Creative
How many creatives do you need per month in 2026?
A simple formula for sizing creative volume from spend, target CPA and test share, plus how to split concepts from iterations and when to kill or scale an ad.
Tomás Lindqvist
Head of Paid Social
Most creative plans start from the wrong end. Someone picks a number, usually “more than last month,” and production scrambles to hit it. The right volume falls out of four inputs you already have: monthly spend, target CPA, how many audiences and offers you run, and how fast your ads wear out.
Here is how we size it for clients.
Start with the cost of a read
Every new creative needs enough spend to tell you something. Judge too early and you kill ads that would have worked. Judge too late and you pay for losers.
Our rule of thumb: give each new creative about 5× your target CPA before calling it. Clear losers can be cut sooner (more on that below). At a €50 target CPA, one read costs about €250.
Then decide how much of your budget goes to testing. For most accounts we use 15–20%. The rest runs on proven ads.
The sizing formula
New creatives per month = (monthly spend × test share) ÷ (target CPA × 5)
Say you spend €50k a month with a €50 target CPA and a 20% test share. That’s €10k of test budget ÷ €250 per read = 40 new creatives a month.
The same math across spend bands, assuming a €50 CPA and concepts shipped in batches of four variants:
Monthly spend | Test budget | New creatives / month | New concepts / week |
|---|---|---|---|
€15k | €3k (20%) | ~12 | 1 |
€30k | €6k (20%) | ~24 | 1–2 |
€50k | €10k (20%) | ~40 | 2–3 |
€100k | €20k (20%) | ~80 | 4–5 |
€200k | €30k (15%) | ~120 | ~7 |
Two things move the number. First, CPA: a cheaper conversion makes a cheaper read. In our Kelpie Skin demo case (illustrative figures; results vary), CPA fell from €40 to €25, which buys 60% more reads from the same test budget. Second, test share: above roughly €150k a month, proven winners carry more of the load, so the share can drop.
Check the floor: audiences and offers
The formula tells you what you can afford. It doesn’t tell you what you need.
Every audience-offer pair that needs its own message needs its own fresh creative. If you sell to three distinct audiences with two offers each, that’s six combinations. We plan at least one new concept per combination per month: six concepts, or roughly 24 creatives once you add iterations.
If the floor is higher than the formula allows, you have more segments than budget. Consolidate audiences or offers before you add production.
Plan for fatigue
Winners wear out. The signs are consistent: frequency climbs, hook rate and CTR slide, and CPA drifts up week over week. At higher spend, ads saturate their audience faster, so the refresh rate rises with budget.
A practical trigger: when an ad’s 7-day CPA runs 25% or more above its own first two weeks, start replacing it. Don’t wait for it to die. Have its successor already in testing.
Concepts vs. iterations vs. formats
Not every “creative” counts the same.
Concept: a new reason to buy. A different angle, story or proof. “Founder explains why” and “customer shows the before and after” are two concepts.
Iteration: the same concept with one variable changed: the hook, the creator, the first frame, the length or the headline.
Format: the same asset resized for a placement: 9:16, 4:5, 1:1, static or video.
Formats are distribution, not tests. Don’t count them in your volume number. A 9:16 video recut to 4:5 is one creative, not two.
For the mix, we aim for about one slot in four going to new concepts and the rest to iterations of what’s already working. Iterations win more often. Concepts find the next winner.
When to kill, when to scale
Write the rules down before the test starts, so nobody argues with a chart afterward.
Kill when:
It has spent 2× target CPA with zero conversions.
It has spent 5× target CPA and its CPA is more than 30% above target.
Its hook rate (3-second views ÷ impressions) sits well below your account median after a couple of thousand impressions. The opening isn’t stopping anyone.
Scale when:
It hits target CPA or better over at least 10 conversions. Ads that survive the first read get extra budget to get there.
Its performance holds when you move it into your main campaign next to proven ads.
Then iterate on it straight away. A winner earns 3–5 new hook variants the following week. That’s where most of next month’s creatives should come from.
A quick sizing checklist
Write down monthly spend, target CPA and test share.
Run the formula.
Count audience-offer pairs and check the floor.
Set a refresh trigger for winners.
Split the number into concepts and iterations, then brief production.
If you’d like this sized against your own account, our Growth Audit covers creative volume as part of its 90-day plan.
