Creative KPIs: what to measure, what to ignore, and how to act on every metric
Measuring everything is measuring nothing. Creative operations run on four metrics — and a handful of anti-metrics best kept off the radar so they don't poison the team.
Why creative teams measure the wrong things
The temptation is to measure output: how many pieces were produced this month. It's the worst possible metric, because it optimizes for volume at the expense of quality and reuse. A well-measured team produces fewer but better-focused pieces, and reuses more assets from the library.
The metrics that matter in creative operations measure flow, not output: how much friction there is between brief and published.
The four metrics that matter
1. Average approval time (AAT)
What it measures: days from when a brief enters the pipeline to when the piece is approved.
How to act: if it climbs, audit the approval stage. The bottleneck is almost always the client, but sometimes it's internal review in disguise.
Benchmark: a healthy agency = under 7 days for standard pieces; a healthy in-house team = under 5 days.
2. Asset reuse rate
What it measures: % of new pieces that reuse at least one previously approved asset.
How to act: if it's low (<30%), the problem is usually that the library isn't searchable. Better search and better tagging come before more new photography.
Benchmark: mature teams reach 50-70%.
3. Revisions per piece
What it measures: average number of iterations (versions) per piece before approval.
How to act: too many (>5) = a weak brief. Too few (<2) = lax approval or a disengaged client.
Benchmark: 3 iterations is the sweet spot for most piece types.
4. Coordination vs. creation
What it measures: % of the team's time spent coordinating (meetings, emails, status updates) vs. creating (designing, writing, producing).
How to act: if coordination exceeds 30%, the operation is badly designed. Audit where the time goes (run two weeks of lightweight time-tracking).
Benchmark: healthy teams sit at 15-25% coordination.
Anti-metrics: what you should NOT measure
- Number of pieces per month. It optimizes for volume and kills quality. Measure demand absorbed, not pieces produced.
- Hours worked per person. It rewards presence, not results. It breeds a culture of performative late nights.
- Likes / engagement per piece. You're not measuring the creative team, you're measuring the algorithm. Useful for marketing, not for operations.
- Raw speed without context. Producing an ad in 2 hours isn't a good KPI if quality dropped 40%.
How to start measuring from scratch
- Pick a single metric to start. Recommendation: AAT. It's the easiest to capture and the most revealing.
- Capture a baseline for a month without changing anything. What you measure at first will be worse than you expect.
- Identify the biggest bottleneck in that metric.
- Apply a single intervention. Measure the next month.
- Once the metric is under control, add the next one.
How Polimake helps you measure
The four metrics above are calculated automatically when the entire cycle lives on one platform. Studio records stage-by-stage timing and counts versions, and Media measures asset reuse.
Free account