How Multi-SKU Brands Should Structure Meta Ads Campaigns

Most Meta Ads guides assume you sell one product. That advice breaks the moment you have 15 SKUs with different margins, different buyers, and different sell-through priorities.

The standard recommendation is to consolidate everything into one campaign so Meta's algorithm can optimize across the full data set. That works when every product earns roughly the same margin and targets roughly the same person.

It does not work for multi-SKU brands.

If you run supplements with six formulas, a fashion line with 50 SKUs across pants, tops, and dresses, or a beauty brand with separate skincare and makeup lines, your meta ads structure for a multi-SKU brand needs to account for margin differences, persona differences, and catalog complexity. Get the structure wrong, and you either starve your best products of budget or feed your worst ones too much of it.

This post gives you a diagnostic framework: five questions that determine whether to group or split, how to handle DPA (Dynamic Product Ads) catalogs, and how to size your testing budget across categories. Every recommendation comes from Flighted's general structural principles and our hands-on work managing multi-SKU accounts.

Key Takeaways

1. Consolidation wins in general, but fails when SKUs have different margins, different personas, or low conversion volume per category.

2. Five diagnostic questions determine your structure: category count, margin spread, persona overlap, regional complexity, and creative throughput.

3. DPA catalogs need separate attribution treatment. Assess DPA ROAS (Return on Ad Spend) on 7-day click, never 1-day view.

4. Push category segmentation to the adset level, not the campaign level, to preserve data signal.

5. Size your testing budget per concept adset. If the math does not support 20 simultaneous tests, run 3-5.

1. Why Consolidation Generally Wins but Fails for Multi-SKU Brands

Meta's conversion optimization algorithm pools data at the campaign level. More conversion events in a single campaign means faster learning and better optimization.

This is why consolidation is the default recommendation. It works.

But the one-campaign, one-adset approach is wrong for roughly 90% of multi-SKU brands. Here is why:

• Different unit economics across SKUs. A supplement brand selling a $30 protein powder at 60% margin and a $15 vitamin pack at 30% margin cannot afford the same CPA (Cost Per Acquisition) for both. Consolidation treats them as interchangeable.

• Non-overlapping personas. A beauty brand's acne-prone 22-year-old and its anti-aging 45-year-old are not the same buyer. Grouping them dilutes targeting for both.

• High AOV (Average Order Value), low conversion volume. A furniture brand doing $20-30M might only generate 5-10 conversions per day per category. Splitting those across too many campaigns leaves each one below the 50-conversions-per-week threshold Meta recommends for stable optimization.

Here is what this looks like in practice. Consider a $20-30M furniture brand running a single consolidated cold CBO (Campaign Budget Optimization) where 57% of the budget flowed to one DPA adset showing a roughly 7x surface ROAS. Sounds strong. But when you evaluate that same adset on 7-day click attribution, the real ROAS was 1.9x. The algorithm was chasing view-through conversions that would have happened anyway.

That is the consolidation trap. The numbers look good until you measure correctly.

For a deeper dive into how margin should influence your Meta Ads account structure, we break down the math in a dedicated post.

2. Five Diagnostic Questions That Determine Your Structure

Before you build or rebuild your campaign structure, answer these five questions. Your answers map directly to whether you should group SKUs into shared campaigns or split them apart.

Question 1: How many product categories do you have?
If you sell across 2-3 categories, a single campaign with category-level adsets is usually sufficient. At 5 or more distinct categories, you likely need separate campaigns for your top 2-3 performers and a catch-all for the rest.

Question 2: How wide is the margin spread?
If your lowest-margin SKU earns 25% and your highest earns 70%, you cannot use a single target CPA. Group products with similar margin profiles (within 10-15 percentage points of each other) and separate those with meaningfully different economics.

Question 3: Do you have multiple non-overlapping personas?
If your categories serve different buyers who would not respond to the same creative, split them. A supplements brand selling "energy" formulas to gym-goers and "sleep" formulas to stressed professionals is targeting two distinct audiences. Group them, and your creative speaks to neither.

Question 4: How many regions do you serve?
Brands selling in the US, UK, and EU often need separate campaigns per region due to currency, language, and regulatory differences. Multi-region, multi-SKU means the structure compounds quickly. Simplify where you can.

Question 5: What is your creative throughput?
Each campaign-adset combination needs fresh creative on a regular cycle (every 2-4 weeks at minimum). If you can produce 5 net-new concepts per month, you cannot support 10 separate adsets. Structure your account to match your production capacity, not your ideal theoretical setup.

The rule: Group when SKUs share similar margins, target the same persona, and you have limited creative volume. Split when margin profiles diverge by more than 15 points, personas do not overlap, or sell-through priorities differ by category.

3. How Catalog Structure in DPA and Advantage+ Should Mirror Business Priorities

DPAs (Dynamic Product Ads) dynamically re-prioritize products a user has already viewed. They sit at the bottom of the funnel, targeting the most aware buyers. They place right before purchase.

That means DPAs get far more credit than they deserve.

This is non-negotiable: assess DPA ROAS on 7-day click attribution or incremental attribution. Never on 1-day view. The 1-day view window inflates DPA performance by claiming credit for conversions that were already going to happen.

The DPA Death Spiral

Here is the pattern we see repeatedly:

1. Brand sees strong DPA ROAS (often 5-10x on surface metrics).

2. Brand keeps bumping DPA budget because the numbers look good.

3. DPA budget increase comes at the expense of creative production and top-of-funnel prospecting.

4. Whole account declines because nobody is filling the top of the funnel.

Do not fall into this cycle. DPA is a conversion capture tool, not a growth engine.

Catalog Strategy for Multi-SKU Brands

For multi-SKU brands at lower spend levels ($10-30K/month), a prospecting catalog can work as a long-tail discovery play. Run it as its own campaign with 3-5 ads. Lead with an intro card that frames the catalog, and vary the intro card and copy across ads to test different entry points.

Keep your retargeting catalog small. Include only your top 10-20 products by margin contribution. Do not dump your entire catalog into retargeting and hope the algorithm figures it out.

One critical note for restricted categories: supplements, health products, and similar verticals often cannot run catalog ads at all. Verify in Commerce Manager that your product feed is approved before building a catalog strategy. If it is not permitted, do not waste time troubleshooting. Build static ad alternatives instead.

For a complete breakdown, see our guide to Meta Dynamic Product Ads.

4. Avoiding Budget Dilution Across Too Many SKU-Level Adsets

The ad-to-budget ratio inside each adset matters more than most brands realize.

If you load 15 ads into an adset with a $50/day budget, the algorithm will concentrate spend on 2-3 ads and starve the other 12. You will never know whether those 12 were weak creative or just outcompeted for budget.

That kills your ability to read test results.

The fix: Split by concept at the adset level. Each adset should represent a single persona-angle-offer combination, with 3-5 ads testing variations of that concept. This gives you readable signal on whether the concept works, not just whether one ad within it won the budget auction.

Do not create individual adsets per SKU unless each SKU genuinely has its own persona and margin profile. A supplements brand with 8 flavors of the same protein powder does not need 8 adsets. The buyer is the same person. The margin is the same. Group those flavors into one adset and test different creative angles (taste, convenience, ingredients) instead.

Budget math per adset: At a minimum, each adset needs enough daily budget to generate 1-2 conversions per day for the algorithm to optimize. If your CPA is $40, that means at least $40-80/day per adset. An adset running on $15/day at a $40 CPA is in permanent learning mode. It will never stabilize.

5. Why Adset-Level Segmentation Beats Campaign-Level Splits

This is the structural principle that ties everything together.

Campaign-level splits are the most expensive way to segment. Every time you create a new campaign, you chunk your conversion data into a separate silo. Meta cannot share learning across campaigns. At smaller budgets ($10-50K/month), this fragmentation can cost you 5-10% in efficiency compared to keeping segmentation at the adset level.

Adset-level segmentation preserves more data signal. Campaign-level data inherits downward into adsets, so all adsets within a campaign benefit from shared learning. At scale, 5-10% efficiency swings are real money.

The rule: Push category and SKU segmentation to the adset level whenever possible. Create separate campaigns only when you have fundamentally different optimization goals (prospecting vs. retargeting), different budget constraints, or regulatory requirements that force separation.

For fashion and apparel brands specifically, we cover this in detail in our Meta Ads account structure guide for fashion brands.