Size Curve

What is a size curve?

A size curve is the planned distribution of units across sizes for a given style or category, expressed as percentages of total units purchased. It determines how many units of each size are bought and how those units are allocated across channels and doors.

For example, a size curve for a women's contemporary top might specify:

| Size | % of Total Units | |------|-----------------| | XS | 10% | | S | 25% | | M | 30% | | L | 22% | | XL | 13% |

Applied to a buy of 500 units, this curve produces: 50 XS, 125 S, 150 M, 110 L, 65 XL.

Why size curves matter in apparel buying

Size curves are one of the most consequential — and most error-prone — elements of apparel buy planning. The size distribution of a purchase directly determines the sell-through rate at the size level and the residual inventory profile at season end.

Incorrect size curves create two simultaneous problems:

1. →

   Stockouts in fast-moving sizes: If a size sells faster than the curve predicted, the brand loses full-price revenue in that size before the selling window closes. This is unrecoverable — you cannot reorder a fashion style on a short lead time.

2. →

   Residual inventory in slow-moving sizes: Over-bought sizes must be cleared through markdown, transfer, or end-of-season liquidation — all at a margin cost.

Because apparel buy decisions are made months in advance with no ability to quickly reorder, size curve accuracy is determined entirely at the planning stage. A size curve error is baked in the moment the PO is submitted.

How size curves are built

Size curves are derived from historical sell-through data — specifically, which sizes sold through fastest and which accumulated residual inventory across previous seasons and styles.

The most accurate size curves are built at a granular level:

• →By category — Women's tops have different size distributions than women's bottoms. Outerwear curves differ from knitwear.
• →By channel — DTC customers in a brand's own online store may skew differently than wholesale customers at a department store. A DTC-specific size curve for a style sold online only will be more accurate than a blended curve.
• →By demographic and region — A brand with a significant customer base in extended sizes (1X–3X) needs size curves that reflect that demand, not just the straight-size distribution.
• →By historical sell-through signal — If size L has accumulated residual inventory for three consecutive seasons in a category, the curve should be revised downward for L before the next buy.

Size curves vs size scaling

These terms are sometimes confused:

Size curve refers to the planned percentage distribution of units across sizes. It answers: of 500 units, how many in each size?

Size scaling or size grading refers to the physical construction of the garment at each size — how the pattern changes from XS to XL. This is a product development concept, not a planning concept.

In apparel planning, "size curve" always refers to the buying quantity distribution.

Size curve errors and their cost

The most common size curve error is applying a generic or category-level curve to styles that have meaningfully different demand profiles. A missy-fit trouser and a slim-fit trouser may share a category but have very different size distributions. Using the same curve for both will produce residual inventory on the under-selling sizes and stockouts on the over-selling ones.

A secondary error is failing to update size curves based on recent sell-through data. If a brand updates its size curves once per year using aggregate historical data, it may miss shifts in customer demand that happened within the past two seasons — particularly relevant in contemporary and trend-driven categories where body proportion preferences can shift.