Selling Colour Without Holding Colour Stock

A buyer called us in August asking what our minimum was on colour. Not on a style, not on a cloth weight, on colour specifically. She had seen teal and aubergine turning up across the autumn/winter 26/27 forecasts and wanted to put a version in front of her list before a competitor did, but the maths she had run assumed she would need to commit to a run before we would touch it. That is how colour has always worked in apparel: a mill dyes a batch, a brand buys enough of it to be worth the mill's time, and if the shade does not sell, the brand owns the leftover fabric until it marks it down or writes it off. She was pricing a bet she did not need to make.

The confusion is reasonable. Almost everyone who has bought fabric before has bought it by the roll, and a roll has a minimum length because a dye lot has a minimum size worth running. What she had not separated out is that the colour test and the fabric purchase are not the same transaction here, because the cloth she wanted was already sitting in a mill's book. Nobody needed to dye anything.

The stock risk was never the colour, it was the roll

Picture the two ways a new shade can reach a rack. In the first, a brand asks a mill to dye a colour that does not exist yet, in a quantity the mill sets, because a dye vat has a minimum worth firing up. That length becomes the brand's problem the moment it's cut, sold or not. In the second, a mill has already dyed and woven the colour as part of its standing range, holds it as stock cloth, and will sell a length as short as a single garment's worth to anyone who asks. The first path carries real stock risk. The second one does not, because the mill absorbed that risk when it built the colour into its book in the first place, priced across everyone who might eventually buy a metre of it.

Our minimum order is one piece per style. That number describes the garment, not the fabric, and it only works because most of what a buyer wants to test already exists as a bunch on a shelf rather than a colour somebody has to invent. Teal, aubergine, chocolate brown and the deeper burgundy shades that WGSN and Coloro named among the key colours for autumn/winter 26/27 are not new inventions. They are variations a mill has already woven into a book that also holds navy and grey, at the same weight, from the same loom. A single garment cut from that book costs what a single garment costs. No roll gets bought on your account, because it was already bought, in bulk, by the mill.

What actually happens at the cutting table

When an order comes in for a colour we already hold, the process looks identical to an order for navy. The cloth is on the shelf, it goes to the cutter, the piece is measured against your spec sheet, and it ships inside our normal 14 to 20 working day window. Nothing about the shade changes the sequence. We keep Zegna's Trofeo and Traveller books and Vitale Barberis Canonico's Four Seasons book cut and ready in Shanghai for exactly this reason, across 202 styles between the two, so a colour test in either book starts moving the same day it's confirmed.

A colour from one of our other 33 mills and merchants works the same way in every respect except timing. That cloth is not sitting on our shelf, so it's ordered in against your confirmed order, and the mill's own shipping time sits on top of our production window before the piece is cut. We wrote about that gap in more detail in held versus ordered in, because it changes your timeline more than it changes your risk. Either way, the order is still one piece. The minimum never grows because the shade you picked lives in a book we ordered rather than a book on our own shelf.

Where a colour still costs you something

I want to be direct about the one place this stops being free, because pretending otherwise would be the kind of thing that comes back on an invoice later. If the colour you want does not exist in any mill's book yet, testing it is a different order altogether. A custom shade means asking a mill to run a dye batch specifically for you, and that batch has a minimum length set by the mill's own equipment, in metres, not in garments. That minimum does not bend for a small brand any more than it bends for a large one, because the cost the mill is covering is the vat, not the buyer.

Practically, this means the decision that matters is not "can I test this colour" but "does this colour already exist somewhere in a supplier's range." A quick way to check before you ask us for a quote:

• Ask which mill's book the shade is closest to, by name and bunch number, not by a Pantone reference alone.
• Ask whether that mill's cloth is held in Shanghai or ordered in, because that changes your timeline, not your minimum.
• If nothing close exists, ask what the mill's own minimum length is for a custom dye run before you build a launch plan around the colour.
• Order one piece in the closest existing shade first, even if it isn't a perfect match, before committing to a custom run of anything.
• Get the bunch number in writing on your invoice. A bunch number is checkable against the swatch in your hand; a colour name on its own is not.

Testing one piece is not the same as ordering fifty

There is a second question that shows up right after the first one, usually a few weeks later, once the single test piece has come back and the reaction was good. If you go from one piece in the new colour to a real batch, does anything change? Mostly not, but one thing is worth knowing before you place that second order rather than after.

A single garment can prove the shade reads well, the cloth handles the way you expected, and your list responds to it. It cannot prove dye lot consistency across a run, because that question only exists once there is more than one piece to compare side by side. Two lengths cut from the same bunch at different times can sit a shade apart even when both are correctly labelled with the same colour name, which is why a batch order is worth confirming against a single reference length rather than against a colour description alone. Ask for everything in a batch to be cut from one length where the mill can offer it, and ask what happens if it can't, before the pieces are already in production. This is a smaller version of the same discipline that applies to a first sample generally: one piece answers some questions and leaves others for the batch that follows it.

None of this changes the minimum. A batch order still starts from the same one-piece floor, and nothing forces you to order more than a handful once the colour has proven itself with your customers. It just means the second order deserves the same five minutes of checking as the first one did, for a different reason.

Why this matters more for colour than for cut

A cut or a construction choice is forgiving in a way colour is not, because a buyer can usually tell from a photo whether a silhouette will land with their list. Colour is the one variable that genuinely needs to be seen on a body, under real light, before anyone commits to featuring it. That is precisely the variable a stock-heavy model makes hardest to test cheaply, and precisely the one a one-piece minimum was built to de-risk. The buyer who called in August didn't need a smaller minimum. She needed to know the minimum she already had covered the question she was actually asking.

We laid out the wider version of this argument, on measurement rather than colour, in what actually breaks a no-inventory MTM launch: the stock problem was solved years ago, and what's left is the parts a one-piece minimum doesn't automatically fix for you. Colour is one of the few places where the minimum really does do all the work, provided the shade is already sitting in a book somewhere. Check that first, and the bet you were pricing turns out to be a lot smaller than it looked.

If you have a shade in mind and aren't sure whose book it's closest to, send us the reference and we'll tell you whether it's a one-piece test or a mill's dye minimum, before you build a season around either answer.