CBO vs ABO: which should I use for Facebook ads?
Use ABO when you need each ad set to spend a guaranteed amount, which is what testing requires, and CBO when you want Meta to allocate one campaign budget toward whatever is converting, which suits scaling proven structures. The common working pattern is ABO for tests, then promoting winners into a CBO campaign. Neither is universally better; they distribute control differently.
Last updated 2026-08-11
What actually differs
With ABO, ad set budget optimisation, each ad set carries its own budget and spends it regardless of how siblings perform. With CBO, campaign budget optimisation, the budget lives at the campaign level and Meta continuously shifts spend toward the ad sets its models expect to convert cheapest. The difference is who decides distribution: you in ABO, the delivery system in CBO. Everything else follows from that. CBO's allocation is also not gently even; it will happily send nearly all spend to one favoured ad set, which is exactly what you want when scaling and exactly what ruins a controlled comparison.
Why testing wants ABO
A test's value depends on every arm getting enough spend to be judged, and CBO does not owe your test that. It allocates on early signals, so an audience or creative that starts slowly gets starved before producing a readable result, and you end the test knowing which arm Meta favoured in hour six rather than which arm is actually best. ABO guarantees each ad set its budget, so every arm accumulates decision-grade data. The cost is inefficiency, since you knowingly fund losers to completion, but that inefficiency is the price of information. Minimum-spend rules inside CBO partially imitate this, but plain ABO is the simpler and more predictable instrument for tests.
Why scaling wants CBO
Once you hold proven audiences and creatives, the question flips from which is best to how to extract the most from a budget across them, and that is CBO's home ground. Meta reallocates faster than any human checking dashboards, follows intra-week demand shifts automatically, and raising one campaign budget is a gentler scaling motion than editing several ad set budgets, each of which is its own learning-phase event. CBO campaigns also degrade more gracefully as they scale, since spend flows around a weakening ad set rather than being trapped in it. The requirement is trust: every ad set in a CBO campaign should be something you are content to see receive most of the budget.
The hybrid pattern most accounts settle on
Run a standing ABO test campaign where new creatives and audiences get guaranteed spend and clean reads, and one or a few CBO scaling campaigns holding the proven winners. Promotion is deliberate: a creative that passes your thresholds in the test campaign gets duplicated into the scaling campaign, ideally carrying its post id so the accumulated social proof travels with it. This separation keeps each mechanism doing what it is good at, and it makes reporting legible, because the test campaign's mediocre blended numbers are expected and quarantined rather than diluting the scaling campaign's economics.
Mistakes that break each mode
The classic CBO mistake is stuffing the campaign with many small or overlapping ad sets and then fighting the allocation with heavy minimum-spend constraints, which recreates ABO with extra steps and confuses delivery. CBO wants few, distinct, individually viable ad sets. The classic ABO mistakes are spreading budget so thin per ad set that nothing exits learning, and manually reallocating between ad sets so often that each adjustment resets stability. In both modes, switching an existing campaign between CBO and ABO is a significant structural change; do it at a natural break in the campaign's life, not mid-test or mid-scale.
How to decide in one minute
Ask what you would do if one ad set consumed 90% of the budget tomorrow. If the answer is that this would ruin what you are trying to learn, you are testing, use ABO. If the answer is that this would be fine provided results held, you are scaling, use CBO. Ask also how much structure you genuinely need: accounts running broad targeting with strong creative increasingly need fewer ad sets, which shrinks the allocation question and with it the stakes of this whole choice. The fewer ad sets you run, the less CBO and ABO differ in practice.
Meta's campaign setup defaults keep evolving toward automated budget allocation, and the interface labels shift over time. The control-versus-allocation logic here is durable; the exact toggles and names may not match your Ads Manager this quarter.