Sold out in a 10 and 12, drowning in 6s and 18s
At the end of the season, the leftover stock tells the story. Every style has a few size 6s and 18s left, and every bestseller sold out in 10 and 12 weeks before the end. The size split on your purchase orders is the same for every style: something like 1, 2, 3, 3, 2, 1. Someone set it up years ago, and it has been copied onto every order since.
Meanwhile your jeans sell to a different size curve from your dresses, and your knitwear in S, M, L sells differently again. The single ratio suits none of them.
Why the ratio never gets fixed
- Sales reports show what sold, not what would have sold if the size had been in stock.
- Returns are not taken off, so bracketed sizes look more popular than they are.
- Size curves differ by range, but most brands use one ratio for all.
- The buyer has no time to rebuild a curve before each factory order.
The first point matters most. If the 10 sold out on day twenty, the sales report only shows twenty days of 10s. The data understates exactly the sizes you need more of.
The cost of the wrong split
| Result | Effect |
|---|---|
| Middle sizes sell out early | Lost sales on your best styles |
| End sizes left over | Markdowns and clearance |
| Broken size runs | Styles that look available but are not |
| Reorders of one size | Factory minimums make small top-ups expensive |
How we build size curves you can use
- We pull sales by style, colour and size from your store and any other channels, net of returns, so the curve reflects what customers kept.
- For each size, we note the days it was out of stock, and estimate what it would have sold in that time from its own sales speed while it was available. The adjustment is shown separately so your buyer can see it.
- Curves are built per range, and per fit where it matters, because a slim tailored trouser sells differently from a relaxed one.
- For a new style, the buyer picks the range or a similar past style, and the tool suggests a split for the order quantity, rounded to the factory's pack or ratio rules.
- The buyer adjusts and approves the split, and it can flow into the purchase order.
- After the season, a report compares the ratio ordered with the ratio sold, so the curves improve each time.
When the buyer sits down to order
The buyer is placing an order for a new wide-leg trouser. They choose 'trousers, relaxed fit' as the reference, enter the total quantity and see a suggested split with the reasoning next to it: past sales, returns taken off, adjustment for sell-outs. They move a few units from the 18 to the 12 because they know the fabric runs small, then approve it. Nobody has to build pivot tables from scratch.
Honest caveats
The tool does not know about things that are not in the data, like a new customer group you are trying to reach or a sizing change to the pattern. That is why the buyer always has the final say. Small ranges with little history get a curve borrowed from a similar range until they build their own. Extended sizes launched recently need special care, because low early sales may reflect low stock rather than low demand.
Signs your size ratios need work
- The same size split goes on every purchase order.
- Middle sizes sell out while end sizes are left over.
- Your size data does not account for sell-outs or returns.
- Jeans, dresses and knitwear share one ratio.
- Reorders are for single sizes that ran out.