Campaign analysis template for agencies
A practical template for closing a campaign with verifiable data, explaining results to a client, and turning lessons into decisions for the next cycle.
The team behind Polimake. We explore the intersection of technology, creativity, and automation.
What this analysis should achieve
A post-campaign report is not a collection of screenshots or a defense of completed work. It should answer four questions:
- What were we trying to change?
- What happened according to the agreed sources?
- Which explanation has support, and what remains a hypothesis?
- What will we do differently in the next cycle?
Complete the document with strategy, media, production, and account management. If a number cannot be reconstructed from its source, do not present it as definitive.
Copyable template
1. Identification
- Client:
- Campaign:
- Period analyzed:
- Report owner:
- Close date:
- Version:
- Sources used: ad platform, web analytics, CRM, sales, survey, or others.
- Known limitations: attribution, reporting delays, tracking changes, small sample, or others.
2. Objective and success criterion
- Business problem:
- Priority audience:
- Behavior we wanted to influence:
- Primary metric:
- Baseline:
- Agreed target:
- Diagnostic metrics:
- Out of scope:
Do not turn every available metric into a KPI. Reach can diagnose distribution; it does not independently prove awareness, consideration, or sales.
3. What actually ran
| Element | Plan | Actual execution | Difference and reason |
|---|---|---|---|
| Audience | |||
| Channels | |||
| Assets and formats | |||
| Media spend | |||
| Production cost | |||
| Dates | |||
| Experiments |
Include late changes. Comparing outcomes with a plan that never ran produces false conclusions.
4. Results
| Metric | Source | Baseline | Target | Result | Status |
|---|---|---|---|---|---|
Use the same attribution window and currency throughout. If two systems disagree, retain both figures, explain the difference, and identify which one informs the decision.
5. Creative performance
For each relevant asset, record:
- asset link or ID;
- creative hypothesis;
- audience, channel, and format;
- spend or distribution received;
- comparable metric;
- decision: retain, adapt, stop, or retest.
Do not call an asset a winner when it received more budget, more time, or a different audience without saying so. Creative comparisons need reasonably equivalent conditions.
6. Cost and return
- Revenue attributable under the agreed source:
- Media spend:
- Production and adaptation:
- Tools and suppliers:
- Discounts, returns, or margin considerations:
- Attribution method:
If the objective was awareness or learning, do not force a financial ROI. Explain the expected value and measurement method agreed before launch.
7. Reading the results
Separate fact from interpretation:
- Observation: directly visible in the data.
- Likely explanation: supported by more than one signal.
- Open hypothesis: requires another test.
- External factor: price, stock, tracking, competition, or timing change.
Example: Version B had a higher click-through rate is an observation. The product close-up caused the improvement is a hypothesis if copy and audience also changed.
8. Decisions
| Decision | Evidence | Owner | Date | Status |
|---|---|---|---|---|
Limit the closeout to three to five decisions. A list of twenty ownerless recommendations will not change the next project.
Hypothetical example: a B2B acquisition campaign
This example demonstrates the template; it is not a Polimake customer or result.
An agency runs four ads to generate assessment requests. The agreed objective is to increase qualified forms without exceeding the client's maximum acquisition cost.
At closeout, it observes:
- form volume increased, but several leads did not meet the company-size requirement;
- the ad showing the deliverable earned fewer clicks and more qualified requests;
- two days of tracking were lost after a website change;
- sales took longer than planned to log call outcomes.
The team does not claim that the deliverable image “converts better” in general. It decides to retest that direction with two messages, change the form, and add a mandatory CRM field. The report retains the tracking failure as a limitation.
A 45-minute closeout meeting
- 5 min: objective and limitations.
- 10 min: execution against plan.
- 10 min: results and creative performance.
- 10 min: explanations and hypotheses.
- 10 min: decisions, owners, and dates.
Share the document beforehand. The meeting should make decisions, not force everyone to read tables live.
Checklist
- Every KPI has a source and definition.
- Costs include media and production where relevant.
- Comparisons use compatible periods and conditions.
- Facts are separated from hypotheses.
- Data failures appear as limitations.
- Analyzed assets retain their version and link.
- Every decision has an owner and date.
- The client has approved what may be reused in case studies or sales materials.
Analysis ends when decisions enter the next brief, calendar, or experiment—not when someone exports the PDF.
If the team has not yet defined objectives and owners, start with the monthly content strategy template. To turn lessons into a repeatable system across clients, see content operations for agencies.