Building a Better Size Curve for Your Collections

The Size Curve Is a Quiet Margin Decision

Every time you place a buy, you split it across sizes — so much Small, so much Medium, so much Large. That split is the size curve, and it's one of the most consequential and least examined decisions in merchandising. Get it right and inventory flows cleanly to full-price sell-through. Get it wrong and you stock out on the sizes customers want while the rest sits until markdown. Most brands use a default curve — often a generic bell shape applied to everything — and never question whether it matches their actual customers. This guide covers how to build a size curve grounded in real demand instead of habit, why the default fails, and how fit data improves it. It pairs closely with forecasting demand by size.

Why the Default Bell Curve Fails

The standard approach assumes demand follows a neat bell across the size range, peaking in the middle. Sometimes it roughly does, but often it doesn't — and applying the same bell to every style and every brand is where the trouble starts. Your customer base has a specific body distribution, and it may not be centered where the default assumes.

A brand whose customers skew toward larger or smaller sizes, or whose specific category attracts a particular body type, will systematically mis-buy if it uses a generic curve. The bell is a convenient assumption, not a measured fact, and convenient assumptions are exactly what cost margin at the size level. The first step to a better curve is to stop treating the default as ground truth and start measuring what your demand actually looks like.

Segment Your Curve, Don't Use One for Everything

A single house-wide size curve is almost always wrong, because different categories and fits sell different distributions. A relaxed-fit style sells a different curve than a fitted one even to the same customers, and outerwear, dresses, and basics each have their own shape. Applying one curve across all of them guarantees error somewhere.

Segment your curves by category, fit type, and where relevant customer group. The more specifically you match the curve to the kind of product and the kind of buyer, the closer your buy lands to real demand. This takes more analysis than a single default, but it's exactly the analysis that prevents the simultaneous stockouts and markdowns a one-size-fits-all curve produces. Tagging your history by fit type makes this segmentation possible, which is why that discipline pays off repeatedly.

Build From Net Sell-Through, Not Gross Buys

The data to build your curve from is historical sell-through by size — but net of returns, not gross sales, and ideally at full price rather than markdown. Gross buys are contaminated by past curve errors (you can't sell a size you didn't stock), and gross sales are inflated by units that came back. Net full-price sell-through is the cleanest signal of what customers actually wanted and kept.

A size that sold well but returned heavily isn't strong demand — it's a fit problem masquerading as one, and buying more of it repeats the error. Correcting your curve for returns is one of the highest-impact refinements available, and it directly links returns data to merchandising. We explore that connection in turning returns data into better-fitting products.

Use Customer Body Data Directly

The most direct way to build a size curve is from your customers' actual bodies. If you can see the distribution of body measurements across your customer base — from a fit tool, sizing widget, or order history — you can build the curve from the demand side rather than inferring it backward from contaminated sales. That's a fundamentally stronger foundation.

Most brands never collect this data and so default to assumptions. Those that do can build curves grounded in who their customers really are, which is exactly the kind of insight a fit tool surfaces. This turns the size curve from a habit into a measured reflection of your market, the same demand-side advantage we describe in how fit data improves inventory planning.

Handle the Tails Deliberately

The extreme sizes — the smallest and largest in your range — deserve specific attention, because they're where curve errors are most expensive and most visible. Under-buy the tails and you alienate customers who already struggle to find their size; over-buy them and you create predictable markdown. Neither is acceptable, and a generic curve handles the tails worst of all.

Decide your tail strategy deliberately based on your real demand data, not by lopping off a standard percentage. If your customer data shows genuine demand in the extended sizes, stock to it — serving those customers well builds loyalty in segments competitors neglect. This is part of building a size assortment that matches your customers, and the tails are where that match matters most.

Test, Measure, and Refine Each Season

A size curve is not a set-and-forget decision. Treat each season as an experiment: buy to your best curve, then measure size-level sell-through and returns against it, and refine. The brands with the best curves aren't the ones who guessed right once — they're the ones who tightened the curve season after season based on what actually happened.

This requires size-level reporting, not just style-level totals, and a habit of reviewing it. Watch which sizes stocked out early and which lingered to markdown, and feed that straight into the next buy. Continuous refinement is how a curve goes from roughly-right to genuinely tuned, and it's the same iterative discipline behind A/B testing sizing and fit changes.

Connect the Curve to Margin

A better size curve isn't an abstract merchandising nicety — it's a direct lever on margin. Buying closer to true demand means fewer stockouts (more full-price sales captured) and fewer markdowns (less margin given away), plus less working capital trapped in dead inventory. Those are real financial outcomes worth measuring and reporting.

Make size-curve accuracy a tracked metric, and tie it to the financial results it drives so the organization invests in getting it right. When merchandising, planning, and finance all see the curve as a margin decision rather than a buying detail, it gets the attention it deserves. This is part of treating fit and sizing as strategic, covered in making fit a company-wide priority.

Where YourSizer Fits

YourSizer surfaces the actual body distribution of your shoppers and the fit signals behind your returns — the exact demand-side data a better size curve needs. Instead of building your buy on a generic bell curve, you can ground it in who your customers really are and which sizes truly perform after returns, tightening the curve season over season. See the technology in our product overview, explore the merchant side on for brands, and check pricing.

Frequently Asked Questions

What's wrong with using a standard bell curve?

It assumes demand peaks in the middle of your size range and follows a neat distribution, which is a convenient assumption rather than a measured fact. Your customer base has a specific body distribution that may not be centered where the bell assumes, so applying it everywhere systematically mis-buys — especially in the tail sizes.

Should I have one size curve or several?

Several. Different categories and fits sell different distributions even to the same customers — a relaxed style sells a different curve than a fitted one. Segmenting your curves by category, fit type, and customer group lands your buy much closer to real demand than a single house-wide default ever could.

How do returns affect my size curve?

They reveal that some sizes selling well are actually fit problems, not real demand — a size with high gross sales but high returns shouldn't be bought more heavily. Building your curve from net sell-through, after returns and ideally at full price, gives a far truer picture of what customers actually want and keep.