TL;DR: Most brands spending serious money on Meta are producing a fraction of the creative their budget demands, and it’s quietly capping their growth. The algorithm needs a steady supply of fresh, diverse creative to keep finding winners, and systematic testing drives two to three times better performance than intuition-led launching. The right volume follows from your budget, your target CPA, and how the maths of testing works, not from guesswork. This is the equation we use.
The Question Behind the Question
“How many ads do we need?” sounds like a production question. It’s really a growth question. Creative volume sets the ceiling on how fast the algorithm can learn, how quickly you find winners, and how long you can scale before fatigue catches up. Underproduce, and you starve the engine. The campaign doesn’t fail dramatically. It just plateaus, and everyone blames the targeting.
The honest answer is that there’s no universal number. Ten ads a month is wild overproduction for one brand and negligent underproduction for another. The number depends on what you’re spending and what each acquisition costs you, so let’s build it from those two inputs.
Start With How Many Conversions an Ad Needs
Every ad that enters a testing campaign needs enough conversions to prove whether it works. Too few, and you’re reading noise. As a working benchmark, a creative needs roughly 5 to 10 conversions before the result means anything, fewer at lower spend and more as budgets grow and you can afford cleaner reads.
One caveat matters here. That 5-to-10 range holds for low and mid target CPAs, roughly up to £50 to £70. When your tCPA is high, waiting for 5 to 10 conversions on every test makes each one punishingly expensive, so dial the requirement down. On a £200 CPA product, two or three conversions per creative is often a sensible read; higher still, you may judge on fewer and lean on upper-funnel signals like thumb-stop and hold rate while the conversion data accumulates.
That single benchmark drives everything. If each test needs, say, 5.5 conversions to reach a verdict, and your target CPA is £83 (near the top of the standard band), then each test creative consumes about 5.5 × £83 ≈ £457 of testing budget before it’s earned a decision. Your testing budget divided by that figure is how many ads you can genuinely test in a month. Test more than that and you spread budget so thin that nothing reaches significance, which is the most common and most expensive volume mistake there is.
Split the Budget Before You Count the Ads
You don’t pour your whole budget into testing. You split it between testing (finding winners) and scaling (pouring spend into proven winners). As a rough guide rather than a rule set in stone, the split tracks how large your budget is relative to your CPA:
- If your monthly budget is around 500 × your target CPA or lower, you’re still in discovery, so weight toward 80% testing.
- If it’s around 2,500 × your target CPA or higher, you’re a scaling account, so drop toward 20% testing.
- Between those points, slide the testing share down proportionally, and we rarely let it fall below 10%.
The logic is intuitive once you see it. A smaller account hasn’t found its winners yet, so it should spend most of its money looking. A large, mature account already has winners and should spend most of its money pressing the advantage, while still reserving enough to keep the pipeline fresh.
The Full Equation
Put it together and the model looks like this:
Testing budget = Monthly budget × testing % Scaling budget = Monthly budget − testing budget Ads to test monthly = Testing budget ÷ (Target CPA × conversions-per-ad) Scaling ads = Scaling budget ÷ (Target CPA × 2)
Scaling ads need roughly twice the per-creative budget of test ads, because they’re carrying real spend rather than just earning a verdict.
A worked example
Take a B2C brand spending £10,000 a month with a £83 target CPA.
- 500 × £83 = £41,500. The £10k budget is well below that, so testing share = 80%.
- Testing budget = £8,000. Scaling budget = £2,000.
- Conversions needed per ad at this spend ≈ 5.5, so each test ad costs ≈ £83 × 5.5 = £457.
- Ads to test monthly = £8,000 ÷ £457 ≈ 17.
- Scaling ads = £2,000 ÷ (£83 × 2) ≈ 12.
That’s roughly 17 fresh test creatives every month, plus a working set of scaling ads, for a brand many would assume needs “a few ads.” If you’re spending £10k a month and shipping four creatives, you’re not testing. You’re guessing with a small sample.
Volume Without Diversity Is Just Noise
Here’s the trap. Once people see a number like 17, the instinct is to hit it the cheap way, taking one concept and making 17 near-identical variations. That defeats the entire purpose. The algorithm doesn’t want 17 versions of the same ad. It wants 17 genuinely different inputs to match against different people and placements.
So volume and diversity are the same instruction. Those 17 test ads should be spread across formats. For a B2C brand we’d split them roughly 30% hi-fi static, 30% low-fi video, 20% hi-fi video, and 20% low-fi static, which works out to about five statics, five low-fi videos, three produced videos, and three quick text-led pieces. (B2B weights more heavily to polished statics and produced video.) We unpack why that weighting works in our guide on creative diversity.
The point: produce to the number, but make the number diverse. Seventeen different doors, not one door seventeen times.
Do You Have the Infrastructure to Feed the Engine?
This is the question that actually decides whether volume helps you. Producing 17 diverse, on-brand creatives every single month, and the same again next month before this batch fatigues (the Creative Fatigue Score tells you when that moment is coming), is an operational capability rather than a one-off effort. Most brands hit a wall here not because they don’t know what to make, but because they have no system to make it at cadence.
That’s the difference between treating creative as a series of projects and treating it as a production line. We’ve written the full playbook for building that line in how to build a scalable creative strategy for e-commerce. The summary: map your required monthly volume against your real production capacity, and if there’s a gap, close it by staffing up, bringing in a partner, or systematising, because the cost of underproduction is always higher than the cost of production.
Where to Start Tomorrow
Take your monthly budget and your target CPA and run the numbers, or use our Creative Volume Calculator to do it in ten seconds and see the format breakdown. Compare the result with what you’re actually shipping. If you’re producing less than half the recommended volume, that gap is your single biggest, most controllable growth lever right now.
Then build the capacity to hit it consistently. Done well, volume gives a learning system enough to learn from rather than just flooding the feed, and most brands simply aren’t supplying that.
Sources
- ATTN Agency: Systematic Creative Testing. Analysis of 50,000+ creative tests across 200+ DTC brands showing a 2 to 3x performance advantage from systematic testing.
- SuperAds: Why Creative Diversity Is the #1 Performance Lever in 2026. Data on creative diversity outperforming volume-of-sameness, and the algorithm’s appetite for varied inputs.
- Ryze AI: Meta Ads Dynamic Creative Fixes 2026. Refresh-cadence data underpinning the need for sustained creative volume.
At Toco Marketing, we specialise in growth and marketing strategies that deliver measurable results. Want to know the exact creative volume your budget demands? Book a chat today, and let’s build a strategy that works for your business.