How to calculate size curves for apparel buying
A size curve is the proportion of demand that falls to each size within a style or colour. You calculate it by taking historical unit sales by size, dividing each size by the total to get a percentage, and applying those percentages to your planned total buy — turning one total quantity into the right depth per size.
Done well, the size curve is what stops core sizes selling out early while fringe sizes sit until markdown.
- Definition — Size curve
- A size curve (also called a size ratio or size profile) is the proportion of demand that falls to each size within a style or colour — for example 10% XS, 25% S, 30% M, 25% L, 10% XL. It always sums to 100%, and applying it to a planned total buy turns one quantity into the right depth per size.
- size quantity = (size sales ÷ total sales) × total buy
- Used by: Buyers and planners breaking a total buy into size-level quantities
- Related: Buy plan, sell-through, assortment plan, demand by size
What is a size curve?
A size curve (sometimes called a size ratio or size profile) is the set of percentages describing how demand for a style splits across its sizes. For a five-size run it might read 10% XS, 25% S, 30% M, 25% L, 10% XL. The curve always sums to 100%, and multiplying it by a total buy gives the quantity to order in each size.
Why size curves matter
Demand is never even across sizes. If you split a buy evenly, you over-order the sizes at the edges of the range and under-order the middle, so the sizes most customers want sell out first. Early stock-outs on core sizes cost full-price sales, and the leftover fringe sizes drive end-of-season markdown. A good size curve directs depth to where the demand actually is.
The simple calculation
- 1. Gather historical unit sales by size for a comparable style.
- 2. Divide each size’s units by the total units sold to get its percentage of demand.
- 3. Multiply each percentage by your planned total buy.
- 4. Round to whole units and adjust so the sizes sum back to the total.
size quantity = (size sales ÷ total sales) × total buy
Example
A comparable tee sold 1,000 units last season, split as XS 100, S 250, M 300, L 250, XL 100. Dividing each by 1,000 gives the curve: 10% / 25% / 30% / 25% / 10%. Applying it to a planned buy of 1,200 units:
- XS = 10% × 1,200 = 120
- S = 25% × 1,200 = 300
- M = 30% × 1,200 = 360
- L = 25% × 1,200 = 300
- XL = 10% × 1,200 = 120
The five sizes sum back to 1,200 units, now ordered in proportion to real demand rather than evenly.
Common mistakes
- Using sales after stock-outs. If core sizes sold out mid-season, the history understates their true demand — the curve looks flatter than reality.
- One curve for everything. Applying a single curve across product types, channels, and regions misses real differences in how each segment buys.
- Ignoring fit changes. A re-graded or restyled fit can shift the curve, so an old style may not be a clean comparable.
- Forgetting to re-balance after rounding. Rounding each size independently can leave the total off by a few units.
Why size curves vary by region, channel, and product
Sizing conventions and body shapes differ by region, and channels attract different shoppers — a flagship store, a wholesale account, and an e-commerce site can each skew to different sizes. Product type matters too: a fitted dress curves differently from an oversized knit. The more your curve reflects the specific segment that buys the product, the more accurate the buy. Download the size curve template to build curves by segment in Excel.
How connected planning helps
From line plan to production, RetailNorthstar connects merchandising, design, buying, sourcing, and operations in one workflow. Size curves apply automatically as buys are built, vary by channel and region where it matters, and flow straight into purchase orders — so size-level quantities stay consistent from plan to PO.
Frequently asked questions
- What is a size curve?
- A size curve is the proportion of demand that falls to each size within a style or colour — for example 10% XS, 25% S, 30% M, 25% L, 10% XL. It lets you translate a total buy quantity into size-level quantities that match how the style actually sells.
- How do you calculate a size curve?
- Take historical unit sales by size for a comparable style, divide each size’s units by the total to get a percentage, and apply those percentages to your planned total buy. The result is the quantity to order in each size.
- Why do size curves vary by region and channel?
- Body shapes, sizing conventions, and customer mix differ by region, and channels attract different shoppers — a flagship store, a wholesale account, and e-commerce can each skew to different sizes. A single global curve over-buys some sizes and under-buys others, so curves are best set by the segment that actually buys the product.
- What happens if the size curve is wrong?
- If the curve is off, core sizes sell out early while fringe sizes sit, so you lose full-price sales on the sizes customers wanted and take markdown on the ones they did not. Getting the curve close protects both sell-through and margin.
- A size curve is the percentage of demand for each size within a style; the percentages sum to 100%.
- Multiply total planned units by each size’s share to get size-level units that match real demand.
- Base the curve on comparable history, and re-balance rounded quantities back to the total buy.
- Size curves vary by region, channel, and product — a single global curve over- and under-buys sizes.
- RetailNorthstar applies size curves automatically as buys are built and carries them into purchase orders.
Your calculator result is one number. RetailNorthstar keeps the whole plan connected — line plan, OTB, assortment, buy, POs, and production.