Account structure is the foundation that everything else in Meta advertising sits on top of. You can have the best creatives, the sharpest copy, and the most compelling offer, but if your account structure is fragmented, overlapping, or misaligned with how Andromeda allocates budget and learns, your performance will suffer. In 2026, Meta's Andromeda algorithm penalizes structural fragmentation more aggressively than ever. This guide breaks down exactly how to structure your account for maximum ROAS, including the CBO vs. ABO decision, the 3-campaign framework, and the most common structural mistakes to avoid.
Why Account Structure Matters More Than Ever
Meta's Andromeda algorithm makes delivery decisions at the account level, not just the campaign level. It considers cross-campaign frequency, creative signal diversity across your entire account, audience overlap between ad sets, and budget distribution efficiency. An account with 20 campaigns, each with 3 ad sets targeting slightly different variations of the same audience, looks like chaos to Andromeda. The algorithm cannot efficiently learn and optimize when its signals are fragmented across too many competing elements.
The penalty for fragmentation is real and measurable. Fragmented accounts, those with many small campaigns and ad sets, consistently show higher CPAs, longer learning phases, and more volatile performance than consolidated accounts running the same total budget. The mechanism is straightforward: fragmentation divides your budget into pieces too small to generate the 50 conversions per ad set needed to exit the learning phase, creates audience overlap that makes your own ads compete against each other in the auction, and prevents Andromeda from building a coherent model of your best audience segments.
CBO Explained: Campaign Budget Optimization
CBO (Campaign Budget Optimization) means you set your budget at the campaign level, and Meta's algorithm decides how to distribute that budget across the ad sets within the campaign. The algorithm shifts budget toward ad sets that are producing better results, lower CPAs, higher ROAS, in real-time. You give up control over individual ad set budgets in exchange for Meta's algorithmic budget allocation.
The advantages of CBO are significant. It automatically identifies and funds your best-performing ad sets without manual intervention. It prevents budget waste on underperforming ad sets (the algorithm reduces their spend). It simplifies campaign management, you set one budget instead of managing multiple ad set budgets. And it aligns with how Andromeda wants to work: the algorithm performs best when it has maximum flexibility to allocate resources.
**The catch with CBO** is that the algorithm can be aggressive about budget concentration. It may allocate 70-80% of your campaign budget to a single ad set if that ad set is outperforming, even if the performance difference is marginal. This can starve other ad sets of the data they need to prove themselves. It can also lead to premature scaling of a "winner" that was simply first to get lucky conversions.
ABO Explained: Ad Set Budget Optimization
ABO (Ad Set Budget Optimization) means you set individual budgets for each ad set. Each ad set gets exactly the budget you assign, regardless of relative performance. You maintain full control over spend distribution, but you also take on the responsibility of reallocating budget manually based on performance data.
ABO is better for controlled testing because every ad set gets a fair, equal chance to prove itself. There is no algorithmic bias toward early winners. You can ensure specific audiences or creative concepts get tested with adequate budget. The trade-off is manual management overhead, you need to check performance daily and shift budgets yourself, which does not scale well as you add more ad sets.
When to Use CBO vs. ABO
The CBO vs. ABO decision is not philosophical, it is situational. **Use ABO for testing.** When you are testing new audiences, new creative concepts, or new offers and need each test variant to receive equal budget, ABO ensures a level playing field. Set each ad set to the same daily budget and let them run for 5-7 days before comparing results. **Use CBO for scaling.** Once you have identified winning ad sets through ABO testing, move them into a CBO campaign and let the algorithm allocate budget toward the best performers. CBO is most effective when the ad sets within the campaign have already proven they can convert, the algorithm is optimizing allocation among proven performers, not exploring unknowns.
**The hybrid approach** combines both: ABO campaigns for testing new elements, CBO campaigns for scaling proven winners. This gives you controlled testing and efficient scaling in the same account.
The 3-Campaign Framework
The most effective Meta ads account structure in 2026 uses three campaign types, each serving a distinct function.
**Campaign 1: Testing.** This is your ABO campaign where new creatives, audiences, and offers are tested. Budget: 20-30% of total ad spend. Structure: 3-5 ad sets, each testing a different variable, with equal budgets. Creatives: 3-5 creatives per ad set, all new or untested. Targeting: broad or Advantage+ (let the creative signals do the targeting). Goal: identify winning creatives and audiences to graduate to the scaling campaign.
**Campaign 2: Scaling.** This is your CBO campaign where proven winners run at higher budget. Budget: 50-60% of total ad spend. Structure: 2-3 ad sets containing your best-performing creatives from the testing campaign. Creatives: only proven winners (top performers from testing). Targeting: broad or Advantage+. Goal: maximize conversions at target CPA from creatives and audiences that have already proven themselves.
**Campaign 3: Retargeting.** This campaign targets people who have already interacted with your brand, website visitors, engagement audiences, customer lists. Budget: 10-20% of total ad spend. Structure: 1-2 ad sets segmented by funnel stage (e.g., website visitors vs. cart abandoners). Creatives: offer-specific, urgency-driven, or testimonial-based (different from prospecting creatives). Targeting: custom audiences only. Goal: convert warm audiences at low CPA.
How Many Ad Sets Per Campaign
The sweet spot for ad sets per campaign depends on your budget and campaign type. For testing campaigns (ABO), 3-5 ad sets is optimal. Fewer than 3 does not give you enough test variants. More than 5 spreads your testing budget too thin, each ad set needs enough budget to exit the learning phase (remember the 50 conversions in 7 days threshold). For scaling campaigns (CBO), 2-3 ad sets is ideal. The algorithm distributes budget most effectively across a small number of proven ad sets. More than 4-5 ad sets in a CBO campaign creates distribution inefficiency.
Creative Allocation Per Ad Set
Each ad set should contain 3-6 creatives with genuinely distinct Entity IDs. Fewer than 3 does not give Andromeda enough signal diversity to work with. More than 6-8 per ad set causes the algorithm to concentrate delivery on just 2-3 creatives while the rest get starved of impressions, you end up paying for creatives that never get tested. The sweet spot is 4-5 creatives per ad set, each representing a different hook type, visual style, or copy angle. Tools like AdRiseLab can generate these diverse sets quickly by analyzing your product and producing creatives with distinct signal fingerprints.
Common Structural Mistakes
**Too many campaigns.** Every campaign you add fragments your budget and your data. If you have more than 5-6 active campaigns, you almost certainly need to consolidate. Many advertisers run separate campaigns for each product, each audience, or each creative concept, this creates fragmentation that Andromeda penalizes. Consolidate into the 3-campaign framework.
**Too few creatives per ad set.** Running 1-2 creatives per ad set starves Andromeda of signal diversity. The algorithm cannot effectively discover audiences with insufficient creative signals. Every ad set should have at least 3 genuinely distinct creatives.
**Audience overlap.** When multiple ad sets target the same or heavily overlapping audiences, your ads compete against each other in the auction. This drives up CPMs and wastes budget. Use Meta's Audience Overlap tool to check for overlap, and merge or exclude as needed. In 2026, the simplest fix for audience overlap is to use broad targeting or Advantage+, let the algorithm figure out who to reach based on creative signals rather than defining overlapping audience segments.
**Budget misallocation.** Spending 80% of budget on testing and 20% on scaling is backward. Testing should identify winners efficiently (20-30% of budget), and scaling should maximize returns on those winners (50-60%). If you are perpetually testing and never scaling, your structure is wrong.
The Advantage+ Shopping Question
Advantage+ Shopping campaigns (ASC) are Meta's most automated campaign type for e-commerce. They use the broadest possible targeting, CBO by default, and automated placement optimization. ASC simplifies account structure dramatically, you can run a single ASC campaign with all your creatives and let the algorithm handle everything else.
Should you use ASC? For e-commerce advertisers spending $5,000+ per month, ASC is worth testing. Many advertisers report that a single ASC campaign outperforms their manually structured campaigns. However, ASC provides less visibility into what is working and why. You cannot isolate which audiences or placements are driving results. For advertisers who need that data, for creative strategy, for reporting, for understanding their market, the 3-campaign framework provides better visibility while maintaining strong performance.
A common hybrid approach: run ASC as your primary scaling campaign and maintain a separate ABO testing campaign for creative R&D. This gives you the efficiency of ASC for scaling and the control of ABO for testing.
How to Restructure Without Losing Data
If your current account structure is fragmented, restructuring is necessary but nerve-wracking. The fear of losing pixel data and audience learnings is real. Here is how to restructure safely. **Do not delete old campaigns immediately.** Pause them instead. Your pixel data and conversion history are account-level, not campaign-level, they persist regardless of campaign changes. **Migrate gradually.** Create your new 3-campaign structure alongside the old campaigns. Move your best-performing creatives into the new structure first. Run both old and new in parallel for 7-14 days to validate performance. Then pause the old structure. **Expect a short learning dip.** New campaigns enter the learning phase. Budget accordingly, the first 7-10 days may show higher CPAs. This is normal and temporary.
Account structure is not a set-it-and-forget-it decision. Review your structure monthly. Consolidate campaigns that have drifted. Prune ad sets that are underperforming. Graduate winners from testing to scaling. The 3-campaign framework gives you a system for continuous optimization, not just a one-time setup.
Related Reading
Learn how Meta's Andromeda algorithm evaluates creative signals and why account structure affects delivery quality. Understand the learning phase and how consolidated structure helps you exit it faster. See the creative testing framework for structuring your testing campaign's creative sets. And explore bulk launching workflows for efficiently populating your testing campaigns with diverse creatives.