On making less

The first spreadsheet we built after leaving the consultancy had one column that embarrassed everyone in the room: full-price sell-through. Across the labels we had worked on, it hovered around thirty-four percent. Two thirds of everything we drew, sampled, graded, produced, shipped, photographed and merchandised was sold at a discount or not at all.

That is not a design failure. It is a calendar failure. A season demands newness on a fixed date whether or not anyone has had a better idea, so you fill the gap with variations: the same coat in four lengths, the same knit in nine colours. Each variation splits the demand for the others, and every split makes the forecast worse.

What we actually changed

In January 2022 we cut ninety styles to twenty-two, and told our three largest stockists that we would not be showing in Paris that June. Two of them stayed. The one that left was, at the time, nineteen percent of revenue.

The rules we wrote down that month have barely changed since:

• A piece stays in the collection until something demonstrably better replaces it.
• A new piece must retire an existing one. The collection does not grow by addition.
• No markdowns. If something is not selling, it is either wrong or it is priced wrong, and both are our problem to fix at the source.
• Every fabric names its mill and its town on the product page.

Revenue fell thirty-one percent in the first year. We had modelled twenty. It was the right decision anyway.

The part nobody warns you about

Making less does not make the work easier — it concentrates it. When you produce five garments instead of ninety, every one of them is inspected by the person who drew it, and there is nowhere for a mediocre piece to hide inside an assortment. The Volta Overcoat went through eleven fittings. The version before it went through three, and you can feel the difference in the shoulder.

It also changes what you argue about internally. We used to argue about newness. Now we argue about whether a two-millimetre change to a collar stand is worth re-cutting the pattern. It is a much better argument to be having.

Where it landed

Revenue passed the old peak in late 2023. Full-price sell-through is now above eighty percent, returns are down by roughly a third, and the repair bench — which we expected to be a cost centre and a gesture — turned out to be the single best-reviewed thing we do.

We are not going to pretend this is a universal prescription. It works because the collection is small enough to hold in your head and because our suppliers agreed to run quantities most would refuse. But the underlying question is portable, and it is worth asking out loud: how much of what you make is only there because the calendar said so?