How to Test & Scale Meta Ads: ABO First, CBO Later

One of the simplest shifts that’s had the biggest impact on our paid social results?

Testing with ABO. Scaling with CBO.

First, the quick definitions, because Meta keeps renaming these. ABO (ad set budget optimisation) means you set the budget at the ad set level, so each ad set spends exactly what you give it. CBO — which Meta now calls Advantage+ campaign budget in Ads Manager — sets one budget at the campaign level and lets Meta’s algorithm decide how to split it across ad sets based on predicted performance. Same two levers they’ve always been; only the labels have moved.

Testing in ABO gives you control

When I run creative tests, I want each ad to get a fair shot. With ABO, I can push spend evenly across ads so nothing gets buried.

The problem with testing in CBO is simple – you’re handing the steering wheel to Meta far too early.

You might launch 100 new ads. Maybe 20 of them actually have potential. But Meta doesn’t know that yet.

Instead of waiting to see which ads convert, it tends to pick winners based on early signals like CTR or engagement. Not sales.

ABO vs CBO: when each actually wins

The two aren’t rivals — they’re tools for different jobs. ABO is a testing instrument: you’re buying clean, comparable data, so control matters more than efficiency. CBO is a scaling instrument: once you already know what works, you want Meta chasing the cheapest conversions across your proven set, and its automation genuinely does that well.

Factor	ABO (testing)	CBO / Advantage+ (scaling)
Budget control	You set spend per ad set — nothing gets starved	Meta reallocates across ad sets automatically
Best for	Clean creative and audience tests	Efficiency once winners are proven
Learning phase	Each ad set needs its own signal to stabilise	Shared budget can help ad sets reach volume faster
Risk	Slightly less efficient day to day	Meta may over-back one ad set on thin early data

What counts as a winning ad

“Focus on actual sales” only helps if you have a decision rule, so here’s ours. Don’t judge anything until an ad set has had enough spend and events to be reliable. Meta’s own guidance is that an ad set needs roughly 50 optimisation events over about 7 days to exit the learning phase and produce stable numbers — below that, performance is genuinely too volatile to trust (Meta Business Help Centre).

So a winner, for us, is an ad that:

• has cleared a meaningful spend threshold (enough to record several purchases, not one lucky sale)
• is delivering a ROAS comfortably above your break-even — not just a good CTR
• has generated enough purchases to be more than noise rather than a single conversion

If it only wins on clicks or engagement, it isn’t a winner. It’s a distraction.

The ABO-to-CBO scaling process, step by step

1. Test in ABO. Launch your creatives in ABO ad sets with enough budget per ad set to gather real signal. Give each ad room to prove itself — no guessing, no shortcuts.
2. Let the learning phase finish. Hold your nerve for the first several days. Judging an ad before it’s out of learning is how good creatives get killed early.
3. Identify what’s working. Look past vanity metrics. Apply the winning-ad rule above and shortlist only the ads clearing spend, purchases and break-even ROAS.
4. Move winners into CBO to scale. Build a CBO (Advantage+ campaign budget) campaign and bring your proven winners across. Duplicating the winning ads rather than rebuilding from scratch keeps the creative intact.
5. Start conservative, then ramp. Open CBO at a budget in the region of your combined test spend rather than a huge leap, then increase gradually — sharp jumps can tip ad sets back into learning. Once something’s proven, CBO helps maximise efficiency at scale.

That’s it. Don’t overcomplicate it.

Where ABO-first goes wrong

Even the right sequence gets botched in predictable ways:

• Moving winners too early. If you migrate before an ad set has cleared the learning phase, you’re scaling a guess. Wait for the signal.
• Too many ads per ad set. Cramming a dozen ads into one test ad set splits the budget so thinly that nothing gathers enough data to prove itself. Keep tests tight.
• Ignoring frequency. A winner today can fatigue fast at scale. Watch frequency and refresh creative before the numbers slide, rather than after.

For how this fits a full account, see our take on the only Meta ad account structure you need to scale profitably, and if you run mid- and bottom-of-funnel audiences, the hybrid CBO + ABO strategy for MOF and BOF.

If you’re relying on Meta to pick your winners too early, you’re leaving money on the table. Own the testing phase – then let CBO do the heavy lifting when it’s time to scale.

FAQ

Is ABO or CBO better for testing Meta ads? ABO is better for testing. Because you set the budget per ad set, every creative gets a fair, comparable amount of spend, so your winners are decided on sales data rather than on Meta’s early guesses. CBO (Advantage+ campaign budget) is better once you already know what works and want to scale it efficiently.

When should I move a winning ad from ABO to CBO? Once the ad has cleared the learning phase — roughly 50 optimisation events over about 7 days — and is beating your break-even ROAS on real purchases, not just clicks. Move proven winners only; scaling anything earlier means scaling an unproven guess.

What is the difference between ABO and CBO on Meta? ABO (ad set budget optimisation) sets the budget at the ad set level, so you control what each ad set spends. CBO — now called Advantage+ campaign budget in Ads Manager — sets one budget at the campaign level and lets Meta distribute it across ad sets automatically.

Post inspired by: Ben Radack