# Less noise. More signal.

## 01 · Less noise. More signal.

FIELDNOTES / A fictional 30-day social review. Find the posts worth learning from, keep the calculations visible and make the next test concrete.

## 02 · A scorecard you can reconcile.

Rates are weighted from summed counts, never averaged across posts.

## 03 · Compare like-sized windows.

Equal inclusive date windows; totals can still reflect a different posting mix.

## 04 · Find the signal by channel.

These are descriptive comparisons, not proof of causation.

## 05 · Keep unknowns in the room.

1 outside-window records; 1 zero-impression records. Rates with a zero denominator stay unknown.

## 06 · Turn the review into a test.

Proposed experiments require your own assignment and success-window decision.