What’s happening
Perhaps never before in history, apart from war and covid, have brands had to reorganize and respond to the market so quickly.
This time, the change didn’t come from a supply chain crisis, but from something far more intimate: the very body of the person buying.
Drugs like Ozempic and Mounjaro are changing customers’ sizes faster than ever. And since I’m inside brands’ operations every day, I get to see up close how a brand struggles to decide what to produce and in which sizes, while its customers keep losing weight in the meantime.
The problem: the customer changed before the collection was ready
The first sign I noticed was size runs being rebuilt from one collection to the next, because the size mix is planned based on the history of past collections. But the customer who used to buy a 44 now wears a 40, and the one who wore an XL now asks for an M. In other words, there’s leftover stock in the larger sizes and shortages in the smaller ones. And it gets worse because apparel manufacturing works 12 to 18 months ahead, while the customer’s body has been changing in six.
Source: Los Angeles Times.
This is already showing up in the numbers: Impact Analytics projects that by 2027, more than 400 million garments a year will be misaligned with demand in the United States. Torrid, the largest plus-size retailer in the country, closed 151 stores in 2025, although it attributes the closures to the shift to digital, not to Ozempic (yeah, right!).
And what happens there is, obviously, already spreading to the rest of the world.
How AI helps
Planning size runs has always meant looking in the rearview mirror, repeating what sold in the previous collection with a small adjustment. That worked because the customer’s body changed slowly, but when it starts changing in a matter of months, the rearview mirror shows a person who is no longer there.
That’s where AI changes the logic, because it stops looking only at what sold and starts reading what’s happening right now. Every customer who exchanges a XL for a M is telling the brand, before any report does, that the size curve has shifted, and the same goes for the size that sells out first and for returns due to fit. AI cross-references all these signals at once, something no team can do by hand every collection.
With that, the brand can adjust its size run before cutting fabric and restock what’s moving, instead of betting everything on the initial order. McKinsey estimates that AI-driven forecasting reduces errors by 20 to 50% and cuts lost sales from stockouts by up to 65%. Still, it’s not magic, because it depends on data, and a brand that doesn’t track exchanges and returns gives the tool nothing to read.
And let’s be honest, very few brands have full command of their own data.
My take
The pandemic taught fashion to sell online, the war taught it to look at the supply chain, and Ozempic is teaching it to keep up with the customer’s body as it changes. In my view, the brands that will get through this period well are the ones that shorten the distance between who their customer is today and what is being produced for her, even if that means no longer treating history as a forecast of the future. And on this point, I think smaller brands may have more of an advantage than large groups with long supply chains.
Fashion has always known how to react to behavioral trends, but it has never had to keep up with a body that changes size faster than a collection gets made.