Why Is My Advantage+ Campaign Not Converting?

By Susan, Head of Paid Ads at ThriveX
Reading time: 6 minutes. Last updated: October 2026.

Four connected glass panels on a dark navy background with gold light flowing between them, one panel darkened with its connections severed, showing how pausing an Advantage+ ad set concentrates budget.

Author's Note

I used to think Advantage+ was a waste of ad budget. Then I started testing it, and my ad results improved. So when an Advantage+ campaign underperforms, I don't assume the automation is the problem. Meta documents most of what goes wrong openly, and much of it comes from habits carried over from manual campaigns that work against how Advantage+ distributes budget. This guide walks through the ones Meta names directly. If you want a direct read on whether the problem is the campaign or the page it sends traffic to, the ThriveX AI Audit checks this specifically.

The Automation Is Not the Problem, the Handling Usually Is

Most advertisers treat an underperforming Advantage+ campaign the way they would treat a manual one. They pause the weak ad sets, cap the spend on the expensive ones, and adjust settings as they go.

Every one of those moves is documented by Meta as something that degrades Advantage+ performance. Often the intervention is the reason the campaign is failing.

That matters commercially because the fix costs nothing. An advertiser doing all three is paying for a system whose main advantage they have switched off, and the money spent while it is switched off buys the same weak result the manual approach was already producing.

What Meta Says Breaks It

Meta's own best-practice documentation for Advantage+ campaign budget is unusually direct about which advertiser habits cause problems.

On pausing ad sets: "Advantage+ campaign budget reserves and spends your campaign budget on active ad sets, and pausing an ad set removes it from consideration. For example, if your campaign budget spend is for two ad sets, pausing one ad set would cause our system to focus on spending the budget entirely on the active ad set. By the time you unpause the other ad set and it starts to run, there might not be any budget left in the campaign."

On pausing weak performers specifically, Meta states that pausing a low delivery ad set "prevents Advantage+ campaign budget from finding tomorrow's opportunities" and that this "can ultimately cause an increase in cost per action (CPA) for your active ad sets."

On spend limits: "The more budget that's locked into specific ad sets, the less flexibility our delivery system has to optimise your campaign budget," with the added warning that combining spend limits with a cost or bid control "can cause your ad set to deliver less than it should, or not deliver at all."

Read together, those three describe the exact routine most advertisers follow when a campaign starts underperforming.

Where Advantage+ Campaigns Actually Break

Four patterns account for most underperformance.

The Ad Sets Get Paused and Unpaused

The most common one, and the most expensive. Pausing concentrates budget into whatever remains active, and by the time the paused set returns there may be no budget left for it to spend. The advertiser sees this as the automation ignoring their ad set.

Every Change Is Made One at a Time

Meta advises inputting changes in bulk, because entering them individually means the campaign "may have to re-enter the learning phase each time." An advertiser making daily single tweaks can keep a campaign in a permanent learning state, which is indistinguishable from the campaign simply not working.

Budget Follows Audience Size, Not Intent

Meta notes that if audiences differ significantly in size, "ad sets with the largest audiences will probably receive the most budget." An advertiser who expects spend to concentrate on their best-converting segment can find it concentrated on their broadest one instead.

Results Are Read at the Wrong Level

Meta advises analysing results at campaign level rather than ad set level, because the system deliberately moves budget between ad sets. Judging individual ad sets and reacting to them is what produces the pausing cycle in the first place.

The Creative Side Is Doing More Work Than the Settings

When the handling is correct and results are still weak, creative is the next place to look, and it carries more weight than most advertisers assume.

A report produced by Meta with AppsFlyer and Dentsu Creative, analysing 1.1 million creative variations, found that campaigns using a diverse set of creatives across formats and messaging types saw up to a 32% improvement in CPA and a 9% increase in incremental reach. The same report found advertisers in non-gaming categories increasing creative output 18% year over year, driven partly by the need to feed learning models.

Bar chart showing that campaigns using a diverse set of creatives across formats and messaging types achieved up to a 32% improvement in cost per acquisition and a 9% increase in incremental reach. — *Figure 1 demonstrates the performance gain from creative diversification found in Meta's joint research with AppsFlyer and Dentsu Creative across 1.1 million creative variations.*

Automation distributes budget across what you give it. Give it four near-identical creatives and it has nothing meaningful to distribute between.

Diagnostic Checklist

Your Advantage+ campaign is likely being held back by handling if:

You have paused and unpaused ad sets inside the campaign in the last two weeks
You have paused an ad set specifically because its delivery was low
Any ad set in the campaign has a spend limit applied
You have a cost or bid control applied alongside a spend limit
You have made single setting changes on separate days rather than in one batch
You judge performance by comparing ad sets against each other rather than at campaign level
Your creatives are variations of one concept rather than genuinely different approaches

Fix Sequence

Stop pausing and unpausing ad sets, and let the current cycle run without intervention
Remove ad set spend limits unless you have a hard requirement that justifies losing the flexibility
Batch any remaining setting changes into a single edit rather than spreading them across days
Switch your reporting view to campaign level, tracking total optimisation events and average cost per event
Add genuinely different creative concepts rather than more variations of the current one
Only after a clean cycle with no intervention, judge whether the campaign itself is underperforming

What This Check Looks Like in Practice

Say an Advantage+ campaign is spending its full budget but cost per purchase is well above what your older manual campaigns achieved. Here's how the checklist above separates a campaign problem from a handling problem.

Open the campaign's change history for the last two weeks. Count how many times ad sets were paused, unpaused, or edited one setting at a time. Each of those interrupts the delivery that Meta's own guidance tells you to leave running.

Next, check every ad set for spend limits. Limits reduce the flexibility Advantage+ uses to move budget toward what's working, which is the main reason to use it at all.

Then line up your creatives side by side. If they're crops or colour variants of the same photo, the campaign has one concept to work with, not several.

Once handling is cleaned up, let one full cycle run without intervention and judge the result at campaign level. Only if it still underperforms is the campaign itself the problem.

Consistency pays off in my own accounts too. On a Penang dental clinic account that started with zero data, keeping the budget steady instead of scaling it aggressively took CPM from RM25.35 in March 2026 to RM9.65 by May 2026, a 62% reduction in two months. That was a messaging campaign rather than Advantage+, but the lesson carries over: the delivery system gets better when it is given time and a stable budget to learn from.

Where the ThriveX AI Audit Fits

Once the handling issues are cleared and the campaign is still not converting, the cause has usually moved past the click. Automation controls who sees the ad; it has no influence on whether the page they land on can close them, which is the same reason more ad spend doesn't mean more conversions when the funnel is the constraint.

If clicks are arriving and sales are not, the checks in why your website gets traffic but no sales apply directly.

This is where the ThriveX AI Audit fits. ThriveX reviews whether the constraint sits in the campaign, the creative, or the destination the traffic arrives at, so the next change goes to the part that is actually losing the sale.

If your Advantage+ campaign spends its budget without converting, the ThriveX AI Audit is priced at $49 during beta, with standard pricing at $99.

FAQ

Should I pause underperforming ad sets in an Advantage+ campaign?
No. Meta's own documentation states that pausing a low delivery ad set prevents the system from finding future opportunities and can increase cost per action for your remaining active ad sets. Pausing also concentrates budget into whatever stays active, which can leave nothing for the paused set when you reactivate it.

Why is my budget going to the wrong ad set?
Meta notes that when audiences differ significantly in size, the ad sets with the largest audiences will probably receive the most budget. Budget distribution follows available opportunity rather than your judgement of which segment matters most, so a broad audience can absorb spend you expected elsewhere.

How do I know if the problem is the campaign or my landing page?
Compare impressions and clicks against conversions. If delivery is healthy and clicks are arriving but conversions are not, the loss is happening after the click, which no campaign setting will fix. If delivery itself is restricted, the cause is more likely spend limits, bid controls, or repeated learning phase re-entry.

Does adding more creative actually help?
Meta's research with AppsFlyer and Dentsu Creative found campaigns using genuinely diverse creative across formats and messaging types saw up to a 32% CPA improvement. Variations of a single concept do not count, since the system needs materially different options to distribute between.