Meta Advantage+ Sales Campaigns in 2026: Setup, Scaling, and Limits

Advantage+ Shopping Campaigns got renamed. Somewhere along the way, in Meta’s push to make automation the default, ASC became Advantage+ Sales, picked up lead generation and app objectives alongside ecommerce, and quietly became the recommended way to run most conversion spend on the platform. If you set up your playbook in 2023 and have been running on muscle memory since, 2026 ASC is a different product sitting under a familiar name.

This guide covers how Advantage+ Sales campaigns actually work now, the prerequisites that decide whether automation helps or flails, a setup walkthrough, budget scaling that does not reset learning, and the honest list of situations where you should keep manual control. No vendor pitch, just the operating manual I wish Meta shipped.

What Advantage+ Sales campaigns automate

An Advantage+ Sales campaign hands Meta control of four things you used to set by hand:

• Audience. No interest stacks, no lookalikes, no age and gender splits. The system finds buyers from conversion signals and from what your creative communicates. You provide geography and, optionally, an existing-customer definition.
• Placement. Delivery across Facebook, Instagram, Audience Network, and everything else, allocated dynamically.
• Budget allocation. Campaign-level budget flows toward whatever combination of creative and audience is converting right now.
• Creative assembly. Dynamic optimization mixes your uploaded assets, adjusts copy and visuals per person, and increasingly applies AI enhancements to them.

Under the hood, this automation runs on Meta’s Andromeda retrieval engine, which reads the creative itself to decide who should see it. That architecture is why ASC lives or dies on creative input, and why everything in this guide keeps returning to that point. I wrote a separate breakdown of how Andromeda changed Meta ads delivery if you want the mechanics.

Recent additions worth knowing: ad set limits inside Advantage+ structures loosened, and Meta added an opportunity score, a 0 to 100 rating in Ads Manager that grades your account setup and suggests changes. Treat the score as a checklist of Meta’s preferences rather than a performance guarantee. Some suggestions are solid, like fixing signal gaps. Others, like “raise your budget,” deserve the skepticism you would give any suggestion that spends your money.

Prerequisites: when ASC works and when it cannot

Automation amplifies the account you already have. Before moving spend into Advantage+ Sales, check four boxes:

• Conversion volume. Around 50 conversions a week on the optimized event is the working floor. Below that, the system learns slowly and erratically. If purchases are too thin, optimize for a higher-funnel event with real volume first.
• Signal quality. Conversions API running alongside the pixel, full event coverage from view to purchase, deduplication verified. The automation is only as good as the data feeding it.
• Creative depth. Multiple genuinely different concepts, in multiple formats. Feeding ASC one hero video and four crops of it starves the system. My creative diversity playbook covers what “different” has to mean for this to work.
• Budget adequacy. Enough daily spend to generate learning data across your creative set. Thirty to fifty dollars a day per campaign is a realistic floor for small accounts; a dollar-a-day-per-concept mindset undershoots badly.

Miss two or more of these and a manual structure will usually outperform ASC until you fix the foundations. That is not a knock on the automation. It is how learning systems behave when starved of data.

Setup, step by step

1. Choose the sales objective and accept the Advantage+ recommended settings when offered.
2. Set geography and language. Resist adding more constraints. Every restriction shrinks what the system can explore.
3. Define existing customers via your customer list, then set the existing-customer budget cap. This is the one meaningful audience control you retain, and it is worth using: without it, ASC happily spends prospecting budget on people who were buying anyway. Note the known limitation that exclusions beyond this are minimal, and demographic controls are gone.
4. Upload creative in depth. Five to ten distinct concepts, mixed formats, plus catalog connection if you sell physical products. Turn on the Advantage+ creative enhancements selectively; review what each one does to your brand presentation before granting it everything.
5. Set campaign budget at a level that can feed every concept you uploaded. Underfunded breadth is the most common self-inflicted ASC wound.
6. Confirm tracking end to end with a test event before launch, not after the first confusing week of data.

Then leave it alone for a week. The urge to tinker on day three is strong, and each structural edit re-enters learning. Judge nothing before seven days and meaningful conversion volume.

Scaling without breaking learning

Budget moves of 10 to 20% every few days are the safe band. Doubling budget overnight routinely knocks a stable campaign back into volatile delivery for days. The rhythm that works: raise by 15%, hold three or four days, confirm cost per result held, repeat. Boring, and it compounds.

Scaling pressure also lands on creative. More budget concentrates more delivery on your winners, and post-Andromeda that burns them out faster. Plan for creative refresh to accelerate as spend grows; the thresholds and rotation cadences in my creative fatigue guide apply doubly inside ASC, where the system’s concentration behavior is at its most aggressive.

Creative strategy inside ASC

Because ASC removes most other levers, creative is where an operator still earns their keep, and there is real evidence about what works. The largest independent dataset on the Andromeda-era transition, Confect’s study of 115.7 billion impressions across 3,014 ecommerce advertisers, found that advertisers running 60 to 100% of spend through catalog ads achieved 44% higher ROAS and 68% lower CPA than those below 30% catalog share. Their top performers separated further through practices with big adoption gaps over the rest: design rules applied to catalog templates, multiple placement-specific formats, video catalog ads, five or more design variants per product, and enriched product assets.

Translated into an ASC feeding strategy for ecommerce:

• Connect the catalog and treat it as a creative surface, with designed templates, seasonal design rules, and clean product data, rather than raw feed images. The catalog is your largest supply of distinct, machine-readable ad units.
• Surround it with five to ten genuinely distinct non-catalog concepts: different personas, angles, formats, and message structures, so the system has ideas to match to people the catalog does not reach. What “distinct” means is precise now, and my diversity playbook covers the taxonomy.
• Check landing destinations. In the same dataset, ads sending traffic to category pages underperformed product-page destinations by 24% through the transition. Where the click lands is part of the creative system, and it is one of the few things the automation will never fix for you.

Troubleshooting the common ASC failures

Spend concentrates on one product or one ad. Working as designed, but audit whether that concentration is profitable concentration. If the favored asset is fatiguing, feed new concepts; the system rotates only when given somewhere to rotate to.

ROAS looks great, growth does not. Classic sign the campaign is harvesting returning customers. Check the new-versus-existing split and tighten the existing-customer budget cap. Attributed ROAS is a claim; the incrementality checks below are the trial.

Learning limited for weeks. Conversion volume is under the floor. Move optimization up-funnel to an event with volume, consolidate budget from parallel campaigns, or accept that the account is not ready for ASC yet.

Sudden delivery drops with no changes. Check creative fatigue flags first, tracking health second, account restrictions third. ASC’s concentration behavior means a single fatigued hero asset can read as a campaign-wide problem.

Volatile results in the first week. Normal. The system explores before it settles. Judge nothing before seven days and real conversion volume, and resist edits that reset the clock.

Measurement that keeps you honest

ASC reports look flattering by default, partly because the automation is genuinely good at finding easy conversions, including people close to buying anyway. Three habits separate real signal from self-congratulation:

• Segment new versus returning customers in every readout. A campaign printing cheap “purchases” that are 60% returning buyers is doing a different job than you think, and the existing-customer cap from setup is your lever.
• Run periodic holdout or lift tests where volume allows, rather than trusting attributed ROAS alone. Cost per incremental conversion is the number that survives scrutiny.
• Read breakdowns weekly: placement, creative, new-versus-existing. You cannot change much inside ASC, but breakdowns tell you what to feed it next, and creative is the input you fully control.

When to keep manual campaigns

The strongest accounts I see in 2026 run hybrids, and the practitioner consensus matches: trust the automation with consolidated, proven spend, and keep manual structures for the jobs it does badly. Keep manual control for:

• Creative testing. ASC allocates budget to likely winners, which is exactly what you do not want while testing, because challengers starve before producing a read. Test in a manual structure, then promote winners into ASC. My creative testing framework, publishing tomorrow, covers that pipeline.
• Genuinely niche or regulated targeting, where legal or practical constraints rule out broad delivery.
• Launches with no conversion history, where the automation has nothing to learn from and manual structure gives you steadier early data.
• Strict brand-control situations, if the creative enhancement options conflict with how your brand must appear. Review each enhancement toggle; they are individually controllable.

One practitioner line from a 2026 ASC roundup stuck with me because it matches my experience: trust the platform automation for adding budget to campaigns with established winners, and hold everything experimental outside it. Automation scales judgment. It does not replace it.

Frequently asked questions

Is Advantage+ Shopping the same as Advantage+ Sales?

Yes. Meta renamed Advantage+ Shopping Campaigns as the format expanded beyond ecommerce to leads and app objectives. Older articles and settings screenshots use the ASC name; the mechanics described here apply to both labels.

What budget do I need for Advantage+ Sales?

Enough to reach roughly 50 conversions a week on your optimized event, with 30 to 50 dollars a day as a practical floor for small accounts. Below that, learning is slow enough that manual structures usually serve you better.

Can I exclude existing customers?

You can define them via customer lists and cap the share of budget that reaches them, which in practice does most of what a hard exclusion did. Granular exclusions and demographic controls beyond that are largely unavailable inside ASC.

Why is my Advantage+ campaign spending unevenly across creatives?

By design. The system concentrates spend on whatever converts best right now. If one asset takes 80% of budget, that is the automation working, though it also means that asset will fatigue sooner. Keep feeding new concepts so the system always has somewhere to rotate.

Should beginners start with Advantage+ Sales?

If tracking is solid and there is some conversion history, ASC is a reasonable default precisely because it removes the settings a beginner would misconfigure. The skill that still has to come from you is creative strategy, and no toggle supplies it.

Sources: Meta Advantage+ product documentation and help center, Birch 2026 Advantage+ Sales guide, practitioner roundups from Top Growth Marketing and others, AdExchanger (January 2026), Meta Engineering (December 2024).