What counts as COGS for an apparel store

Landed cost, trims, packaging, fulfilment and return handling: which apparel costs belong in COGS when you're measuring POAS, and which don't.

Cost of goods as a share of revenue, US-listed apparel companies

43.12%

Source: Aswath Damodaran, NYU Stern: Margins by Sector (US), data as of January 2026, accessed 27 Sept 2026. Full citation

Industry figure (COGS/Sales) across 35 US-listed apparel companies, such as Levi Strauss, PVH, Ralph Lauren and Columbia Sportswear. These are large public brands that sell through wholesale and their own stores, not Shopify-sized stores, so treat it as a reference point rather than a target.

For measuring POAS, "cost of goods" means every cost that scales with each order you sell. Ad spend is not included, because it's the other side of the ratio. For an apparel store, the supplier's unit price is only the start.

What goes in

CostCount it?Why
Unit cost from the factory or wholesalerYesThe base of landed cost
Inbound freight, duty and customs brokerageYesLanded cost: duties on imported apparel can be a meaningful share of unit cost
Trims, labels, hangtags, polybagsYesPer-unit and easy to forget
Pick, pack and outbound shipping you payYesScales with orders, not with time
Payment processing feesYesA share of every order
Return shipping and handlingYes, spread across all ordersNot on any supplier invoice, so easy to miss
Markdowns and discountsNo, they lower revenueCounting them as cost double-counts them
Photography, samples, designNoFixed or per-season overhead
Google Ads spendNoIt's what POAS divides by

Returns deserve their own line. In its 2026 report on apparel sizing, Coresight Research describes the reverse-logistics cost of each return (shipping, restocking, repackaging and inspection) as "often exceeding the margin on the original sale". We haven't found a reliable published figure for what one apparel return costs, so measure your own:

  • the return label, if you pay for it
  • the time to receive, inspect, refold and restock
  • the units that come back unsaleable and have to be written off or marked down

Add these up over a month, then divide by all orders in that month, not just the returned ones. That gives a per-order return cost to carry into every POAS calculation.

Building one unit's cost

Illustrative arithmetic with made-up round numbers, not a benchmark.

A shirt that sells for $60:

  • Factory price $12, plus $2 of freight and duty, plus $0.50 of trims and packaging: landed cost $14.50.
  • Pick, pack and the shipping you pay: $6.
  • Payment fees at 3% of $60: $1.80.
  • Return cost spread across all orders: $2.

Cost per order: $24.30, or about 40% of the selling price. The factory price alone would have suggested 20%.

The gap between those two percentages is the reason apparel POAS gets misjudged. A store that enters only the factory price into Shopify's Cost per item field will report margins it doesn't have.

Three levels of accuracy

Selvra OS works with three tiers of cost data and shows which one a figure is based on.

T1: one flat margin (labelled "estimated")

One percentage for the whole catalogue. This works for a store that sells one kind of product at one price point, like a single-line basics brand. It misleads a store that mixes tees with outerwear, because the flat figure is wrong for nearly every product.

T2: margin by category (labelled "approximate")

A margin per product type: tees, knitwear, outerwear, accessories. When a catalogue mixes basics with seasonal fashion, margins can differ widely between categories, and this is where the extra accuracy starts to change decisions.

T3: cost per product (labelled "verified" once synced)

Shopify's Cost per item field, per variant, read directly. This is the most accurate tier. Watch for sizes: a cost entered on one size doesn't apply itself to the others, and extended sizes can cost more to make.

Reading the benchmark above

The figure above comes from large US-listed apparel companies. It is a reference point, not a target. A store selling only direct to consumers sells every unit at retail price, which lowers cost as a share of revenue compared with selling the same unit wholesale. Smaller production runs can push unit costs the other way. Use the benchmark to check that your own figure is plausible, then trust your own figure.

Common COGS mistakes in apparel stores

  • Using the supplier price and stopping there. Landed cost is higher, sometimes substantially for imported goods.
  • Leaving returns out. The cost of processing returns is real even when the product goes back into stock.
  • Treating sale price as a cost. A markdown shrinks revenue. Record it as lower revenue, and don't add it as an extra cost.

How Apparel & Fashion compares

VerticalCOGS as a share of revenueSource
Apparel & Fashion43.12%Aswath Damodaran, NYU Stern: Margins by Sector (US), data as of January 2026
Home & Furniture69.72%Aswath Damodaran, NYU Stern: Margins by Sector (US), data as of January 2026
Food & Beverage / CPG76.77%Aswath Damodaran, NYU Stern: Margins by Sector (US), data as of January 2026
Electronics & Gadgets61.23%Aswath Damodaran, NYU Stern: Margins by Sector (US), data as of January 2026

The same question in other verticals

Source and how to read this number

Figure
Cost of goods as a share of revenue, US-listed apparel companies: 43.12%
Source
Aswath Damodaran, NYU Stern: Margins by Sector (US), data as of January 2026
Link
https://pages.stern.nyu.edu/~adamodar/New_Home_Page/datafile/margin.html
Accessed
27 Sept 2026
Caveat
Industry figure (COGS/Sales) across 35 US-listed apparel companies, such as Levi Strauss, PVH, Ralph Lauren and Columbia Sportswear. These are large public brands that sell through wholesale and their own stores, not Shopify-sized stores, so treat it as a reference point rather than a target.
Last reviewed
30 Sept 2026