The situation
Impressions rise while clicks are flat. Period data can be measured; the cause of the change cannot be invented.
Decision to review
MEASURE → OBSERVE
Before/after observations around completed work, with the change recorded as correlation or observation rather than claimed causation.
