CBO on Meta in 2026 wins at low spend and loses at scale, and the crossover point is roughly five to seven thousand dollars per campaign per month depending on audience structure. Below that threshold, Meta's budget allocation across ad sets is better than most media buyers' manual allocation and consolidating at the campaign level buys cleaner learning. Above it, CBO starts to flatten performance by starving ad sets that need protected budget to escape learning phase, and ABO reclaims fifteen to thirty percent of scale that CBO leaves on the table. The rule is budget-tier dependent and most account audits get it backwards.
CBO wins below $5-7k per campaign per month because Meta allocates budget better than most buyers do. Above that threshold, CBO starves ad sets of the daily spend they need to escape learning phase and ABO reclaims 15-30 percent of lost scale. The decision is spend-tier dependent — audit the threshold in your account before defaulting.
What does CBO actually do differently from ABO?
CBO (Campaign Budget Optimisation) sets the budget at the campaign level and lets Meta's auction system allocate dynamically across ad sets based on real-time performance signal. ABO (Ad Set Budget Optimisation) sets a fixed daily budget per ad set and runs each one independently of the others. CBO was Meta's default from 2019 until the mid-2020s and remains the default in Advantage+ campaigns. The core difference in behaviour is that CBO can shift the entire campaign budget onto one ad set within hours if that ad set is winning, which is efficient at low spend and problematic at scale when it leaves other ad sets below the fifty-conversion-per-week learning threshold.
Why does CBO win below the five-to-seven-thousand-dollar threshold?
At low to mid campaign spend, the full campaign budget is barely enough to support three or four ad sets at the fifty-conversion learning threshold anyway. CBO's allocation efficiency wins because Meta's real-time read of auction signal is better than a buyer's weekly manual rebalance. The ad sets that would starve under ABO are the ones that were going to starve anyway, and CBO at least concentrates spend on the winner rather than spreading it thin across losers. Below the threshold, trying to run ABO forces the buyer to either underfund every ad set or over-split the campaign into too few ad sets to maintain testing velocity.
Why does CBO lose at scale above the threshold?
Above the threshold, the campaign budget is large enough that each individual ad set could clear fifty conversions per week on its own daily budget, but CBO's allocation prevents that from happening cleanly. The system shifts budget toward the near-term winner, which starves the losing ad sets below their learning threshold, which prevents those ad sets from escaping learning phase, which prevents them from ever being optimised enough to compete, which creates a feedback loop where CBO perpetually underfunds anything that is not already winning. ABO breaks the loop by guaranteeing each ad set the daily budget it needs to accumulate real learning signal independently. This is also why Advantage+ campaigns behave more cleanly above the €5k a month threshold — the automation assumes enough budget to support multiple ad sets at learning scale, which only holds above the tier.
What are the signs CBO is leaving scale on the table in your account?
1) Campaign spend above $5k per month with more than three active ad sets. 2) One ad set consuming more than sixty percent of campaign spend consistently. 3) Two or more ad sets perpetually in learning phase that never clear fifty conversions a week. 4) Overall campaign CPA stable but individual ad set CPAs invisible or unreliable because volume is too low to read. 5) Creative tests dying before they generate enough signal because CBO shifted budget away mid-test. 6) ROAS plateau at the campaign level despite audience expansion attempts. Any three of these in combination is a strong signal to run the ABO test.
How do you audit which structure wins in your account?
The audit is a twenty-eight-day paired test. Duplicate the existing CBO campaign as an ABO campaign with the same ad sets, creatives, and total daily budget split evenly across ad sets. Run both for twenty-eight days with the same optimisation event. At day twenty-eight, compare blended CPA (not platform-reported ROAS, which routinely overstates versus Shopify by thirty percent or more) across the two campaigns. If ABO beats CBO on blended CPA by more than ten percent, the account is above the threshold and ABO should become the default. If CBO wins or ties, stay on CBO until spend grows. Rerun the audit quarterly as spend scales.
When does CBO still win above the threshold?
CBO still wins above the threshold in two specific cases: when the campaign has only two or three ad sets (not enough to need protection, and CBO's efficiency compounds), and when the account is early in a scaling phase where no single ad set has yet established clear winners (CBO's real-time allocation finds the winner faster than a buyer's manual rebalance). Both exceptions compress as the campaign structure matures. By the time the account has four or more ad sets and a clear performance hierarchy, ABO starts to win regardless of CBO's historical efficiency.
Conclusion
CBO versus ABO is a spend-tier decision, not a philosophical one. Below the five-to-seven-thousand-dollar monthly campaign threshold, CBO wins cleanly. Above it, ABO reclaims scale that CBO leaves on the table by protecting ad set learning. The practical next step for any team currently defaulting to CBO at campaign spend above the threshold is to split a single CBO campaign into two ABO equivalents, run for twenty-eight days, and compare blended CPA. In Advize's experience, roughly six in ten accounts above the threshold see measurable gains from the switch.