The return rate you can see is already out of date

Every fashion business watches its return rate. Very few watch it early enough to do anything about it.

The problem is structural. A return rate is only final once the return window has closed, and for a marketplace order that can be sixty days or more after the sale. By then the season has moved on, the reorder has been placed and the next drop has been costed. The number arrives accurate and useless.

What you need is the rate a product is heading towards, measured while the cohort is still open. Three things make that possible.

Read the cohort, not the calendar. A rate calculated across everything sold this month mixes orders that are three days old with orders that are three weeks old. The young orders have not had time to come back, so the rate looks flatteringly low. Group orders by the week they were placed and every group can be compared like for like.

Watch the shape of the arrivals. Returns do not arrive evenly. Every category has a curve, and once you know the shape of yours you can look at a half-finished cohort and see where it lands. A product returning hard in week two is telling you something a settled rate will only confirm two months later.

Treat forty per cent as a diagnosis, not a discovery. By the time a rate reads forty, the only decision left on that buy is markdown. Every decision the number should have informed, the size mix, the reorder quantity, the channel it went to, was made weeks earlier when the rate still looked like eleven.

None of this needs new data. It needs the returns you already hold, joined back to the orders that produced them and grouped by cohort rather than by calendar month. That is the view Reportly is built around.

‍

By the time a return rate reads forty per cent, it is already too late. The decision it should have informed was made weeks ago.

The Reportly team

What this looks like in practice

Open a product in Reportly and the returns panel shows two numbers side by side: the rate booked so far, and the rate the cohort is heading towards once its window closes. Beneath them sit the reason customers give and the size doing the damage.

If the evidence is too thin to call, it says so rather than printing a confident number built on nine units. A rate you can trust at week two is worth more than a perfect one at week ten.