Real-time fashion is here. It will eat your lunch if you are not watching.

You place the order months before anyone will wear it. Ten thousand units, chosen from a sample rail last winter, shipped to a warehouse, and then the wait. Some of it flies. A lot of it sits, gets marked down, and quietly loses money. That has been the deal in fashion for decades, and in most businesses the shape of it is familiar. Commit early, at scale, on a guess.

A new kind of competitor has stopped taking that deal. It does not guess better than you. It has made guessing so cheap that it barely needs to guess at all. That is the real story behind the brand your customers keep mentioning, and it is worth understanding before it reaches your market, because the model travels a long way beyond clothes.

Why does the old fashion playbook suddenly feel slow?

Because the gap it was built around has almost closed. Fashion used to be made to order, one garment at a time. Then ready-to-wear put standard sizes on a rail. Then fast fashion, led by Zara and H&M, cut the cycle from a year to a few weeks and taught us to expect new stock all the time. Each step shortened the distance between a trend appearing and a product arriving. This is the quiet death of an operating model, the same shift underneath the end of business as usual.

Shein removed most of what was left. A design does not start with ten thousand pieces. It starts with a hundred, sometimes two hundred, put online to see what people actually do (Shein Group). The failures cost almost nothing. The moment something moves, the software reorders it across a network of small workshops, and the winners scale within days. An investigation by Rest of World found Shein adding between 2,000 and 10,000 new styles to its app every day across the second half of 2021. A traditional retailer plans a whole season around a few thousand.

It did not make guessing better. It made guessing cheap.

What is the machine really competing on?

Not taste. Not the best designer in the room. It competes on the speed of the loop between a signal and a response. It sees what people want almost as they want it, makes a little, learns, and then makes more of only the right things. The algorithm does not replace the designer. It decides how much risk the supply chain is allowed to take.

That is why this matters even if you never sell a shirt. Every business runs on a version of the same loop. Notice what the market wants, decide what to make, find out if you were right. The firm that closes that loop in hours will out-learn the firm that closes it in quarters, whatever the two of them happen to sell, because the gap keeps compounding in the fast learner's favour. The advantage is no longer the size of your bet. It is the speed of your learning, which is really a question of the loop you design.

Does faster have to mean more waste?

Right now, it usually does. The same speed that clears dead stock also feeds a habit of buying and binning. The wider industry already carries a heavy cost. The fashion sector produces an estimated 2 to 8 per cent of global carbon emissions and drinks water by the trillions of litres (UN Environment Programme). The Ellen MacArthur Foundation put it starkly. The equivalent of a rubbish truck of textiles is landfilled or burned every second, and less than one per cent of clothing is recycled into new clothing.

Here is the turn. The machinery itself is neutral. Making a hundred pieces and reordering only what sells is, in principle, the opposite of overproduction. The technology that made disposable fashion cheaper is the same technology that could make responsible fashion pay. What it serves is a choice, not a fixed outcome.

How do you use the same speed to build something better?

You point the loop at a different target. Instead of racing to the lowest price and the highest churn, you use real-time demand to make closer to what people will actually keep rather than what a forecast merely hoped they might. Make less, make it later, and make it nearer to the customer. McKinsey describes this exact shift, using analytics and on-demand production to match supply to real demand rather than to a forecast set a year out.

Then you pair the speed with substance. Materials that last. A supply chain you are willing to show. The honesty to let a customer see your working. The lower-altitude questions, how do we copy their prices, how do we churn faster, lead nowhere good. The better question is how you build the fastest honest feedback loop in your market, and that is worth a proper conversation, because the answer is specific to your business. It is the same instinct behind building rather than simply cutting.

What could a fast and responsible brand look like?

Picture a business that makes almost nothing on a guess. It reads real demand, produces in small runs, and scales only what people truly want, so the markdown rail and the landfill truck both shrink. Its supply chain is something it will show you, not something it hides. It moves at the speed of the trend, and it can still look its customers, its workers and its own conscience in the eye.

That is not a softer, slower version of fashion. It is the same engine the real-time players built, turned toward keeping rather than discarding. Technology's highest purpose is to serve human life, and the brands that hold onto that will carry both the speed and the story. The machine rewards the fastest learner. Nothing decrees that the fastest learner has to be the most wasteful one.

About the author

Thomas Green

British technology futurist, AI keynote speaker and advisor. Thirty years across enterprise technology and AI strategy, helping leaders navigate the future of work. The futurist who died.