From an AI visibility gap to a measured change
Turn a report gap into a suggested change, review the brief and its hypotheses, approve it, then measure before and after publishing to see what moved.
The brands and numbers in this video are invented. It has on-screen captions and no sound.
Transcript
After the French press page went live: recommended in 8 of 15 answers, up from 2.
Each gap in a report can become an action, tracked stage by stage.
Open one to review it.
The brief quotes the answers it was built from.
Anything the answers do not prove is labelled a hypothesis.
Edit it before you approve it.
Approve it when it is ready.
Then measure before you publish: a baseline runs first, so there is a before to compare against.
Two follow-up measurements after publication. The result is counts, before and after.
And it says plainly that a result shows a change, not what caused it.
When the control questions moved just as much, the result is marked inconclusive.
