Short-Form Campaign Retrospective Template

1. Scope

Anchor the review in the decision the campaign was meant to support.

Define the campaignName the objective, date range, audience, channels, and product or offer.Campaign scope

Record what changedList the control and every tested change. Mark cells that changed more than one important variable.Control and tested changes

2. Evidence

Record comparable evidence and its limits.

Campaign evidenceEnter spend, impressions, clicks, installs, conversions, watch signals, and each metric window you relied on.Evidence table

Limits and confoundersName missing events, unequal delivery, attribution differences, small samples, or campaign changes that weaken comparison.Evidence limits

3. Learning

Separate creative facts from the story you tell about them.

ObservationsList what the data and creative review directly show.Observed facts

InterpretationsList plausible explanations and what additional evidence would raise confidence.Interpretations and needed evidence

4. Decisions

Make an explicit decision for the campaign and its source material.

KeepName what should continue unchanged because it is useful or still needed as a control.Keep

StopName what should stop because it is contradicted, unsafe, too costly, or no longer useful.Stop

StartName the smallest new behavior, asset, or test the team should begin.Start

Reuse the right footageSort hooks, UGC, demos, proof, and CTAs into reuse now, needs work, archive, and rights unclear.Asset disposition

5. Next cycle

Carry one clear question into the next cycle.

Define the next testWrite the next hypothesis, changed variable, control, evidence floor, owner, and review date.Next-cycle hypothesis

Close the campaign without losing the learning.

Start with the objective and tested changes so the review does not drift into every opinion the team has about the campaign.

Treat metrics as evidence only when their definitions and windows match. Keep observations separate from explanations, especially when spend or event volume is limited.

The most useful retrospective ends with asset decisions and a smaller next hypothesis—not a long list of vague lessons.

What to know before you use it.

Does this connect to my ad accounts?

No. You enter the evidence you trust, and the worksheet stays in this browser session until you copy or download it.

What if the campaign did not produce a clear winner?

Record that honestly. An inconclusive result can still reveal tracking gaps, weak controls, reusable footage, or a better next question.