Clothing Brand Meta Ads Strategy: $491K in 30 Days (6.36x ROAS)

The clothing brand Meta ads strategy we used to generate $491,675 in revenue on $77K in ad spend (6.36x ROAS) in a single month is built on one principle: split campaigns by Average Order Value (AOV), let the creative do the targeting, and structure the account so Meta’s Andromeda algorithm can sequence winning ads across the funnel. This is the exact account structure, testing framework and creative strategy we run for fashion brands in 2026, after Andromeda (Meta’s ranking update rolled out in late 2024) and GEM (Generative Emergent Model) changed how targeting and attribution work.

At Skaleit we manage Meta ads for ecommerce brands. This case study covers a clothing brand where interest targeting, single winning creatives and one-format ad strategies had stopped working. Here is what replaced them.

TL;DR

• Result: $491,675 revenue on $77K spend, 6.36x ROAS, 30 days, one clothing brand.
• Account structure: CBO campaigns split by AOV band ($80 to $100, $100 to $150), ad sets split by product category.
• Ad volume: 15 to 30 ads per ad set. Winners are stacks, not singles.
• Catalog: DPA split by category or occasion, never “all products” in one ad set.
• Scaling: one cost cap campaign with 10 to 15 winning ads, duplicated by post ID.
• Retargeting: a separate campaign at 5 to 8% of spend. Existing customers excluded everywhere else.
• Creative: clothes on ground, UGC pictures, split screen, offer-led statics, aesthetic shoots, founder video.

Prefer to watch? This is the video the article is based on.

What Stopped Working After Andromeda for Clothing Brands

Interest targeting is dead for fashion accounts. Stacking magazine interests, competitor brand names or lifestyle interests no longer moves performance, because after Andromeda the targeting is done by the creative itself. Meta reads the image, identifies the style, the product category and the demographic cue, and delivers accordingly.

Three more things broke at the same time:

1. Two or three creatives per ad set is not enough. Ad sets need 15 to 30 ads to give Andromeda and GEM enough material to sequence.
2. The “winning creative” no longer exists. What wins is a stack of ads that convert together. The last ad gets the attribution, but remove the top-of-funnel ads and the last-click ad stops converting.
3. Single-format accounts underperform. Statics only, or video only, starves the sequence. Meta needs format and concept diversity.

The AOV-Split CBO Account Structure That Hit 6.36x ROAS

Fashion brands sell products at very different price points, and that is the trap. Drop a $40 t-shirt and a $180 outerwear piece into the same CBO (Campaign Budget Optimization) campaign and Meta’s default bidding, highest volume at lowest cost, pushes spend to the cheapest conversions, not the most profitable ones. The structure below fixes that by keeping every campaign inside one margin band, so one ROAS target is valid for everything in it.

Campaign 1: CBO for AOV $80 to $100

• Ad set 1: t-shirts, 15 to 30 ads.
• Ad set 2: hats, 15 to 30 ads.
• One ROAS target for the campaign, because the margins inside it are similar.

Learning happens at the creative level, then aggregates at the ad set level. That aggregate is what decides who the ad set reaches, so the ad set has to hold one product category or one occasion of use. People interested in hats are rarely the people interested in t-shirts, and mixing them blurs the signal.

Campaign 2: CBO for AOV $100 to $150

Same logic, different price band, different ROAS target, 15 to 30 ads per ad set.

Campaign 3: DPA split by category

This is where most fashion brands go wrong. They build one Dynamic Product Ad set called “all products” and wonder why it dies. Instead:

• Ad set 1: DPA with t-shirts only.
• Ad set 2: DPA with outerwear only, or split by occasion of use.
• Add as many ad sets as you have categories. The only rule is one category or occasion per ad set.

Splitting the catalog tells Meta to optimise delivery for people interested in that specific category, which lifts click-through and conversion rate at the catalog level.

Campaign 4: cost cap with winning ads (scaling)

• Bidding: cost per result cap, set for the AOV of the products inside.
• One ad set with 10 to 15 winning ads, all from the same price band so one cap makes sense.
• Winners are duplicated from campaigns 1 and 2 using the post ID. Without the post ID the ad loses its engagement, and a lower-engagement ad ranks lower in the auction.

For how we set the cap and step it up, see our bid cap strategy for scaling Meta ads.

Campaign 5: manual retargeting for existing customers

• Spend cap: 5 to 8% of the account budget.
• Ad set 1: DPA cross-sell to existing customers.
• Ad set 2: 20 to 50 stacked creatives across product categories. The audience is already defined, so mixing products is fine here.
• Existing customers are excluded from all four acquisition campaigns.

Skip the exclusion and Meta spends against people who already bought. CAC looks great, the account never learns to find new customers, and scaling stalls. Email and SMS are cheaper ways to sell to buyers you already paid for.

Why the Winning Creative No Longer Exists

Ads Manager assigns each purchase to the last ad a person engaged with. In a sequenced account that means one ad collects the credit for work the whole stack did.

So the decision rule changes. You do not turn off an ad because its own attributed purchases look thin. You judge the ad set and the campaign, and you keep the stack that produces the account-level ROAS.

How to Test and Scale Inside This Structure

Every new ad launches inside the existing ad sets, not in a separate test campaign. That keeps the sequence intact and puts the new creative in front of the audience the ad set already learned.

Ad sets hold 50 to 150 ads, so freeing slots for new tests matters. As performance climbs we raise the campaign budget and let the CBO redistribute across ad sets.

Highest Volume vs Maximize Conversion Value

We stay with highest volume inside AOV-split campaigns because it gives control over ROAS per margin band. Maximize conversion value only makes sense in a single consolidated CBO, and even there it treats a high-value, low-margin sale as a win.

Creative Strategy: the Creative Is the Targeting

If an ad shows a street-style t-shirt on a man, Meta will deliver it to men who like street style. That is how GEM works, so creative diversity is the targeting plan. Different people need to see themselves in the ad before they buy the same product.

Four levers sit behind every format that sells clothes:

1. Association. Your product next to the lifestyle, and sometimes next to other brands, the buyer already wants.
2. Aesthetics. Old money, streetwear, dark academia. Show the aesthetic, not just the garment.
3. Mission. “Luxury quality made affordable” is a reason to choose you over the same t-shirt with another logo.
4. Community. Running crews, gym culture, a scene. People buy the membership.

The creatives that return 3x or more on client accounts right now: clothes-on-ground photos, plain UGC pictures, split screens with the outfit and a product detail, offer-led statics in single, grid and split layouts, eye-catching bright photos with a sharp headline, aesthetic shoots that place the product in the buyer’s world, audience call-outs with the matching outfit, and founder or collection videos for the top of the funnel.

Related Results From Skaleit

The same structure and creative logic runs across our apparel accounts. See how we increased GNJ Underwear revenue by 357% in three months, how we sell out fashion drops in 48 hours, and the $827K swimwear playbook. For the algorithm side, the complete Andromeda guide for ecommerce covers attribution and structure in depth.

FAQ

Want This Structure in Your Account?

If your clothing brand is stuck with interest targeting, thin ad sets or a catalog campaign that never scales, get in touch with us and we will look at your account before the first call.