- Advantage+ Shopping runs about 4.5x ROAS vs 3.7x manual, ~32% lower CPA, when signal is strong. [1]
- That ROAS is blended, not incremental. It over-credits retargeting and existing demand. [2]
- Threshold to use it: 25/week (reported, April 2026), 50 for dependable, 100+ for reliable. [3][2]
- You keep geography, a customer-mix cap, signal quality, and creative. You lose audiences, placements, and budget control. [3]
- Cap existing customers at ~25-30%. Keep one manual retargeting line as your measurement control.
Meta spent 2026 quietly removing the choice. Manual and Advantage+ flows merged into one interface. New Sales campaigns default to AI optimization. The weekly-conversion bar to qualify dropped to 25. The message to advertisers is not subtle: this is the road now. The scale behind the push is public: Advantage+ passed a $20 billion annual revenue run-rate in late 2024, growing 70% year over year. [6] And the performance case for taking it is genuinely strong, as long as you're honest about the number it hands back.
~3.7x ROAS. You control audiences, placements, budgets. And you do the work.
~4.5x ROAS, ~32% lower CPA. The AI controls almost everything. You control the inputs.
The lift is real and it compounds at scale: the CPA advantage runs near 38% above $10,000/month but only ~14% below $2,000/month, because the system needs volume to learn. [5] One caveat before you plan against any of these figures: they come from vendor benchmarks, not Meta's own data. None of that is the catch. The catch is what the 4.5x is measuring.
The number is blended. The decision needs the incremental one.
Advantage+ is very good at two things that flatter attribution: retargeting warm audiences and absorbing demand other channels created. Its reported ROAS folds those in. Meta's own framing ("17% more purchases per dollar" than manual) is a blended comparison, not an audited lift on net-new customers. [3] No surfaced source publishes a full incrementality methodology, so treat the headline as directional. [2]
You can get a rough floor yourself in 10 seconds.
Blended ROAS → a rough incremental floor
Strip out the existing-customer conversions that would likely have happened anyway. What's left is a rough floor for incremental ROAS on new customers, the number that actually justifies the spend.
Model: incremental ≈ blended × (1 − existing-customer share). A directional estimate only. It assumes existing-customer purchases are largely non-incremental. Validate with a geo or audience holdout before you act on it.
At a blended 4.5x with 40% of conversions coming from existing customers, the incremental floor is closer to 2.7x. Still fine for many brands. But a very different number to plan against, and the reason you keep a manual retargeting line: it's your measurement control, not just a campaign.
What you actually keep control of
You keep: geography and minimum age, an existing-customer budget cap (start 25-30%), conversion-signal quality (pixel + CAPI, Event Match Quality), creative inputs, and how many ASC vs manual campaigns you run.
You lose: detailed audience segmentation, placement exclusions, and creative-level budget allocation. Advantage+ treats interests and lookalikes as suggestions, not rules.
So the levers moved from targeting to inputs. The two that matter most: conversion density (get above the threshold before you scale) and the customer-mix cap (stop ASC quietly spending your budget on people who already own your product). UGC-style creative earns its place here too. Inside ASC it beats polished brand creative by roughly 48% on CTR and 26% on CPA in one widely cited vendor benchmark, and the system wants 15+ live ads to work with. [1]
The thresholds are worth memorizing: Meta's floor reportedly dropped to 25 conversions a week in April 2026, but 50 is where optimization gets dependable and 100+ where it's reliable. [3][2] Cap existing customers around 30% [4], and expect the CPA advantage to run near 38% above $10,000/month against only ~14% under $2,000. This system needs volume to earn its edge. [5]
How to run it in 2026
- Qualify first: get to 25+ weekly conversions with clean CAPI signal before ASC gets majority budget. Below that, fix signal and density, not budget.
- Cap existing customers at 25-30% so ASC hunts new demand instead of re-buying your own.
- Keep one manual retargeting campaign, for cart/checkout control and as your incrementality holdout.
- Consolidate: 1-2 ASC prospecting hubs, 6-8 active ad sets total. Over-segmentation is now a liability.
- Feed it creative: 15+ live ads, UGC-leaning, with catalog custom labels for the AI to sort.
- Judge it on incremental ROAS and new-customer share after the 7-10 day learning phase, not the blended headline. [3]
Advantage+ is a good system being handed a lot of trust by default. Take the performance. Refuse the framing. Run it to the threshold, cap the customer mix, keep a manual control line, and measure the number Meta doesn't put on the dashboard. The advertisers who lose here aren't the ones who adopted ASC. They're the ones who believed the first ROAS it showed them.
Sources
- Skale Strategy · Meta Advantage+ Shopping campaigns in 2026
- The Interconnections · Meta ads in 2026: what changed for e-commerce
- 1ClickReport · Advantage+ Shopping: the 25-conversion rule (2026)
- Alex Neiman · Meta Advantage+ Shopping campaigns guide
- Digital Applied · Facebook ads benchmarks 2026: CPC, CPM, CTR by industry
- AdExchanger · Q4: Meta minted money and improved its monetizationreporting Meta's Q4 2024 earnings: Advantage+ surpassed a $20 billion annual run-rate, up 70% year over year
Frequently asked questions
Is Advantage+ Shopping actually better than manual campaigns?
On blended numbers, usually yes, when signal and volume are strong. Independent 2026 benchmarks put Advantage+ Shopping around 4.5x ROAS versus roughly 3.7x for manual, with about 32% lower CPA, and the CPA advantage widens at scale (around 38% above $10,000/month, only ~14% below $2,000/month). But those are blended figures, not measured incremental lift on new customers.
What is the conversion threshold to use Advantage+?
Meta reportedly lowered the threshold to 25 conversions per week per campaign in April 2026. Practitioners still treat 50 purchase events per week as the point where optimization is dependable, and 100+ as reliable. Below 25, ASC tends to underperform. Fix conversion density and Conversions API signal quality before shifting majority budget.
How much control do I actually lose?
Most of the granular levers. Advantage+ treats interests and lookalikes as suggestions; only location and minimum age are hard constraints. You lose detailed audience segmentation, placement exclusions, and creative-level budget control. What you keep: geography, an existing-customer budget cap (start ~25-30%), conversion-signal quality, creative inputs, and how you structure ASC versus manual campaigns.
Why is the reported ROAS misleading?
Because Advantage+ is strong at retargeting and broad reach, it can attribute conversions that would have happened anyway: existing customers and demand already created by other channels. The blended ROAS looks great without proving incremental lift on net-new buyers. The only honest read comes from a holdout: a geo split, an audience split, or a time-based test.
What's a good 2026 account structure?
For $30,000 to $100,000/month, the common shape is one to two Advantage+ Shopping campaigns for broad prospecting plus a single manual retargeting campaign for cart and checkout, with no more than six to eight active ad sets in total. Smaller accounts start with one ASC campaign once they're consistently near 25 weekly conversions, and keep manual campaigns for control and measurement.
How do I measure the incremental ROAS that Meta doesn't show?
Run a holdout: a geo split, an audience split, or a time-based test that withholds ads from part of your market and compares outcomes. A quick floor is to strip existing-customer conversions out of the blended number, so a 4.5x blended ROAS with 40% of conversions from existing customers implies a rough incremental floor near 2.7x. Keep one manual retargeting line as your measurement control, not just a campaign.
What creative works best in Advantage+ Shopping?
UGC-style creative tends to outperform polished brand work inside ASC, beating it by roughly 48% on CTR and 26% on CPA in one widely cited vendor benchmark. The system also wants volume to sort through, so feed it 15 or more live ads plus catalog custom labels. Creative is one of the few real levers you keep, so treat it as an input worth optimizing.
Is it a mistake to run lots of narrow ad sets in Advantage+?
Yes, over-segmentation is now a liability rather than a safeguard. Advantage+ needs conversion volume to learn, so splitting spend across many narrow ad sets starves the system of signal. For most accounts the shape is one or two ASC prospecting hubs plus a single manual retargeting campaign, with no more than six to eight active ad sets in total.
Mikołaj Salecki
Writes about media, tech, and AI business for people who actually run digital. Former agency lead. Skeptic of frameworks that read better than they perform.
Get the next issue by email.
One letter, once a week. Sharp coverage of media, tech, and AI business. No filler.
