Meta Advantage+ or manual campaigns: which should you run?

What Advantage+ means today

Advantage+ shopping campaigns have been renamed Advantage+ sales campaigns (5). More important than the name, Meta no longer makes you choose between an Advantage+ campaign and a manual one. Advantage+ now runs inside one setup that keeps the full range of ad features (1).

A sales campaign counts as Advantage+ ON only when it uses three features together (6):

• Advantage+ campaign budget
• Advantage+ audience
• Advantage+ placements

So the real question is no longer which campaign type to pick. It is which of these levers you hand to the algorithm and which you keep. If you manage campaigns through tools connected to the API, note that since Marketing API v24.0, released on 8 October 2025, new ASC campaigns can no longer be created through the legacy APIs (7).

Automation is also central to Meta's ad business. Its CEO said the annual revenue run rate for its end-to-end AI-powered ad solutions, including the Advantage+ suite, had passed $60 billion (8). It is better to learn to use it on your terms than to avoid it.

What head-to-head tests found

The most direct comparison available comes from Haus, which ran tests pitting manual campaigns against Advantage+. Manual campaigns delivered a higher incremental ROAS (iROAS) for 58% of brands (2). That is a majority, not a verdict. It also means a large share of brands did not get that result.

Timing matters too. Advantage+ did 9% better than manual campaigns on average at the experiment midpoint, but 12% worse by the end (9). A test you read halfway through can point you the wrong way.

On new versus existing customers, the gap was smaller than many owners assume. Advantage+ drove a slightly higher share of its impact on new customers than manual setups, 70% versus 65%, and the share of spend on retargeting was essentially flat at about 15% (10).

One caveat: these findings do not account for recent personalisation enhancements to Andromeda, Meta's AI ad-recommendation model (11). Treat them as a reason to run a test on your own account, not as a fixed rule.

The factors that decide it for your store

Whether Advantage+ or manual wins varies by brand, and it depends on three things (3):

• How long you have been spending on Meta
• How long your customers think before buying
• Your price point

The research does not say which way each factor tips the result. Use these factors to decide what to test, not as a reason to skip the test. If your store differs sharply from another on any of them, a result that held for that store may not hold for yours.

How to stop Advantage+ spending your budget on existing customers

Meta retired the Existing Customer Budget Cap at the same time as the ASC API (12), so the old setting is gone. Meta's replacement is a split inside one campaign (4):

• One ad set targets a custom audience of your existing customers, with a daily spend cap.
• A second ad set excludes that same audience.

For the split to hold, the existing-customer audience must not be treated as a suggestion the algorithm can widen, so set targeting_relaxation_types to 0 for custom audiences (13). Without that setting, the algorithm can treat the audience as a suggestion and widen it, which undoes the split.

Keep your customer list up to date. An exclusion only covers the people who are actually in the audience.

The control you keep inside Advantage+

Automation does not mean giving up cost control. Advantage+ campaigns still support bid strategies that give you some control, including cost cap, bid cap and minimum ROAS (14).

• Cost cap: suits you when you know the CAC your margin can carry.
• Bid cap: suits you when you want a hard limit on what you pay in each auction.
• Minimum ROAS: suits you when you have a clear return floor per order.

Because Advantage+ ON status requires all three features together (6), and the single setup keeps the full range of ad features (1), you can also keep one lever manual, such as placements, when you have a clear reason to.

How to judge the test

Judge on incremental results, not on the ROAS the platform reports. The Haus comparisons measured incremental ROAS (2), which asks what sales a campaign actually added, not which sales it can claim. If you are unsure which number should run your store, read the difference between MER, ROAS, POAS and ROI.

Run the test to the end. The midpoint and the end of the same experiments told opposite stories (9). Before you call a winner, also check how much of your spend is reaching existing customers. Spend on people who would have bought anyway can make any setup look better than it is.

Sources

1. (1) Meta for Developers · 2025-06-03
2. (2) Haus · 2025-07-28
3. (3) AdExchanger · 2025-10-29
4. (4) Meta for Developers · 2026-05-21
5. (5) Social Media Today (Industry Dive) · 2025-02-06
6. (6) Meta for Developers · 2025-06-03
7. (7) PPC Land · 2025-10-26
8. (8) Marketing Dive (Industry Dive) · 2025-10-30
9. (9) Haus · 2025-07-28
10. (10) Haus · 2025-07-28
11. (11) AdExchanger · 2025-10-29
12. (12) Meta for Developers · 2025-06-03
13. (13) Meta for Developers · 2026-05-21
14. (14) PPC Land · 2025-10-26

Read also

• How to Run Structured Creative Testing on Meta and TikTok