For years, ABO vs CBO was a preference. Some media buyers trusted their own hands on the budget, others trusted the algorithm, and both could point at accounts that worked.

That framing stopped being useful in February 2026. Meta merged its manual and Advantage+ campaign flows into one setup, and AI-driven budget optimization is now switched on by default for new sales, app promotion, and lead campaigns. You did not lose the ability to control spend at the ad set level. You lost the neutral starting point.

So the question is no longer which one is better. It is when you override the default, and what you need to see before you stop overriding it. This post lays out a plain framework: test creative fairly with ABO, hand budget to Advantage+ once a winner has earned it, then read the results at the creative level, where the answer actually lives.

What Changed in Ads Manager

Three practical changes matter for budget strategy.

Campaign Budget Optimization is now called Advantage+ campaign budget. Same mechanic, new label: one budget at the campaign level, distributed across ad sets in real time toward whatever converts cheapest.

The old binary choice between a manual campaign and an Advantage+ campaign is gone. There is one creation flow, with individual toggles for audience, placements, creative, and budget.

Those toggles start in the on position. Turning off Advantage+ campaign budget gives you ad set level control, which is ABO. It still works exactly as it always did. It is simply now a deliberate act rather than a default state.

The Two Jobs a Budget Structure Has to Do

Your budget structure is doing one of two jobs at any moment, and it cannot do both at once.

The first job is producing a clean read. You want to know whether a hook works, whether a new angle converts, whether a format outperforms the one you have been running. That requires every contender to get comparable exposure, which is exactly what ABO guarantees and exactly what Advantage+ refuses to do.

The second job is allocating money quickly toward what already works. Advantage+ reallocates within hours. No media buyer checking dashboards once a day can match that, and trying to is a waste of your attention.

Most confusion about ABO vs CBO comes from asking one campaign to do both jobs. It will do neither well.

Phase 1: Test With ABO

Structure so only one thing varies

One campaign, one audience, four to five ad sets, one creative concept per ad set, equal budgets. If creative is the variable you are testing, keep targeting identical across every ad set. If audience is the variable, keep creative identical.

This sounds obvious and gets violated constantly, usually by testing a new hook against a new audience at the same time and having no way to attribute the result. Consistent ad naming conventions are what make the read survive into reporting a month later.

Size the budget against your target CPA

Round numbers are a bad default. Set each ad set at roughly twice your target CPA per day, so the test can actually produce conversions rather than impressions.

The reference point is Meta's learning phase: an ad set generally needs around 50 optimization events within seven days to stabilise. If your budget makes that arithmetic impossible, you are not running a test. You are collecting noise.

Let it run before you judge

Three to five days minimum, and no edits in between. Every budget change or pause restarts learning and throws away the data you were waiting for. The urge to intervene on day two is the single most expensive habit in creative testing.

The Graduation Test Most Accounts Skip

Here is the step that separates accounts that scale from accounts that stall.

A good CPA at $15 a day proves nothing about what happens at $150. Plenty of ad sets look like winners purely because the budget was boxed into them and had nowhere else to go.

Before you hand anything to the algorithm, raise the winning ad set's budget meaningfully and hold it for three to four days. If CPA holds, you have a real winner and it is ready for Advantage+. If it breaks immediately, that ad set was never a scaling candidate, and no amount of automation will change that.

Run this gate every time. It costs you four days and saves you from pouring campaign-level budget into something that only worked in a container.

Phase 2: Hand Budget to Advantage+

Duplicate into a fresh campaign

Do not flip your existing ABO campaign over to Advantage+ budget. Changing budget structure in place resets learning on everything inside it, and it destroys the testing campaign you spent a week building.

Duplicate the proven ad sets into a new campaign instead. Your testing structure stays clean and keeps running.

Group only comparable ad sets

Three to five ad sets is the practical ceiling. Beyond that, the weaker ones receive almost nothing and you are paying for a structure you cannot read.

More important than the count is that the ad sets are peers. Putting a proven winner next to an untested angle guarantees the algorithm funds the winner and starves the newcomer, which tells you nothing you did not already know. If two ad sets are not at comparable maturity, they belong in different campaigns.

Use a cost cap if you need a ceiling

Plain lowest-cost bidding hands Meta the wheel completely. A cost cap or bid cap lets the algorithm scale spend while you hold a hard limit on cost per result, which is usually the real answer for anyone hesitant about giving up control.

Ad set spend minimums exist and can be useful for the first week to guarantee initial exposure. Leave them on indefinitely and you are paying for automation while preventing it from working.

Reading the Winners After You Hand Over Control

There is a cost to letting Advantage+ allocate, and most guides skip it: your ad set report stops being a scoreboard.

When Meta pushes 70% of budget into one ad set, that ad set's numbers are not a verdict on its creative. They are a verdict on allocation. The creative in the starved ad sets was not judged and found wanting, it was simply never given the chance to be judged. Campaign-level reporting cannot tell those two situations apart, and if you treat spend share as a quality signal, you will retire creative that never actually failed.

What you need instead is performance read at the creative level, across campaigns, independent of how budget happened to distribute. Which hook is carrying the account. Whether the winning ad works because of the format or the offer. Which angles keep reappearing in your top performers.

That is the job Adnova's Creative Analytics does. It tags every creative across dimensions like hook, format, angle, and messaging, then connects those tags to ROAS and CPA, so you can see which creative variables drive returns rather than which ad set the algorithm happened to favour this week.

This matters more in 2026 than it did in 2023. As Meta absorbs targeting, placement, and budget decisions, creative is the last lever you fully control. Knowing precisely what is working inside it is no longer a reporting nicety.

Mistakes That Break the Framework

• Testing inside an Advantage+ budget. The algorithm picks a winner in the first few hours on almost no data, and your other creatives never run.
• Judging during the learning phase. Days one and two tell you nothing. Reacting to them costs you the days that would have.
• Mixing proven and unproven ad sets. The proven one absorbs the budget and the test is dead on arrival.
• Running ABO with a low creative hit rate. If most of your concepts are weak, equal spend just funds weak creative evenly. Fix the creative pipeline first.
• Expecting automation to rescue bad creative. Advantage+ scales what works. It does not manufacture a winner.
• Editing constantly. Every significant change restarts learning. Set the structure, then leave it alone for a week.

Conclusion

ABO vs CBO is not a strategy choice. It is a lifecycle. ABO exists to find out what is true, Advantage+ exists to press on what you have found, and the graduation test in the middle is what stops you from confusing the two.

In practice they run in parallel, permanently. A testing campaign on ad set budgets never switches off, because creative fatigues and the pipeline has to keep producing candidates. Proven ad sets graduate into Advantage+ campaigns as they earn it.

Adnova supports both halves of that loop. The Bulk Ad Launcher gets structured tests live in minutes instead of an afternoon of manual ad set duplication, and Creative Analytics tells you which creative variables are actually driving performance once Advantage+ is deciding where the money goes.

FAQs

1. Is CBO the same thing as Advantage+ campaign budget?

Yes. Meta rebranded Campaign Budget Optimization as Advantage+ campaign budget. The mechanic is unchanged: one budget set at the campaign level, distributed across ad sets in real time based on performance.

2. Can I still use ABO after the 2026 changes?

Yes. New campaigns default to Advantage+ campaign budget, but the toggle can be switched off, which returns budget control to the ad set level. That override is exactly what makes ABO useful for testing.

3. Can I run ABO and Advantage+ campaigns at the same time?

You should. A permanent ABO testing campaign feeding winners into Advantage+ scaling campaigns is the standard structure for accounts spending seriously. Keep them as separate campaigns so the testing budget cannot be reallocated away from your test.

4. How many ad sets should one Advantage+ campaign have?

Three to five, all at comparable maturity. More than that and weaker ad sets receive near zero spend, which means you are paying to run a structure you cannot draw conclusions from.

‍