How to Split Sizes on a First Order
By The Velocity Wear Team
The size split is the part of a first order that gets the least thought and causes the most dead stock. It defaults to an even spread because that feels neutral and requires no decisions, and an even spread is almost always the wrong answer — it guarantees you run out in the middle while holding the ends.
Demand is middle-weighted
Across almost every apparel category, sales cluster in the middle sizes. A five-size range typically sells far more M and L than XS and XXL, and the shape of that curve is the single most useful piece of inventory knowledge a brand has.
A reasonable starting point without data: a small number at each extreme, a solid block at each of the sizes either side of centre, and the largest quantities in the middle. Something like a 1 : 2 : 3 : 3 : 2 : 1 shape across a six-size range is a defensible first guess.
Then shift it for your actual audience
That generic curve is a starting point, not an answer, and the shift matters more than the shape. Menswear-leaning streetwear skews larger. Women's studio and fitness brands skew smaller. Merch tables and band audiences skew notably larger. Workwear skews larger still.
If you already have an audience — a following, a gym membership, a club roster, an email list — you have real information that no general curve provides. Use it. A gym owner knows their members; a club knows its roster; a brand with an engaged following can simply ask.
Fit changes the curve
A specific effect that catches brands out: if a garment is designed oversized, customers size down, and the curve moves. An intentionally boxy fit sold to people who normally wear a large will sell more mediums than the same brand's regular-fit product.
Which is another argument for saying explicitly on the product page what the intended fit is and whether to take your usual size. Ambiguity there does not just cause returns, it distorts your next order.
Mixed sizes inside the minimum
Mixed sizes are included inside our 20-piece minimum per design, which is what makes any of this possible on a first run. A minimum applied per size would force a brand into either a narrow range or a much larger order, and both of those are worse than carrying a realistic curve at low volume.
Twenty pieces across six sizes is not depth in anything, and it is not meant to be. It is enough to find out where the demand actually sits.
Reading the result properly
- Track sell-through per size, not just overall. An aggregate of 70% can hide mediums selling out in a fortnight while XXLs never moved.
- A size that sold out is not necessarily a size that had more demand — it may just have been under-ordered. Note when it sold out, not only that it did.
- A size that did not move is clearer information, provided it was actually available for the whole period.
- Do not over-correct on one run. Adjust toward what sold rather than swinging the whole curve to match one month.
The reorder is where you get paid for this
The first order is a probe. The reorder is where the size split becomes real money, because you are now ordering against evidence instead of a general curve.
This is the main thing a reorder should change. Same garment, same print, same colour — a different distribution. Get that right two or three times and your dead stock problem largely disappears, which is a bigger effect on profitability than most brands get from negotiating unit price.
