- Have found a winning ad in testing but are unsure how to best scale it further
- Mix testing and scaling campaigns in the same structure, without a clear transition between the two
- Want to understand the difference between ABO and CBO, and when to use which
- Find that scaling a winner often means rebuilding the campaign structure from scratch
ABO and CBO are not two competing ways of doing the same thing; they solve two different tasks in the ad's lifecycle.
During testing, ABO is ideal because you set a fixed budget for each ad set, giving you full control over how much each variant has to prove itself. This is crucial during testing because you want a clean, comparable picture of how each individual hypothesis performs in isolation.
Once a winner is found, the task changes. Now it's not about comparing variants under controlled conditions, but about giving the algorithm the opportunity to allocate budget dynamically to what performs best, and that is exactly what CBO is built for. CBO lets Meta distribute the total campaign budget between ad sets based on expected performance, instead of keeping it fixed per set.
Why simply increasing the budget in ABO is a mistake
A common mistake is to find a winner in an ABO test campaign and simply increase the budget in the same ad set, believing that scaling is now solved.
The problem is that ABO is not designed for that task. Without letting the algorithm dynamically distribute the budget between multiple ad sets, you lose the flexibility that makes scaling effective, and you risk simply forcing more budget into a structure that was built for something else.
The right approach is to let the winner graduate: it is moved into a new, dedicated scaling campaign built on CBO, where it can be combined with other proven winners and allowed to compete for budget dynamically.
Structuring this transition correctly requires a clear process that is the same every time a winner is found.
1. Name campaigns so that testing and scaling can never be confused
A test campaign should be clearly labeled as a test in its name, and a scaling campaign clearly as scaling, so there is never any doubt about which structure a given ad belongs to.
2. Define what qualifies a winner for graduation
Set a clear criterion, for example, proven performance above a given threshold, for when an ad is ready to be moved from the ABO test to a CBO scaling campaign.
3. Duplicate the ad into the scaling structure, keep the original
Instead of moving the ad and risking the loss of its accumulated data and learning, duplicate it into the new CBO campaign while the original test version remains active.
4. Let multiple winners compete for budget in the same CBO campaign
When multiple proven winners are gathered in the same scaling structure, CBO can actually do its job: allocate budget dynamically to those performing best in the given period.
CBO is not always the right structure. There are still situations where ABO is the best choice:
- You are still in an active testing phase and need a clean, controlled view of each variant
- You want to control exactly how much budget goes to a specific segment, regardless of performance
- The budget is low enough that CBO's dynamic distribution does not yet have sufficient data to work with
Outside of these situations, CBO is typically the strongest structure when a winner needs to be scaled.
Three signs that testing and scaling are not clearly separated in your account structure:
- Winning ads remain in the original test campaign, just with a higher budget
- The campaign naming does not make it clear whether something is in testing or scaling
- You find it difficult to see which ads are actually competing for the same scaling budget
If you recognize one or more of these, it is likely time to implement a fixed ABO-to-CBO graduation in your account structure.
We never let a winning ad remain in its original test campaign, just with a higher budget. We build a fixed graduation into the account structure: ABO for controlled testing, and CBO for scaling, where proven winners are consolidated and compete dynamically for the budget.
It is part of the same systematic logic found in our campaign naming standard, where it is always clear whether a given campaign is in testing or scaling. It is part of the approach we build into our Paid Social work through E-COM OS.
1. What is the difference between ABO and CBO?
ABO (Ad Set Budget Optimization) sets a fixed budget for each ad set. CBO (Campaign Budget Optimization) allows Meta to dynamically distribute the total campaign budget across ad sets based on expected performance.
2. Why is it a mistake to just increase the budget in an ABO test campaign when you find a winner?
Because ABO is not designed to dynamically distribute budget between multiple ad sets. You lose the flexibility that makes scaling effective if you simply force more budget into the test structure.
3. Should you move or duplicate a winning ad when scaling?
Duplicate. By duplicating the ad into the new CBO scaling campaign, you preserve the original test version's data and learning, while allowing the winner to compete for the scaling budget in the new structure.
4. Is CBO always better than ABO?
No. ABO is still the right choice during active testing, where you need a controlled, comparable view of each variant. CBO is strongest when scaling already proven winners.
5. How do you avoid confusing test and scaling campaigns?
By using a consistent naming convention where it is clearly stated in the campaign name whether something is in testing or scaling.
.png)
