How to Build a Size-Level P&L

Most fashion brands run their P&L at the style or SKU level, and stop there. But a single style sold in six sizes can contain six very different businesses — some sizes converting well and returning rarely, others eating margin through discounting, extra shipping, and repeat returns. If you never break profit down by size, you are averaging away the exact signal you need to fix your buy.

Why a Size-Level P&L Matters More Than Style-Level Numbers

A style-level P&L tells you a jacket made 40% gross margin. It does not tell you that size Small made 55% margin while size XXL made negative 8% after returns, markdowns, and the extra freight from being restocked twice. Averages hide the sizes that are quietly subsidized by the rest of the curve.

This matters because size curves are not symmetric in cost. Core sizes (typically M and L) tend to sell through at full price with fewer returns. Tail sizes — the smallest and largest in a run — are more likely to be bought as a "safety" purchase by a size-uncertain shopper, more likely to be returned, and more likely to end up marked down. Industry fit research has repeatedly linked size uncertainty to elevated return rates, and returns are one of the largest hidden costs in apparel margin structures, a point echoed in retail industry analysis from NRF on the true cost of returns.

What "Profit by Size" Actually Requires

Profit by size is not just revenue divided across a size curve. It requires allocating four cost buckets down to the size level:

• Cost of goods sold (COGS), which is usually flat per unit regardless of size unless up-sizing costs more fabric
• Returns cost: processing, restocking, damaged/unsellable inventory, and reverse shipping
• Markdown cost: the gap between original and final sold price for units that didn't move at full price
• Fulfillment cost: pick, pack, and ship, which can vary slightly for plus-size or petite packaging

Building the Size-Level Margin Model

Step 1: Pull Size-Level Order and Return Data

Start with every order line at the size level, not the style level. You need units sold, gross sales, discounts applied, and units returned per size, per style, over a consistent window (a full season is best, to average out early-cycle noise).

Step 2: Allocate Returns Cost Per Size

This is the step most P&Ls skip. Each return has a processing cost (inspection, restock or write-off, refund handling) and a lost-margin cost if the unit is marked down or scrapped on the second pass. Multiply your average per-unit return handling cost by the size-specific return rate, not the blended rate.

Step 3: Allocate Markdown Cost Per Size

Pull the final sold price for every unit that did not sell at full price, grouped by size. Tail sizes disproportionately end up in markdown because slower sell-through pushes them into end-of-season clearance.

Step 4: Build the Table

Once you have the four cost buckets, the size-level P&L table looks something like this for a single style:

This is a synthetic example, but the shape is common: a bell curve in margin that mirrors the bell curve in fit confidence. The tails are where uncertainty, and therefore returns, concentrate.

What to Do Once You See the Numbers

Fix the Buy, Not Just the Price

If XXL is consistently margin-negative after returns, discounting it further will not fix the problem — the issue is fit confidence, not price. The better move is usually a combination of tighter size-curve buying (fewer tail units relative to core sizes) and better pre-purchase sizing guidance so fewer tail-size shoppers buy the wrong size in the first place. This connects directly to how you forecast demand by size for the next buy — a size-level P&L is the input that makes that forecast profit-aware instead of just volume-aware.

Watch for the Out-of-Stock Trap

There is a real tension here: cutting tail-size units to protect margin can starve genuine demand and push those shoppers to a competitor. Before cutting a size, check whether the negative margin is a return-rate problem or a stock problem. If you're already seeing this size stock out early, read our guide on how to manage out-of-stock sizes without losing sales before you cut the buy.

Reduce Return Rate at the Source

The highest-leverage move is usually reducing the size-specific return rate itself, since that number multiplies through every other cost bucket. Tools like YourSizer's fit-recommendation engine are built specifically to reduce size-driven returns by giving shoppers a confident size recommendation before checkout, rather than after a wrong-size box shows up. Lowering the tail-size return rate by even five to ten points can move a size from margin-negative to margin-positive without touching price.

How Returns Quietly Erase Size-Level Margin

It is worth being explicit about how large this effect can be. Returns are not just a refund; they carry restocking labor, reverse logistics, and in apparel, a meaningful share of returned units that cannot be resold at full price due to try-on wear or packaging damage. Analysis of retail returns economics from sources like McKinsey has highlighted how returns processing and devaluation can consume a large share of otherwise-healthy margin — and that effect is not evenly spread across a size run. If you want the fuller breakdown of how this shows up in unit economics, see our piece on the hidden returns cost in your unit economics.

Building This Into a Recurring Process

A size-level P&L is only useful if it's refreshed regularly, not built once as a research project. The practical approach:

1. Rebuild the size-level margin table every season, per core style
2. Flag any size falling below a minimum margin threshold (e.g., under 15%)
3. Route flagged sizes into either a buy-curve adjustment or a fit-confidence intervention
4. Re-check the following season to confirm the intervention worked

Frequently Asked Questions

What data do I need to build a size-level P&L?

You need order-line data broken out by size (not just style), including units sold, gross and net sales, discount amounts, and returns per size. You also need your cost of goods, average return-handling cost, and markdown records tagged to size where possible. Most of this already exists in your order management or Shopify/Ticimax/ikas backend; the work is in joining it at the size level instead of the style level.

How often should I rebuild the size-level P&L?

Quarterly is a reasonable minimum for most apparel brands, with a full rebuild at the end of each selling season before the next buy is finalized. Fast-fashion or high-velocity brands may want a monthly view since size-level return patterns can shift with promotional cycles and new customer mix.

Is a size-level P&L only useful for brands with a lot of SKUs?

No. Even a brand with a handful of core styles benefits, because the size curve within a single style already contains meaningful profit variation. The insight scales down to a single best-selling style just as well as it scales up across a full catalog — the math per size stays the same either way.

Can reducing return rates really fix a negative-margin size?

Yes, in many cases. Since return-handling cost and markdown exposure are both driven by return rate, cutting the return rate for a tail size directly reduces the two largest cost buckets in the size-level P&L. Better pre-purchase size guidance is typically the fastest lever, since it prevents the return before it happens rather than processing it after the fact.