Brand and creative teams
Decide from evidence, and stop chasing the handoff.
Two problems, and only one is about creative. Nobody can actually read a quarter of work, so what to make next is answered by whoever is most confident. And requests arrive from performance as messages, then leave as surprises.
- Evidence, not opinion
- No causal claims
- Your own sources
What did the creatives that worked last quarter have in common?
Reading 214 assets, scene by scene, against what each ran with
One pattern holds across the top performers.
| Element | In top set | In the rest |
|---|---|---|
| Face in first 2s | most | few |
| Price on screen | few | most |
| Under 15 seconds | most | about half |
A pattern, not a cause. The test that would settle it comes with it.
Read fromYour creative libraryMeta AdsGoogle Ads
What it costs today
The bottleneck is not production capacity. It is the space between production and approval.
A quarter of work
The pattern is invisible however long you stare at the thumbnails.
Platforms report on the ad, never on what is inside it. So the elements that actually drive the result are not in any report you have.
Five of these outperformed the rest, and what they share is not visible in any platform's reporting.
Schematic · a quarter of work, read in full rather than sampled
The treadmill
You need thirty a month. You can make ten.
Fatigue lands three to four weeks in, which is roughly when the replacements were due.
- Five to ten new creatives a sprint
- What to make next, decided by whoever argues hardest
- Results read at reach and engagement, then the trail goes cold
- Every asset read against what it ran with
- The winning pattern named, with the test that would settle it
- Each brief built on everything made before it, not on the last one
What this is made of
One system for the work, one for the queue.
Before you ask
Three that come up every time.
Will this reduce our work to a metric?
The opposite. It refuses to tell you a creative caused a result, because how the platform chose to serve an ad is tangled up in how that ad performed. What it gives you is the pattern across everything you made, and the test that would settle it. That is more honest than a scorecard and more useful than an opinion.
Who has to keep the tracker current?
Nobody. The reason trackers die is that updating them is somebody's unpaid job, so status moves as a side effect of the work: a requisition raised, a deliverable finished, a handover due, posted to the right group without anyone typing.
Does it read video, or only statics?
Both, and the text inside them. A video is examined scene by scene with a timestamped transcript, and a static is read including its copy, then keyed to the ad, campaign, product, offer or audience it ran against.
Start here
Bring last quarter's creative and the handoff you argue about.
We will read the work against what it ran with, and map the requisition seam properly. You keep both.