A creative report explains which advertising ideas received useful response, what the evidence can support, and what the team should test next. It connects performance numbers to the message and execution people actually saw.
A list of thumbnails sorted by ROAS is a starting point. It does not explain whether the comparison is fair or what a designer should make next.
Define the question and reporting scope
Start with a specific question, such as “Did a product demonstration bring more qualified enquiries than a customer quote?”
State the platform, objective, dates, audience, placements, and attribution rules. Note any changes to the offer or landing page. Without that context, a creative shown to returning customers can appear better than one introducing the product to strangers.
Separate the campaign’s business outcome from the diagnostic metrics used to understand it.
Label the ideas behind the files
Give each asset a concept label, opening message, format, and version. Include the thumbnail or a link to the file so the reader does not have to interpret an obscure asset ID.
For example:
- Concept: saves setup time.
- Hook: show the finished workspace first.
- Format: vertical product demonstration.
- Version: demonstration B, revised opening.
Use a stable file naming convention so those labels can connect to your production records.
Include enough delivery context
Show spend, impressions, the relevant response measure, and the chosen outcome. A rate without its volume can be misleading.
Here is a fictional example:
| Creative | Spend | Link clicks | Qualified leads | Cost per qualified lead |
|---|---|---|---|---|
| Demonstration A | €400 | 500 | 10 | €40 |
| Quote B | €100 | 150 | 2 | €50 |
The demonstration has a lower observed cost per qualified lead. The uneven spend and small outcome counts mean this table alone cannot establish a reliable winner or isolate the reason for the difference.
Describe observations before explanations
Observation: “The demonstration generated 10 qualified leads at €40 each in this period.”
Possible explanation: “Showing setup may have answered a common concern.”
Next test: “Compare two demonstrations that keep the offer and format consistent but show different setup steps.”
Keeping those statements separate prevents a plausible story from becoming an unsupported fact.
Avoid misleading comparisons
Check whether one asset ran for fewer days, reached a different audience, or benefited from a different promotion. Conversion delays can also make the newest creative look worse before all results arrive.
Advertising systems may allocate delivery unevenly. Observed differences from normal delivery are not automatically equivalent to a controlled experiment.
Use CTR and conversion rate to describe where response differed, and ROAS versus ROI to keep revenue claims separate from profit claims.
End with production decisions
Finish with a small set of actions: continue a concept, make a specific variation, investigate a tracking issue, or gather more data.
Assign each action an owner and a review point. A useful report leaves the team knowing what to make and what it expects to learn from that work.