By the time you've watched five videos you're more confused than when you started. And probably a bit annoyed.

‍

So let us settle this.

‍

They're all right. And they're all wrong.

‍

After scaling over 250 ecom brands, many of them past 8 figures, we can tell you with complete confidence that there is NO single correct way to test creative on Meta.

‍

Anyone telling you there's one universal answer is either selling something or hasn't worked with enough accounts to know better.

‍

The right approach completely depends on where YOUR brand is right now. Your spend level, your margins, your appetite for risk, and whether you're building for next month or next year.

‍

It all comes down to one trade-off.

‍

Do you want cleaner data or faster results?

‍

Because you rarely get both at the same time.

‍

Here are the 5 ways we test creative across our client accounts and exactly when each one makes sense.

‍

1. ABO with one ad per ad set

‍

Best for: Brands that want complete control and the cleanest possible data.

‍

Run each ad in its own ad set with its own budget. Yes it costs more to test this way. But you'll know exactly which ads are performing and why. Every decision you make after that is based on real data, not Meta's guesswork.

‍

If you're spending serious money and you can afford to test properly, this is the gold standard.

‍

2. ABO with concepts or personas grouped

‍

Best for: Brands that want clean data but at the concept level rather than individual ad level.

‍

Instead of one ad per ad set, group 3-5 ads together by concept or persona. So an ad set might contain 5 ads all targeting the same pain point with different angles.

‍

You'll know which CONCEPTS are resonating without needing a separate ad set for every single ad. Cheaper than full ABO but you lose some granularity at the individual ad level.

‍

3. CBO with minimum spends

‍

Best for: Brands that want a bit more efficiency without losing all control.

‍

Run a CBO campaign but set minimum spend rules on each ad set. Every ad still gets a fair shot to prove itself but Meta can start shifting budget towards the early winners.

‍

You're letting the algorithm do some of the work but with guardrails so it doesn't blow your entire budget on one ad in the first 3 hours.

‍

4. Cost caps

‍

Best for: Brands with unpredictable performance who need to protect their downside.

‍

Set a strict cost cap so Meta only spends when it can hit your target CPA. You won't get volume but you also won't waste money on ads that aren't going to convert.

‍

This is the conservative approach. Slower but safer.

‍

5. Consolidate into prospecting / scaling

‍

Best for: Brands with tight budgets who just need to know what works fast.

‍

Throw new creative straight into your prospecting/scaling campaign and let it compete with your current winners. Most ads will die quickly but the ones that survive are genuinely strong AND you haven't spent loads on separate tests.

‍

It's brutal but it's efficient.