The most common question we get from ecommerce brands is the same one every time. How do you actually test and scale Facebook ads? Should you test in ABO and scale in CBO, or test and scale in the same campaign? Cost cap, bid cap, or highest volume?
There is no single answer. The right structure depends on your budget, your offer, your SKUs, and how comfortable you are letting Meta make decisions for you. We spend $3M+/month across our client portfolio, and we run four different structures depending on the brand. This is the framework we use to pick which one.
TL;DR
There is no one-size-fits-all Meta ads structure. The right setup depends on whether you are willing to “waste” budget on testing in exchange for cleaner learnings (ABO), or whether you want Meta to optimize testing budget for you in exchange for biased learnings (CBO). After Andromeda, four structures work consistently across our client base: ABO test plus CBO scale, single-campaign single-adset test and scale, multi-adset CBO test and scale, and multi-adset CBO test plus single-adset CBO scale. We pick based on monthly ad spend, SKU count, and how much testing rigor the brand needs.
The Two Worlds: Separate Test/Scale vs. Test and Scale Together
Historically every media buyer ran a testing campaign and a separate scaling campaign. Test in ABO, find winners, duplicate into a CBO scale. That worked for years.
After Meta’s Andromeda update, a second approach started outperforming the classic split for a lot of our brands. Test and scale inside the same campaign. Either as a single CBO with one adset, or a single CBO with multiple adsets feeding the same scale.
Why the shift? Two reasons.
First, audience overlap. When you split into separate test and scale campaigns, your adsets compete with each other for the same audience. CPMs go up. Meta has been pushing consolidation hard since Andromeda for exactly this reason. Fewer adsets, more ads inside each one.
Second, learning consolidation. Meta now learns at the creative level, then pushes that learning up to the adset level. The more creatives sit inside one adset, the more data points Meta has to refine targeting. Spreading creatives across many adsets fragments that signal.
The trade-off: separate campaigns give you more flexibility to test specific angles or landing pages with clean variables. Consolidated campaigns give Meta more data to optimize, but you give up some control.
Complexity Scales With Budget
There is one thing nobody talks about that drives this decision more than anything else. Management complexity.
The more you spend, the more campaigns and adsets you end up running. An ABO setup at $50K/month might mean 20 to 50 adsets you have to monitor, edit, document, and revisit every few days. A CBO setup at the same spend might be 3 to 5 adsets where Meta handles most of the budget allocation.
That gap in the middle, the red arrow on the chart, is the time you reclaim by trusting Meta’s optimization. We use that time on creative production and offer building, which is where 80% of Meta ad performance actually lives.
The Decision Framework: ABO or CBO?
Here is the framework we walk every brand through before picking a structure.
The first question is whether you are willing to lose money on testing in exchange for cleaner data. If yes, ABO is the right tool. You launch one adset per concept, persona, or angle, with 4 to 6 ads inside each. You force-spend the testing budget and you learn, even when the learning is “this angle does not work.”
If no, you cannot afford to waste budget, then CBO is the answer. Meta forecasts which creatives are most likely to perform before spending a dollar on them, then allocates the budget accordingly. You keep more of your budget on winners. The trade is biased learning. Meta picks the creatives Meta wants to push, which usually correlates with engagement, not always with conversion.
90% of the time Meta is right. 10% of the time you miss winners that would have converted at higher CPMs but better conversion rates.
The 4 Structures We Run for Clients
Across our portfolio, four structures cover almost every scenario. Each one fits a specific budget level and risk profile.
Structure 1: ABO Test + CBO Scale
Classic separation. ABO testing campaign with one adset per concept, persona, or angle, 4 to 6 creatives in each. Winners get duplicated via post ID into a single-adset CBO scale campaign.
Use this when you have the budget to genuinely test angles cleanly and you want maximum control over what gets scaled. Run all of it on one bidding strategy. Either everything in highest volume, or everything in cost cap or bid cap. Mixing bidding strategies inside the same logical setup usually breaks things.
Structure 2: 1 Campaign, 1 Adset, Test & Scale
The simplest possible structure. One CBO campaign, one adset, all your creatives inside.
This works very well for brands spending up to $1K to $2K/day, especially single-product brands or brands launching one product at a time. You can also duplicate this structure per product, so each product gets its own dedicated test-and-scale campaign with its own adset and 50 to 60 creatives.
The risk: Meta might turn off creatives with low ROAS that are actually driving incrementality. Read the data carefully and look beyond pure ROAS at the ad level.
Structure 3: 1 Campaign, Multiple Adsets CBO
One CBO campaign, multiple adsets, each adset holds a creative batch grouped by angle, persona, or concept. Meta distributes the budget across adsets and you scale by lifting the campaign budget.
Split adsets by angle and concept, not randomly. Even though Andromeda has made mixed-angle adsets perform surprisingly well, giving Meta cleaner audience signals at the adset level still helps it consolidate learning. Use this when you have enough creative volume to populate 3 to 5 angle-based adsets and you want CBO to handle budget allocation.
Structure 4: Multi-Adset CBO Test + Single-Adset CBO Scale
Same as Structure 3, but with a separate single-adset scaling campaign sitting on top. The scale campaign holds the post IDs of all winning ads from the test CBO. You get more control over scale spend without losing the consolidated testing benefits.
Use this when you are spending $5K+/day and want explicit control over how much budget pushes into proven winners. The scale campaign here is also where you choose your bidding strategy. Highest volume for aggressive growth, cost cap or bid cap when you need stricter ROAS guardrails. The more you spend, the more we lean toward bid cap on the scale campaign.
The Engagement Rate Ranking Metric Almost Nobody Watches
One quick aside that matters more than people realize. Meta exposes a metric called Engagement Rate Ranking inside Ads Manager (custom columns → engagement rate ranking). It tells you how Meta scores the user-experience of each ad: below average, average, or above average.
Above-average ads get pushed harder, get lower CPMs, and reach more people. Below-average ads get throttled. If you are running CBO and wondering why Meta is not spending on your “obviously better” creative, check engagement ranking first. Often the under-spent ad is below average and Meta is protecting platform experience over your conversion rate.
This is also why CBO sometimes hides winners. An ad with a higher CPM but a much higher conversion rate can outperform on net ROAS, but Meta will not push it because user experience is mediocre. ABO is the only way to force-spend on those creatives and find out.
How to Pick Your Structure
A simple decision tree from our experience:
- Spending under $2K/day, single product: Structure 2 (1 campaign, 1 adset, test and scale together)
- Spending $2K to $5K/day, multiple SKUs: Structure 3 (multi-adset CBO, scale via budget)
- Spending $5K+/day, want testing rigor: Structure 1 (ABO test, CBO scale)
- Spending $5K+/day, want consolidated learning + control: Structure 4 (multi-adset CBO test + single-adset CBO scale)
These are starting points, not rules. We have brands that defy every one of them because their offer or SKU mix changes the math.
The Bottom Line
There is no single best Meta ads structure. Anyone selling you “the one setup that works” is selling you their setup, which works for their business or maybe for the few accounts they have run. We have run 70+ brands across every vertical and spend level, and we use four different structures depending on the brand.
The right question is not “ABO or CBO.” It is: how much testing rigor do I need, and how much complexity can I manage at my current spend? Answer those two and the structure picks itself.
Want Help Picking the Right Structure for Your Brand?
This is exactly what we do at Skaleit. Paid Meta ads management, campaign structure, and creative strategy for ecommerce brands scaling profitably. Get in touch here and we will audit your current setup and recommend which of the four structures fits your spend, your offer, and your team.
FAQ
Should I test in ABO or CBO for my ecommerce brand? If you can afford to spend $200 to $500 per adset on testing without immediate ROAS pressure, ABO gives you cleaner learnings. If you cannot afford to lose budget on bad creatives, CBO lets Meta forecast and protect your spend. Most brands under $2K/day should default to CBO; brands over $5K/day with creative testing rigor should consider ABO.
Does CBO actually work for testing, or is it only for scaling? CBO works extremely well for testing post-Andromeda. Meta forecasts likely-winning creatives before spending and allocates budget accordingly, which is exactly what you want when you cannot afford to waste budget. The trade is that you lose visibility into creatives Meta refuses to push, even if some of them would convert.
How many ads should I put in a single adset? For testing-focused adsets, 4 to 6 ads grouped by angle or concept. For consolidated test-and-scale adsets, 10 to 15 active ads. For single-product CBO test-and-scale campaigns, you can push 50 to 60 creatives if your daily budget supports it ($1K+/day minimum).
Should I use cost cap, bid cap, or highest volume for scaling? Highest volume is the default and works for most brands. Cost cap and bid cap give stricter ROAS guardrails when you spend $5K+/day and need predictable returns. Bid cap is the tighter of the two. Whichever you pick, do not mix bidding strategies inside the same logical test-and-scale flow, it confuses delivery.
What is Engagement Rate Ranking and why does it matter? It is a Meta metric showing how user-friendly your ad is judged to be (below average, average, above average). Above-average ads get lower CPMs and more reach. Below-average ads get throttled. If a CBO is starving your “best” creative, check this metric first.
How do I know when to switch from ABO to CBO? When the management overhead of running 20+ testing adsets starts eating more of your team’s time than the testing rigor is worth. Most brands cross that threshold somewhere between $5K and $10K/day in spend, but it depends on team size and creative volume.