Many apparel brands think their Meta or Google campaigns are underperforming because the audience is wrong. In reality, the campaign may be spending on the wrong products.
This is one of the biggest hidden leaks in fashion performance marketing. Large catalogs often contain hundreds or thousands of SKUs across sizes, colors, collections, categories, fabrics, discounts, and stock levels. When all products are pushed into one broad campaign, ad platforms may spend on items that look clickable but are commercially weak.
That is how brands end up with traffic, add-to-carts, and low ROAS.
A campaign can be technically correct and still lose money.
Why? Because the product mix is wrong.
For example:
A low-AOV product gets clicks but cannot absorb CAC. A kurta has only XS and XXL left in stock. A sandal gets engagement but has a high return rate. A dress has a beautiful image but poor size availability. A clearance SKU attracts bargain hunters but does not improve contribution margin. A bestseller is under-promoted because the algorithm did not give it enough learning spend.
This is why fashion brands need product performance management, not just campaign management.
Use this framework to classify your catalog.
High spend, high ROAS. These products deserve more budget, better creatives, landing-page support, and inventory protection.
Low spend, high ROAS. These are often the biggest opportunity. The product works, but the algorithm has not given it enough delivery.
High spend, low ROAS. These look good in ads but do not convert profitably.
Low spend, low ROAS. These should usually be excluded from paid campaigns unless there is a strategic reason to revive them.
A public AdYogi case study on Bombay Shirt Company is a strong example of why catalog-level control matters. The brand had 500+ products and faced product/category insight issues, low page views on 200+ SKUs, and skewed ad-spend distribution. AdYogi reported that a catalog-linked ad showing only Business Shirts delivered 33% higher ROAS than generic ads, and customised product sets delivered 61% higher ROAS than catalog ads without customisation.
That is the core lesson: the same media platform can perform very differently depending on which products are being promoted.
Start with a product-level report, not a campaign report.
Check:
Then ask:
Which products are getting spend but not orders? Which products are profitable but under-spent? Which high-AOV products need more creative support? Which SKUs should never enter paid campaigns because inventory is broken?
For apparel and footwear brands, do not push the full catalog blindly. Build structured product sets.
Create separate product sets for:
The goal is not to overcomplicate the account. The goal is to stop algorithms from wasting money on products that should not be advertised.
This is especially important in fashion.
A product can be technically “in stock” but commercially dead if only one odd size is available. For example, a kurta with only XS left should not be promoted the same way as a kurta with S, M, L, XL, and XXL available.
Create feed rules that exclude products where:
AdYogi’s profit-first optimisation blog also highlights low-inventory SKUs, low-conversion designs, broken links, and poor creatives as common causes of catalog waste, and recommends cleaning the catalog before setting automated guardrails.
Fashion brands should run a weekly SKU-level performance review.
The meeting should answer:
A performance agency that does not discuss SKU-level performance is only managing the surface of the account.
It is the process of tracking and optimising paid media performance at the SKU, product, category, size, and margin level instead of only looking at campaign-level ROAS.
No. Large apparel catalogs should be cleaned and segmented. Broken-size products, low-AOV products, poor converters, and low-stock SKUs should usually be excluded or controlled.
A hidden winner is a product with strong conversion or ROAS but low ad spend. These products often need dedicated product sets, creatives, and budget allocation.