The style that sold 60 pieces and lost money

The arithmetic that gets hidden

A style comes in at 80 pieces across five sizes. It sells 60. On any report the shop is looking at, that is a success and worth repeating.

Read by size, it is not a success at all. Medium and large sold out, probably early, and the shop lost sales it never knew about. The twenty pieces sitting in S, XL and XXL will be marked down at season end, and the markdown eats most of what the fast sizes earned.

Reorder that style on its headline number and the same thing happens again.

The data already exists

Worth being clear about what is actually broken here. The shop is not missing information. Every one of those sixty sales was rung up with a size on the label. The information passed through the counter and was not kept in a form anyone could read back.

That is why this is worth doing even for a small shop. It is not new data collection. It is retaining what already moves through the till.

Where the work actually lands

Not at the counter, which surprises people. Billing at variant level is barely slower than billing at style level once barcodes are on.

The work moves to receiving. A delivery booked in as "40 pieces of style X" undoes the whole thing, and the person in the back room now has to break that down by size and colour. That is a genuine change in someone's job, it takes longer than what they did before, and it is the step that quietly kills this in most shops.

Any shop considering this should decide who does that and whether they have time, before buying anything.

One category at a time

1. Pick the highest-value category, not the largest. Fewer lines, more learning per line.
2. Count it properly at variant level once. This is the tedious part and there is no way around it.
3. Bill that category at variant level for a month, then physically recount and compare.
4. Change receiving to variant level for that category's next purchase order.
5. Add the next category when the first one's count still agrees after a month.