Advantage+ sales campaigns (formerly Shopping): when they pay off and when they just inflate your ROAS

As of: 22 October 2026.

What Advantage+ sales campaigns are — in one sentence

Advantage+ sales campaigns (formerly Advantage+ Shopping Campaigns, ASC) are automated Meta campaigns with the Sales objective, in which you don't build audiences by hand: you upload creative and a budget, and the algorithm decides who sees the ad, where, and at what cost. According to Meta's help center, the strongest optimizations kick in when a sales campaign uses Advantage+ audience, placements, and budget. You'll still see the older name „Advantage+ Shopping” in guides and in this article's title, but the interface now says „Advantage+ sales campaigns”. The mechanics haven't changed: minimal manual settings and an algorithm that allocates the budget itself.

In short: you hand Meta the targeting, placements, and budget allocation in exchange for the promise of better performance with less work. It's the same bargain Google offers in Performance Max. As there, the biggest risk is a dashboard result that looks better than the actual one. The rest of this article is about that gap.

Why Advantage+ can inflate ROAS

Advantage+ inflates ROAS because the easiest path to a conversion the algorithm can see is people who were going to buy anyway: your existing customers and warm remarketing traffic. A campaign optimized for „as many purchases for as little as possible” naturally pulls toward that cheapest, safest traffic. A sale from someone who knows your brand and has already come back three times lands in the report with a high ROAS, even though it needed no ad at all. The result: the dashboard shows a beautiful number, and a large part of it is sales you'd have had regardless.

Then there's attribution. The default 7-day-click and 1-day-view window generously credits Advantage+ with purchases that started elsewhere: in search, a newsletter, a referral. The ad that „closed” the customer with a last impression gets full credit for the transaction. It's the same mechanic we described with Performance Max and inflated ROAS: the automation is only as good as the data and the counting you feed it.

This example doesn't mean you should drop Advantage+. It means you shouldn't make budget decisions on the raw number in Ads Manager. On the Gabo Wool account we managed the Meta budget through CRR, i.e. the ratio of ad cost to the revenue generated, instead of panel ROAS alone; year over year, CRR fell there from 18.28% to 11.15%. Separating existing-customer budget and running a lift test, both described below, turn that inflated figure into a result you can rely on. Whether your incremental ROAS covers your margin, you can check in the ROI calculator.

Existing-customer budget — what replaced the retired cap

Until recently, the most important setting in these campaigns was the existing-customer budget cap: it defined the maximum share of budget the campaign could spend on people from your customer lists. Meta has retired that setting and says you can get the same effect manually: with a campaign that spends nothing on existing customers, or one that splits budget between new and existing customers. The goal is unchanged: force the algorithm to chase incremental sales instead of the cheapest sale from a loyal customer. Without that split, the campaign will default to wherever conversion is easiest.

For the split to work, Meta needs to know who an „existing customer” is, so build a custom audience from an up-to-date customer list (uploaded or connected via CAPI) and use it as an exclusion or in a separate ad set with its own budget. A brand that wants to grow with new audiences puts a small share of budget on existing customers; if the main goal is retention and selling to your base, that share can be higher, but then you knowingly accept lower lift. Make the share your own decision. A setting left on autopilot hands it to the algorithm.

That last point is underrated: if your measurement leaks at the consent and cookie layer, customer lists and conversion signals reach Meta incomplete, and even a well-set exclusion works on patchy data. That's why airtight measurement (server-side tracking and proper server-side tagging) is a precondition for Advantage+ automation to get a true picture at all.

How to measure real lift, not dashboard ROAS

You can only measure lift from Advantage+ by comparing sales with the ad and without it. The dashboard ROAS won't show it, because by definition it claims everything it managed to touch. That's what incrementality tests are for: you hold out a control group that sees no ads and check how much more the exposed group sells. The difference is the lift, the number that tells you how much the campaign actually added to sales.

In practice you have three routes, from simplest to most credible. First, Meta's built-in Conversion Lift test: easy to run, but measured inside the same ecosystem that grades itself. Second, a geo holdout: turn the campaign off in selected regions and compare sales with regions where it runs. This test doesn't depend on Meta's attribution and gives a surprisingly readable result. Third, a periodic blackout: switch Advantage+ off for 1–2 weeks and watch how much total sales drop in the store (the drop in the panel doesn't matter here).

You don't need to do this weekly. One solid incrementality test per quarter, plus a constant view of your new-customer share, is enough to know whether Advantage+ drives growth or just books it nicely. If you want it all in one place next to your other channels, see our piece on the marketing dashboard. That's where you see whether marketing earns money.

Advantage+ vs manual campaigns — when each wins

Advantage+ sales campaigns win where you have a broad, clear offer, plenty of creative, and enough conversion volume for the algorithm to learn from; manual campaigns (with hand-picked audiences and ad-set-level budgets) win where you need control: a narrow niche, an expensive product, exclusions, a separate message for cold and warm audiences. You don't have to pick one type: the best accounts combine both, and the line is drawn by how much control you need and how much data you have.

A practical split that works on e-commerce accounts: Advantage+ as the main acquisition engine with existing customers excluded or split out, alongside a narrow manual campaign for the jobs automation won't handle well: a premium line, excluding recent buyers, a pure remarketing campaign measured separately. How exactly to split budget across channels, we laid out in our benchmark on how much to spend on Google Ads and Meta Ads.

How to structure budget and creative so you don't burn it

You set an Advantage+ sales campaign up well when you treat creative as the main lever and keep the structure simple: one campaign per goal, a budget large enough to get the algorithm out of the learning phase (per Meta's help center, usually after about 50 results a week per ad set), and a varied enough set of ads for it to test. Splitting this into many parallel Advantage+ campaigns usually backfires, because they compete for the same traffic, split the learning data, and drive up each other's costs.

Creative is the one area where you still have an edge over the algorithm, because it's the only thing Meta won't invent for you. A steady flow of fresh, varied assets now does more for Advantage+ results than tweaking settings; that's why producing ad creative and UGC content should be an ongoing process. And if you want help tying it all into one system, we're a marketing agency that runs Meta Ads campaigns daily.

FAQ: Advantage+ sales campaigns