Citarra field notes
How AI decides what gets recommended.
Original thinking on citations, measurement, and the gap between being visible and being chosen. Published when the evidence changes our mind.
Issue 01 / AI discoveryResearch, not content volume
Featured / Measurement / 6 min read
What actually decides whether AI names your brand
Most teams optimise for the wrong thing. See what citation data reveals about the sources assistants lean on and why one prompt tells you almost nothing.
Read the field note ↗Latest thinking
Read by question, not chronology.
Why one AI visibility score hides the thing you need to know
Averaging engines into one number destroys the signal that tells you what action to take next.
↗You cannot prove AI visibility drove revenue, and you should stop trying
The honest version of attribution is weaker than most vendor claims and considerably more useful.
↗Our editorial rule
If a claim cannot be traced back to a question, an answer, or a source, it does not belong in the story.The same principle guides the product: evidence first, summary second.
Turn the research into a view of your own market.
Track your brand across three AI engines for 14 days. No card required.
Start free