TL;DR
Advantage+ sales campaigns replaced Advantage+ Shopping Campaigns, and with the rename came a quiet structural change: the old percentage cap on spend toward existing customers is gone. Meta's own documentation says the cap is deprecated, and the replacement is on you: build two ad sets, one that includes your existing customer audience and one that excludes it. If you run a single blended ad set, the algorithm will drift toward the cheapest conversions available, and the cheapest conversions are people who already know you. Your prospecting ROAS will look great and your customer base will quietly stop growing.
The cap is gone. Nobody sent a memo.
Under the old ASC setup, you could cap how much of the budget went to existing customers. It was a blunt tool, but it did a real job: it forced a share of spend to go find strangers. In Advantage+ sales campaigns, that cap is deprecated. Meta's documentation is clear enough about the replacement (two ad sets, one with the audience, one without), but documentation is not a default. Defaults are what most accounts actually run, and the default now is one blended pool where the algorithm picks the easiest wins.
This matters because the algorithm is not being sneaky. It is being rational. Given a choice between a past buyer who converts at a high rate and a stranger who might convert, it picks the past buyer every time. That is literally what you optimised it to do. Blaming the machine for doing its job is a waste of a good rant. The structure is your job.
How the drift shows up in your reports
The symptoms are flattering, which is why they survive so long:
- ROAS climbs while new-customer revenue flatlines. The blended number looks healthy because repeat buyers carry it.
- Frequency rises on a small pool. The same warm audience sees the ad again and again, because that is where the cheap conversions live.
- Prospecting "stops working." It did not stop working. It stopped getting budget.
- Q4 makes it worse. Holiday traffic floods the warm pool with past buyers, cart abandoners, and subscribers. The drift accelerates exactly when you most need new buyers for January.
If you cannot split new vs returning revenue in your reporting, you cannot see any of this. That is step zero, and most accounts skip it.
The two ad set fix
Advantage+ sales campaigns now support multiple ad sets and custom audience exclusions, which is what makes this fix possible. Use them.
Ad set 1: prospecting. Exclude your existing customer audience (purchasers, and anyone else you count as already won). This ad set's job is new customers only. Judge it on new-customer cost and new-customer revenue, not blended ROAS. It will look worse than the blended number. That is the point: it is telling the truth.
Ad set 2: existing customers. Include your customer audience here. Retargeting, repeat purchase, cross-sell, win-back: this is its home. Judge it on repeat revenue and margin, not on whether it "beats" prospecting. Different jobs, different scorecards.
A few practical notes. Keep the exclusion list fresh; a stale customer list leaks past buyers back into prospecting. Do not duplicate the same creative across both ad sets and then wonder why delivery overlaps; tailor the message (strangers need the pitch, past buyers need the reason to come back). And do not set the prospecting budget from blended ROAS targets, or you will starve it the first time it reports honestly.
What to do with the old "it worked fine" objection
Some accounts will say their single blended campaign performs well, so why touch it? Fair question, wrong metric. A blended campaign can perform well for years while slowly converting your business from a growth engine into a retention machine with a shrinking top of funnel. You only notice when the warm pool runs dry, usually during a quiet quarter when there is no holiday traffic to mask it. Splitting the structure now is cheap insurance. Splitting it later, mid-slump, is a rebuild under pressure.
FAQ
What happened to the existing-customer budget cap in Advantage+ sales? Meta deprecated the old percentage cap. The documented replacement is two ad sets: one that includes the existing customer audience and one that excludes it.
Are Advantage+ sales campaigns the same as Advantage+ Shopping Campaigns? Yes, renamed. Advantage+ sales is the newer name, and it now supports multiple ad sets and custom audience exclusions, which the older ASC structure handled differently.
Why does my blended ROAS look good while growth stalls? Because the algorithm drifts spend toward past buyers, the cheapest conversions. Blended ROAS counts those wins without asking whether the buyer was new.
How should I judge the prospecting ad set? On new-customer cost and new-customer revenue, not blended ROAS. It will report lower than the blended number, and that honesty is the feature.
How often should I refresh the customer exclusion list? Continuously, and at minimum before any big push. A stale list leaks past buyers into prospecting and quietly undoes the split.
Does this apply to lead generation too? The same drift logic applies anywhere warm audiences convert cheaper than cold ones. If your goal is new leads, exclude existing customers and leads from prospecting, or the same story plays out.
MetaMaxd is a Meta ads agency. We run the accounts; you run the business. If your prospecting budget is buying the same customers twice, we will find it.
