Sizecurve Watch the 3-minute demo We scanned 298 catalogues Which size goes first? Preview — synthetic store, 1,350 orders over 60 days

What to reorder, in which sizes.

Size curves built from demand net of returns, not gross sales. Broken size runs flagged before the stock goes dead. Purchase orders by size, with every quantity showing the history behind it.

Install free from the Shopify App Store → A read-only Shopify app for apparel brands. Free to find broken size runs; $29 a month for reorder quantities, with a 14-day free trial.

Below: a preview on a generated store. The free size split calculator runs on your own numbers, in your browser.

471units stranded in broken size runs
15%of units returned — 203 of 1,350
519units to order across 8 styles
2styles too new to forecast alone

Check your own store for broken size runs → Reads your public catalogue, names the styles, no login and no install.

What needs attention

Every style

StyleSize curve — net of returnsDemandReturnsTo order
Heavyweight Teebroken run
SMLXL
SMLXL
245net 10% 189
Oxford Shirt
SMLXL
SMLXL
141net 15% 28
Slim Jean
SML
SMLXL
138net 32% 78
Wool Overcoat
SMLXL
SMLXLXXL
58.04net 8% 33
Ribbed Tank
SMLXL
XSSMLXLXXL
128net 9% 61
Pleated Chinobroken run
SMLXL
SMLXL
101net 14% 71
Knit Polo
SMLXL
SMLXL
90net 12% 10
Linen Shirt
SMLXL
SMLXLXXL
64net 20% —
Crewneck Sweat
SMLXL
SMLXL
129net 11% 49
Field Jacket
SMLXL
SMLXL
41net 16% —
Pleated Trouser (new)borrowed curve
SMLXL
SMLXL
9net 10% —
Shawl Cardigan (new)borrowed curve
SMLXL
SMLXL
5net 17% —

Hatched segments are sizes that are out of stock. A style is only called broken when the missing sizes carry more than 30% of its own curve and stock is still stranded in the sizes that remain — not simply because something sold out.

Purchase orders Download CSV

Heavyweight Tee — 189 units

245 net units in the window is 4.1 a day. 75 days of lead time and cover needs 306 to hold. Each size takes its share of that, less what it has on hand and on order, never below zero.

SizeShareReturnsOn handOn orderOrder
S16.7%5%45—6
M25.7%14%0—79
L33.9%5%0—104
Oxford Shirt — 28 units

141 net units in the window is 2.4 a day. 75 days of lead time and cover needs 176 to hold. Each size takes its share of that, less what it has on hand and on order, never below zero.

SizeShareReturnsOn handOn orderOrder
M28.4%15%31—19
L30.5%14%45—9
Slim Jean — 78 units

138 net units in the window is 2.3 a day. 75 days of lead time and cover needs 173 to hold. Each size takes its share of that, less what it has on hand and on order, never below zero.

SizeShareReturnsOn handOn orderOrder
S18.1%36%6—25
M28.3%40%16—33
L30.4%21%33—20
Wool Overcoat — 33 units

58 net units in the window is 1.0 a day. 75 days of lead time and cover needs 73 to hold. Each size takes its share of that, less what it has on hand and on order, never below zero.

SizeShareReturnsOn handOn orderOrder
M34.5%13%20—5
L29.3%11%12—9
XL · sold out since 2026-08-3113.8%0%0—10
XXL · sold out since 2026-08-3012.1%0%0—9
Ribbed Tank — 61 units

128 net units in the window is 2.1 a day. 75 days of lead time and cover needs 160 to hold. Each size takes its share of that, less what it has on hand and on order, never below zero.

SizeShareReturnsOn handOn orderOrder
S19.5%7%21—10
M28.1%8%6—39
XL12.5%6%8—12

Styles with too little history of their own borrow a curve pooled from styles that do have it — weighted by units sold, and never from other thin styles. Those quantities are marked borrowed curve so you can see which numbers are estimates before you spend money. When there is no pool to borrow from, the mark is too little history instead: that quantity came from the style's own handful of units and is the weakest number on this page.

Why net of returns

A style selling 30 M and 40 L, with half the L coming back, reads as L-dominant on gross sales and M-dominant net of returns. Forecasting on gross sales over-orders exactly the sizes you are drowning in, and the error compounds every cycle. Refunds are attributed to the period of the sale, not the refund, so a January mistake does not corrupt March's seasonality.