Take a plain knitted shirt and follow it backwards. The yarn was spun somewhere with room for bales and machines that run for days. The fabric was knitted on large circular machines somewhere else, in a building with the height and the floor strength to carry them. It was dyed and finished in a third place, with water, steam, drainage and treatment on site. It was cut and sewn in a fourth, surrounded by people and by rooms full of spare needles and thread. The buttons, labels, elastic and polybags came from a market district where a hundred variants are sitting on shelves. Then it was packed, and only then did it become one object with one brand on it.
The obvious explanation is that labour is cheaper in some towns than others. That is part of it, but it misses the more interesting reason: each of these steps wants a building that is actively wrong for the others. Putting them all under one roof sounds efficient until you price the roof.
The machine decides the building
A large circular knitting machine is tall, heavy, and vibrates. It wants a high ceiling, a strong floor, and a steady supply of yarn cones that can be wheeled to it. It also runs best when it runs continuously, because stopping and restarting a fabric changes the tension and leaves a visible band. A room full of these machines is a factory designed around vertical space and uninterrupted rotation.
A sewing floor is almost the opposite. It wants dense rows of light machines, good daylight or even lighting, and a lot of people close together, with easy access in and out. Floor strength matters far less than circulation. It is a horizontal operation, measured in minutes per garment rather than hours per roll. You could put both in one building, but you would be paying for a high, reinforced hall in the rooms where nobody needs it, and squeezing people into a space designed for machines.
Dyeing goes where the water is allowed to go
The decisive split is wet processing. Dyeing and finishing consume water, produce warm effluent, need steam, and require treatment and discharge permissions that are not granted to just any address. A workshop that dyes fabric is tied to its pipes, its permits and its treatment plant in a way a sewing room is not. It also wants to run large batches: a dye vessel is cheapest when it is full, and a colour is most repeatable when one batch is large enough to cover a whole order.
This is the step that most often sets the pace. A shirt can be cut in an afternoon, but the fabric may be waiting for a dye vessel of the right size to come free, and then for a finishing range to set the width and the shrink. When a clothing order is described as delayed, a surprising amount of the delay is not people sewing slowly. It is cloth sitting between two wet processes that are booked in batches.
Why two "identical" shirts are slightly different
Dye lots are batches, and batches are not perfectly identical. Fibre from different bales takes dye a little differently, and a second vessel is a second chemical event even when the recipe is unchanged. The industry manages this by cutting a whole garment from one lot and by keeping roll and lot records, but if you buy two shirts of the same colour months apart and hold them together in daylight, you may see a faint difference at the seam or under the collar.
That is not necessarily a defect and not necessarily a fake. It is the visible edge of batch processing. The same logic explains why a matching set is usually cut together: the only reliable way to guarantee a match is to avoid the comparison entirely by using the same roll. Colour consistency is a logistical achievement before it is a chemical one.
The trims are a catalogue, not a purchase
A shirt is also a collection of small decisions: ribbing at the neck, a label at the back, a button, a thread colour, a size sticker, a bag. None of these is worth making for a single order, and all of them have to arrive before the sewing line starts. They come from districts that function as living catalogues, where a designer can choose among existing options rather than commissioning new ones.
That changes what a design is. A great deal of the variation between two shirts is selection from what is already on a shelf, not invention. It is the same principle that makes a toy district fast: shared stock makes short runs possible. The difference is that here the shared stock is a set of tiny components whose variety is effectively free to browse.
What the washing label does not say
The label tells you the fibre content and how not to ruin the garment. It does not tell you that the fabric was knitted in one town under a high ceiling, dyed in another beside a treatment plant, cut in a third, and given its buttons from a market two streets away from the packing table. It also does not tell you that most of the calendar between order and delivery was spent waiting for a vessel, a roll, or a lorry between towns.
Which is a useful thing to know the next time a garment seems inexplicably cheap. It is cheap partly because no single building had to be all of those things at once. The region is the factory, and the road between its towns is one of the machines.