Field notes
Practical thinking for AI discovery teams.
Research notes on crawler behavior, answer visibility, citation evidence, and the measurement habits that make them useful.
Published for people who need to explain what changed.
Measurement
Why a mention is not an outcome
A framework for separating answer presence, source selection, measurable referral traffic, and inferred influence.
- Direct versus inferred attribution
- Evidence boundaries
- Useful reporting language
Crawlers
Crawled is not the same as cited
How to diagnose the gap between machine access and inclusion in an AI-generated answer.
- Access and coverage checks
- Source preference signals
- Repeat-run validation
Operations
Build a prompt set that survives reporting
A stable approach to prompt ownership, provider coverage, evidence retention, and change review.
- Prompt grouping
- Run consistency
- Reviewable answer evidence
Early access