SOURCE
9 supplied post records
Skillrove / Reviewed source
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.
BOUNDARY
Fictional practice data. No measured customer claim.
Skillrove / Reviewed source
A scorecard you can reconcile.
Rates are weighted from summed counts, never averaged across posts.
85,000
Impressions
4.15%
Likes + comments + shares / impressions
1.25%
Clicks / impressions
Skillrove / Reviewed source
Compare like-sized windows.
Equal inclusive date windows; totals can still reflect a different posting mix.
4 to 5
Observed posts
56,000 to 85,000
Impressions
2.80% to 4.15%
Weighted engagement
Skillrove / Reviewed source
Find the signal by channel.
These are descriptive comparisons, not proof of causation.
63,000 impressions · 4.33% engagement
22,000 impressions · 3.64% engagement
Skillrove / Reviewed source
Keep unknowns in the room.
1 outside-window records; 1 zero-impression records. Rates with a zero denominator stay unknown.
NOT REACH
Impressions can include repeated views.
NOT REVENUE
Clicks and listed spend do not prove sales.
Skillrove / Reviewed source
Turn the review into a test.
Proposed experiments require your own assignment and success-window decision.
Test two different process openers
Keep topic, format and 7-day observation window matched.
Publish a controlled checklist draft
Do not infer purchases from click counts.