Facebook ad structures post-Andromeda are the campaign architectures (test and scale split, single CBO multi-adset, single CBO winning adset, single CBO single adset) that align with Meta’s Andromeda algorithm update and GEM (the architecture that builds winning ad sequences instead of winning single ads). As of 2026, there is no one-size-fits-all structure. The right pick depends on budget, number of products, AOV (Average Order Value), industry, and KPIs. After spending over $10M for ecommerce clients, we built an internal account structure matrix to decide which framework fits each brand we onboard.
In this guide we break down the five Meta pillars that justify each structure, the four campaign setups that still work post-Andromeda, the pros and cons of each, and the exact budget x product matrix we use inside the agency.
TL;DR
– Post-Andromeda, Meta optimizes the winning ad sequence, not the winning ad. Turning off top-of-funnel ads breaks the entire sequence. – Four structures still work in 2026: test and scale split, single CBO multi-adset, single CBO winning adset, and single CBO single adset. – Use budget + product count to pick: low budget + 1 hero product favors single CBO single adset; high budget + 10+ products favors single CBO multi-adset per product or collection plus bid cap or cost cap. – Always pair scaling structures with bid cap or cost cap campaigns running simultaneously, especially during predictable high-performance windows. – Creative diversification inside the same adset is the new key to scale because GEM sequences ads automatically.
Why Andromeda Changed Facebook Ad Account Structure
The Andromeda update rewired how Meta delivers and learns from ads. Before Andromeda, you had to give the algorithm clarity by isolating angles into separate ad sets. After Andromeda, GEM builds the winning sequence inside a single consolidated ad set, so creative diversification and consolidation now drive scaling.
Five pillars define every structure decision we make:
- Meta’s business model: Meta sells impressions. Ads must be engaging, not just selling, or Meta will reallocate budget toward higher-engagement creatives even if their CPA is worse.
- Cost vs value: Meta bids by highest volume, lowest cost. It pushes budget toward the lowest CPA (Cost Per Acquisition), not the highest ROAS (Return on Ad Spend).
- Learning phase and consolidation: Learning happens at the ad set level and typically requires 50 conversion events per week. Below $10K to $20K monthly spend, consolidate into fewer ad sets so Meta learns faster. See Meta’s official guidance on the learning phase.
- Creative signals: Post-Andromeda, mixed angles and formats inside one ad set work because the system optimizes delivery per creative.
- Ad sequence: GEM optimizes the sequence. Turning off a low-ROAS top-of-funnel ad can collapse the whole account.
For the full breakdown, read our pillar on Meta’s Andromeda algorithm and what it means for ecommerce.
The 4 Facebook Ad Structures That Still Scale in 2026
1. Test and Scale Split (OG Setup)
One ABO (Ad Set Budget Optimization) campaign for testing creatives in batches, plus an Advantage+ Shopping Campaign with cost cap or bid cap for scaling. You can also use a CBO (Campaign Budget Optimization) for testing and Advantage+ Shopping plus cost cap for scaling.
Process: test angles, personas, landing pages, offers in ABO or CBO. Identify winners. Copy the post ID. Scale with cost cap or bid cap. When you move winners over, move the full sequence, not single ads.
– Pros: manual control, force budget into creatives Meta would otherwise starve. – Cons: goes against Andromeda because you optimize single ads, not sequences. More time and precision required.
2. Single CBO Multi-Adset
One CBO per product, per country, or per funnel. One ad set inside per batch, with statics, videos, and angles mixed freely.
– Pros: Andromeda-friendly, full-funnel budget distribution, less management time, avoids creative overlap. – Cons: Meta may push budget toward high-engagement creatives instead of true winners. You can end up with too many ad sets and lose focus.
3. Single CBO Winning Adset
A variation of the multi-adset setup. Once creatives prove themselves, duplicate all winners into a dedicated “winning ad set” inside the same campaign. This forces Meta to push budget toward proven sequences.
– Pros: Andromeda-friendly, isolates proven creatives, less management overhead. – Cons: potential audience overlap between the original ad set and the duplicated winning ad set, especially at high spend.
4. Single CBO Single Adset
One CBO per product or funnel with one ad set holding every creative. Meta limits ad sets to roughly 50 ads, though that ceiling is often soft.
– Pros: 100% sequence-driven, fully Andromeda-friendly, the simplest structure to manage, and naturally forces spend onto the best performers. – Cons: every new ad restarts the learning phase. In our testing, performance often improves anyway because Meta uses existing ad set data to find new audiences.
The Budget x Product Matrix We Use After $10M Spent
This is the decisional matrix our team applies during onboarding. Two variables: daily budget and number of products.
| Scenario | Budget | Products | Recommended Structure | |—|—|—|—| | Starter | Low | 1 hero product | Test and scale split OR single CBO single adset | | Growth | Medium | 1 hero product | Test and scale split + single CBO winning adset + bid cap or cost cap | | Aggressive | High | 1 hero product | Test ABO/CBO + scale cost cap, or single CBO (multi or winning adset) + bid cap | | Catalog Starter | Under $500/day | 1 to 10 products | Single CBO single adset OR multi-adset, one campaign per product | | Catalog Growth | Medium | 1 to 10 products | Single CBO multi-adset or winning adset + bid cap or cost cap per top product | | Catalog Scale | High | 1 to 10 products | Single CBO multi-adset only (skip winning adset to avoid overlap) + bid cap or cost cap per product | | Fashion / SKU-heavy | Low to medium | 10+ products | Single CBO single adset per collection, then expand to multi-adset | | Fashion / SKU-heavy | High | 10+ products | Single CBO multi-adset per product or collection + bid cap or cost cap per collection |
For SKU-heavy brands, group ads by collection (for example, t-shirts, jeans, outerwear) instead of by single product. If you sell physical drops or limited releases, our framework on Meta ads strategy for fashion drops shows how this matrix adapts to short campaign windows.
Why You Can’t Scale: The Winning Ad Misconception
Most brands stuck at $20K, $30K, or $100K per month believe Meta spends most of the budget on the winning ad. It does not. Meta spends on the ad with the highest engagement and lowest CPA, which is usually a middle-of-funnel creative with a frequency above 2.
A real winning ad has three traits:
– Incremental attribution above target – Frequency below 1.5 – Highest possible spend at that frequency
In one client account we audited, the ad with the highest spend had a 4.38 frequency and a capped CPA. Meta kept pushing budget into it, but increasing the campaign budget did nothing because the audience was already saturated. Meanwhile, an ad with a 1.27 frequency at $7K spent and a $25 CPA was the actual growth driver, because it was bringing new top-of-funnel users who later converted on the higher-frequency middle-of-funnel ads.
The fix: stop scaling middle-of-funnel ads. Build more top-of-funnel creatives with frequencies below 1.5, even if their ROAS looks lower. They expand the lifespan of your high-frequency winners by feeding them new audiences. Meta’s own Performance 5 framework reinforces creative diversification as a primary lever.
This is the same principle behind our bid cap strategy for protecting ROAS at scale: control where Meta spends instead of letting it default to the lowest-CPA pocket.
How to Pick Your Structure in Under 60 Seconds
- Daily budget under $200 and one product? Single CBO single adset.
- Daily budget $200 to $1,000 and proven winners? Single CBO winning adset plus bid cap or cost cap.
- Daily budget $1,000+ and one product? Single CBO multi-adset plus bid cap or cost cap.
- 10+ SKUs at any budget? One CBO per collection, single or multi-adset depending on spend.
- Always pair scaling structures with cost cap or bid cap, especially for promotions, weekends, and Black Friday.
FAQ
About the Author
Antonio Ventre is the founder of Skaleit, a Meta ads agency for ecommerce brands. Skaleit has spent over $10M on Meta ads, scaling DTC brands across fashion, beauty, supplements, and home goods.
Want Skaleit to Build This System for Your Brand?
If you want our team to apply this matrix, audit your current structure, and rebuild your account post-Andromeda, book a call with Skaleit.