Print on demand, and when to stop

There is a moment most apparel brands hit where someone looks at what the print partner takes per shirt, works out what a bulk run of the same shirt would cost, and decides to order five hundred. Sometimes they are right. Often they have just bought a year's worth of a design that sold for six weeks.

We have sat on every side of this. We have printed to order, we have held stock by the pallet, and we built a print-on-demand app that other brands run on. So this is not a pitch for either. It is what we would tell a mate.

What print on demand actually buys you

Not cheap shirts. Freedom to be wrong. Every design you launch costs you nothing until someone buys it, so you can put out the one you are not sure about. Every size is decided by the buyer, so you never sit on a box of larges. Every month you do not carry stock is a month that cash is somewhere more useful.

The price of that is margin per unit, and a partner between you and your customer. For a brand that is still finding out which designs are its designs, that trade is nearly always worth it. The margin you would save by ordering in bulk is smaller than the money you would lose being wrong about what to order.

What stock actually buys you

Margin, mostly. A run of two hundred costs a fraction per unit of what one costs. And control: you pick the blank, you see every print before it goes out, and you ship the day the order lands because the shirt is already on the shelf. There is a reason the brands you admire hold stock. Past a certain point it is the only way the numbers work.

When to make the move

Two signals, and you want both. The first is repeat designs. If most of what you sell is the same twenty things month after month, you are paying on-demand prices for what is really a stock business. The second is a size curve you can predict. If you can say, without looking it up, that a launch will sell mostly mediums and larges with a few of everything else, you are ready to order to that curve.

One signal alone is not enough. Repeat designs with a size curve that swings around means you will run out of one size and drown in another. A steady curve on a range that changes every month means you are forecasting for designs you will have replaced by the time the stock lands.

What stock actually costs

The print run is the cheapest part. Everyone budgets for the print run. Here is the rest of the list, and we have paid for every line of it.

• Forecasting, and you have to be good at it. Not roughly right. Good. Every unit you over-order is cash on a shelf, and every one you under-order is a sale you paid marketing to lose.
• Launches run like a campaign. Stock has a clock on it from the day it arrives. A launch that slips a fortnight, or goes out soft, is not a slow week. It is a pallet getting older.
• Marketing that is dialled in, because the run only pays off if it sells through. On demand, a design that flops costs you a graphic. On stock, it costs you the graphic and four hundred shirts.
• Somewhere to put it, and someone to count it. Storage, picking, packing, a stocktake that is right, and the day you discover it was not.
• People. Buying, receiving, warehousing and fulfilment are jobs. Even when a 3PL does the hands-on part, someone at your end owns the numbers.
• A way out for what does not sell. Every stocked brand ends up with a corner of shirts that were a good idea in March. Outlet, bundle, sample sale, donation. Decide the exit before you order, because you will need it.
• Cash. The bulk run is paid for before the first unit sells. Print on demand let you get paid first. Stock reverses that, and the gap between the two is your working capital.

The honest margin sum

Work out what you pay the partner per unit. Then work out what a unit costs stocked, once you include the run, the storage, the people, the cash you tied up, and the units you will end up clearing at cost. Do this properly and the gap between the two numbers is smaller than the first sum suggested. It is still a gap, and at volume it is a real one. But it is bought with forecasting, discipline and working capital. It is not found.

On demand, being wrong costs you a design. On stock, it costs you the design and a pallet.

Finding the balance

Most brands that get this right do not pick one. They stock the twenty designs they would bet on, in the sizes they would bet on, and leave everything else on demand. The catalogue stays wide because the long tail costs nothing to keep. The core stays cheap because it is printed in bulk. A new design goes out on demand, and if it earns its place it moves to the stocked side on the next run.

There is no shame in either direction, and it is not a one-way door. The mistake is ordering the pallet before both signals are there, or staying on demand a year after they have arrived because a purchase order feels like a commitment. It is one. So is a design that has sold every week for a year and is still being made one at a time.