Know what to test next.
Part of Atlas AI. Creative Radar finds promising ads and signs of fatigue. Follow the evidence to one specific change your team can test.
Fictional PLAYROOM close-up creative: click-through rate fell from 1.4% to 0.9% and cost per acquisition rose from $30 to $40 across matched 7-day periods. A proposed headline, "Pick your next slot.", replaces "Your games. In one place." The proposal is untested.
Read the whole ad.
Fatigue is rarely about one detail. Creative Radar looks at the image, the words and the action together.
Image
The product shot and how close it sits.
Supporting copy
The line that explains the offer.
Headline
The promise a person reads first.
Call to action
The words on the button.
Compare matched periods.
The same measures across two seven-day windows. Audience, spend and campaign setup can also move the result.
| Measure | Close-up, previous | Close-up, recent | Full-phone, previous | Full-phone, recent |
|---|---|---|---|---|
| Click-through rate | 1.4% | 0.9% | 1.2% | 1.8% |
| Cost per acquisition | $30 | $40 | $30 | $24 |
| Spend | $2,400 | $2,800 | $2,400 | $2,400 |
| Acquisitions | 80 | 70 | 80 | 100 |
Synthetic data. The close-up creative brought 70 acquisitions, down from 80, as spend rose.
A proposal is not a result.
A proposed edit has no measured outcome yet. It becomes evidence only after your team runs the test.
Proposed
One change, with the rest of the ad kept the same.
Agreed
Scope, budget and success measures set. The team creates the final ad.
Running
Matched periods. Watch click-through rate and acquisition cost.
Measured
Keep, revise or retire the change. The result informs the next test.
Which creative needs a closer look?
Bring a set of ads and one question your team wants to answer.