How many ads to put in a Meta Ads campaign

Right now everyone is dumping 30 or 40 creatives into a campaign at launch. The thinking behind it isn't dumb, the more material you feed the algo, the more options it has. Except in real life, it almost never plays out that way. I see eCommerce accounts where we shipped a ton of creatives all at once, and in the end nobody knows what's actually working.

I'm going to explain why I think most advertisers should do the opposite, start with a handful of creatives, and above all how to do it cleanly.

How Many Ads to Put in a Meta Ads Campaign

This is the real question everyone asks before launching an ad campaign. And the answer goes against what most advertisers do. You launch a campaign with 30 or 40 ads. You don't know what's going to work, that's normal, that's the whole point of testing. Except Meta is going to pick a small batch of creatives and ignore everything else. Out of your 40 ads, 4 or 5 grab the budget and the rest will never see a single euro. So you worked on thirty creatives nobody will ever see.

And there's something even nastier. The more you dilute your budget, the more you dilute your learning. If you've got 300 euros a day spread across 40 creatives, each ad gets crumbs. You end up with 20 clicks on one creative, 30 on another, 15 on a third. Analysing that afterwards and pulling out a real learning is a nightmare. You don't have enough data per creative to say for sure that this one works and that one doesn't.

Then there's quality. When you're obsessed with pushing out 40 creatives, you cut corners. Quantity takes over and the level of each ad drops. Two good creatives will always beat forty mediocre ones. You tick your "I produced a bunch of creatives" box, thinking volume will do the job on its own.

Last thing, reporting. With 40 ads in a campaign, your performance table becomes unreadable. You spend your time scrolling through rows of creatives that spent three euros. Whereas with a handful of creatives, you open your dashboard and in 30 seconds you know what to do.

The right move, start with 3 or 4 creatives max

What I recommend is starting with 3 or 4 creatives at launch. Sometimes even one or two if you're on Meta's newer creative workflows. And those creatives, you really polish them.

The idea isn't to make the same ad four times. You take different personas, you hit each of their pain points individually, and you vary the formats and hooks. That way you still get creative diversity, but on a small scale. You keep the testing principle, you just cut out the waste.

A concentrated budget learns faster

It's exactly the same logic as with international expansion. On Meta, the right answer is never to put 10 euros on 10 countries. It's to put 100 euros on a single country. With 100 euros concentrated, the algo can test plenty of targeting, plenty of placements, and it pulls out clean learnings. With 10 euros diluted everywhere, it learns nothing.

Same goes for creatives. 4 creatives each getting a decent budget for two weeks gives you real numbers. With 40 creatives scraping crumbs, you just get noise that's impossible to read.

The learning phase hates overload

The more variants you cram into an ad set, the longer you stretch the learning phase. The algo needs a certain volume of conversions per ad set to exit that phase and stabilise delivery. If your budget is split across 40 ads, each one only gets crumbs, the algo takes longer to figure out which ones to push, and delivery takes forever to stabilise. With a few well-funded creatives, you clear that hurdle much faster. It sounds trivial, but it's often what makes the difference between a campaign that takes off and one that stalls.

How to handle rotation afterwards

Once your 3 or 4 creatives are live, you let them run for a week or two. You watch what happens and you draw conclusions. If the results are good, you don't touch anything. I insist on this because a lot of brands have this reflex of swapping their ads every two weeks just to "refresh." That's a mistake.

If a creative has been performing for three months, you leave it. It's better to keep an ad that has proven it works for three months than to cut it for the sake of novelty. On these platforms, only results count. The algo couldn't care less about how fresh a creative is.

When a creative fatigues or starts underperforming, that's when you cut it and turn on a new one. How do you spot the wear? CPM climbing, click-through rate dropping, cost per purchase drifting. Those are your signals. But you know why you're cutting it and why you're activating the new one. That's when you're really steering the ship. Not a change set to the calendar every 15 days because that's the habit.

Test on the side, scale afterwards

I prefer to split into two campaigns, one pure acquisition, one testing on the side. You test your new creatives for a week or two in the testing campaign. The ones that work, you move into Advantage Plus and let them run. That way your acquisition is never disrupted by your tests, and your tests don't eat into your winners' budget.

And when you find one or two winning creatives that really perform, that's when you can scale budgets on them. You can even duplicate them for international. But first you find the recipe, then you scale.

What this actually changes

You might still produce 30 or 40 creatives total over time. The number of creatives itself isn't the problem. Dumping them all at once at launch is. Do it gradually, each time knowing what worked before and why you're adding the new one.

Start with few creatives and concentrate your budget on them, really polishing each ad. You'll add more later if you need to. And if you're producing just to produce, you might as well not bother.