Start with what happened.
Keep observations distinct from estimates, assumptions and missing data so the diagnosis can be examined.
HOW IT WORKS
A controlled operating loop, not a stream of unexplained recommendations.
Observe. Diagnose. Propose. Approve. Measure.
ILLUSTRATIVE SCENE · NO LIVE CAMPAIGN DATABEHIND THE APPROACH
Provider connections and campaign execution are not active on this website.
Keep observations distinct from estimates, assumptions and missing data so the diagnosis can be examined.
A person can approve or reject the proposed change. Rejection ends the action without silent escalation.
Compare the approved group against the agreed baseline before deciding whether the change should remain.
OUTBOUNDFIX / DEMO