Review the decision, not the screenshot.
A useful review explains what happened, what the record can support and what the team should do next.
- File
- Type
- Reading time
- From
- Strategy
- Measurement
- Team leadership
Put the boundary around the result.
Start with the period, currency, platform and event being counted. Purchases, initiated checkouts, messages and completed payments describe different things. Keep their labels intact. If a period is unspecified, do not quietly turn a snapshot into a monthly result.
Separate an exceptional day from the full campaign. The linked March account includes both; the single-day return is useful context, but the thirteen-day period answers a different question. A review should help the reader see that distinction immediately.
Separate observation from explanation.
Observation is what the record shows. Explanation is why the team thinks it happened. Write them separately, then state what could challenge the explanation. A changed offer, a stock issue, attribution timing or a different audience may matter alongside the campaign edit.
Ask the person closest to each part of the work to contribute. Creative can explain the message; operations can explain availability; the media buyer can explain delivery. Respecting those boundaries makes the review more precise and the ownership more honest.
Finish with a decision.
A review needs a next action: continue, change, investigate or stop. Assign an owner and a review point. Say what evidence would support the choice and what would trigger a different one. Keep the first action small enough that the team can understand its effect.
Do not make the meeting a contest for the most impressive number. Ask whether the decision improved the work under its actual constraints. That question leaves room to learn from strong results, weak results and choices that were reasonable with incomplete information.
Keep a record someone else can use.
Save the context, the observation, the working explanation, the decision and the follow-up. Include relevant source links and clearly mark missing information. The next reviewer should be able to understand the reasoning without relying on the memory of the person who presented it.