, citation structure, audience size and the gotcha that trips up reporting.