Notes on how AI search actually works
Short, specific write-ups on the mechanics behind the checks: crawling, citation selection, and what can and cannot be measured.
No posts yet. This route exists so articles have somewhere to land. Nothing is listed here that cannot be read — a blog index full of placeholder titles would be exactly the kind of invented content this product argues against.
What will be published here
- How an AI engine decides which page to cite, and which parts of that you control
- Why blocking GPTBot does not remove you from ChatGPT answers
- What a robots.txt 5xx actually does to every crawler that obeys RFC 9309
- Reading the difference between a memory-grounded and a retrieval-grounded answer
- What "AI referral traffic is undercounted" means in practice, and what cannot be recovered