Shein

The physical heart of the model sits in the manufacturing cluster around Guangzhou, where a large ecosystem of small garment workshops has grown up around Shein. These are not the sprawling mega-factories that stitch millions of identical T-shirts for a single Western brand. Many are modest operations, sometimes only a few dozen machines, and their competitive edge is not scale but flexibility — the ability to run a tiny order of two hundred pieces one week and a larger repeat the next, switching styles constantly.

The trick that ties them together is software. Shein connects its suppliers to a shared system that pushes orders, tracks inventory, and feeds back sales data almost continuously. A workshop doesn't wait for a seasonal plan; it receives a stream of small, specific instructions and is expected to turn them around fast. This is closer to how a tech company manages a logistics network than to how a traditional garment brand runs its supply chain, and it's the reason the days-not-weeks clock is even conceivable.

A three-dollar top is a genuinely strange object. Someone designed it, someone cut and sewed the fabric, it was photographed, listed, packed, and — this is the part that breaks the older logic — flown, often individually, from a warehouse in China to a doorstep in Europe or America. Shein leaned heavily on air freight and on a customs quirk: in the United States, the so-called de minimis rule long allowed low-value parcels, under roughly $800, to enter without duties. Shipping millions of small packages directly to shoppers let the company skip much of the tariff and warehousing cost a conventional importer would carry.

That arrangement drew scrutiny. The de minimis exemption became a political target precisely because it let ultra-cheap direct shipments flood in duty-free, and moves to close or narrow it in the mid-2020s directly threatened one of the model's cost advantages. The prices, it turned out, depended partly on a regulatory loophole as much as on manufacturing cleverness — a reminder that "efficient" and "cheap" are not always the same thing.

Step back from the app and Shein looks less like a clothing company than a demonstration of what happens when a supply chain is rebuilt entirely around demand data. The founding insight — let buyers decide what gets made, in real time — is genuinely elegant, and in narrow terms it reduces waste. Nothing sits unsold because nothing is made until it sells. If you judged the model only by the standard the old industry set for itself, it would look like a solution.

The trouble is that the standard changed. By making each individual garment so cheap to produce and so cheap to buy, the model doesn't just satisfy demand — it manufactures more of it. When a top costs less than lunch, the calculation a shopper makes shifts. Buying becomes casual, keeping becomes optional, and the total number of garments flowing through the world climbs. Global clothing production has roughly doubled over the past couple of decades while the average garment is worn far fewer times before disposal, and ultra-fast fashion is both a symptom and an accelerant of that curve.

Go back to that workshop near Guangzhou, waiting to see whether last night's dress earns a repeat order. What's remarkable about the scene is how little guessing it contains. The design was cheap to try, the first batch was small, the app did the deciding, and the factory reacted rather than forecast. Shein didn't invent cheap clothes or fast turnaround; it wired the whole chain together so tightly that the gap between a shopper's tap and a sewing machine's start shrank to almost nothing. On its own terms — waste from unsold stock, capital tied up in wrong guesses — the model works beautifully.