The stylist didn't rise. It dissolved
For about a decade the personal stylist looked like the future of buying clothes. You would answer a quiz, a human would read it, and a box would arrive. The pitch was that taste could be delivered as a subscription.
That business has largely gone. Nordstrom closed Trunk Club in May 2022. Zalando stopped taking new Zalon customers in July 2022 and wound the service down across Belgium and the Netherlands that autumn. Amazon shut its physical Style stores in November 2023 and ended Prime Try Before You Buy on 31 January 2025.
It would be easy, and wrong, to write that as a failure story. What actually happened is stranger. Styling did not go away. It stopped being a service you subscribe to and became a feature inside somebody else's conversation.
First, the part that cuts against me
Stitch Fix is the obvious counter-example and it deserves to go near the top rather than buried at the bottom.
It is not collapsing. In the third quarter of its 2026 financial year, ended 2 May 2026, it reported net revenue of $340.3m and what chief executive Matt Baer described as a fifth consecutive quarter of year on year revenue growth on an adjusted basis. The quarter before that, revenue grew 9.4% year on year. Revenue per active client has been climbing. In Q3 the company reported sequential growth in active clients and called it a milestone.
Client numbers are still below where they were a year earlier. But a business with fewer customers each spending more, growing revenue five quarters running, is consolidating around a higher-value base. That is a pivot, not a decline, and anyone arguing that the styling service is finished has to explain those numbers first.
Two other things are worth saying plainly. Stitch Fix left the UK: it confirmed its exit in August 2023 and ceased operations here on 31 October 2023, so a British reader cannot use it whatever the numbers say. And in January 2024 the company restructured its styling workforce, moving full-time stylist roles to part-time, affecting several hundred people according to trade reporting at the time.
What replaced the box
On 20 May 2026, ASOS launched ASOS Stylist. Not on ASOS.com. Inside ChatGPT.
Built with the video commerce firm Bambuser, it turns the ASOS catalogue and video library into structured data a language model can retrieve in real time. You can ask for pastel floral A-line dresses for spring and get a curated edit, in video, from across the brand portfolio, then continue to ASOS.com to buy. Melissa Lim, ASOS's Head of Product, framed it as bringing discovery, styling advice and shoppable products into one conversational experience, and as a step toward reducing friction in how customers shop.
It is genuinely good, and it is free, and it sits where people already are. That is the whole point. The subscription box asked you to come to it. This meets you inside a conversation you were having anyway.
The thing that isn't in the announcement
Read ASOS's launch material and you will find category, occasion, trend, styling guidance, complete looks, shoppable video, hundreds of brands.
You will not find size. Or fit. Or measurement.
That absence is worth stating precisely, because it is easy to overclaim. The announcement not mentioning fit is not the same as the tool being unable to handle it. I have not tested it, and until someone does, the honest sentence is the narrow one: a major retailer launched a styling experience on a new platform and described it entirely in terms of taste.
Which is consistent with everything else on this site. Try-on answered the style question at enormous scale while Google's own documentation said the generated image doesn't indicate fit. Body scanning retreated from consumer fashion onboarding into fitness and health. The taste layer keeps getting better, cheaper and more ambient. The body layer has barely moved.
There is one exception worth crediting. In October 2024 Zalando joined body measurement technology to its virtual fitting room so customers could build an avatar from their own measurements rather than a statistical model, initially on Levi's upper garments across fourteen European markets. It was a pilot, and Zalando was careful to describe its results as limited at that stage. Notably it is the retailer, not the assistant, doing that work.
Why this matters more than it looks
Here is the shift that actually counts, and it is not the one about robots replacing stylists.
When styling lived on the retailer's own site, whatever it knew about you stayed there. Now the styling conversation happens on a third-party platform, and the retailer becomes an app inside it. Your preferences accumulate somewhere new, and they are still preferences rather than measurements.
This site argued in July that none of the fit-prediction incumbents holds a profile that belongs to you. That was true when the profiles sat with True Fit, Fit Analytics and Bold Metrics. It is more true now that the conversation has moved one layer further out, because the number of places holding a fragment of you just went up again, and none of them holds your chest, your shoulder or your inseam.
The strongest objections
Three, and none of them is weak.
People may simply prefer this. Logg, Minson and Moore found in a 2019 paper in Organizational Behavior and Human Decision Processes that people show a measurable preference for algorithmic over human judgment. Their experiments concern advice-taking and forecasting rather than shopping, so the result cannot be transplanted directly into a fitting room. But the direction of travel is not obviously against what people want.
Recommendation engines cut both ways on choice. Work by Fleder and Hosanagar, and later by Lee and Hosanagar, found that a fall in what a market as a whole buys can coexist with a rise in what an individual buys. The randomised field experiment in that later work ran across more than 82,000 products at a large North American retailer, and it was general retail, not fashion. It says nothing about whether a styling recommendation is any good.
And Stitch Fix's numbers, again. Assistants have not obviously beaten the human-curated box. The box is growing.
Where that leaves you
Styling is now abundant. Ask any assistant what to wear and you will get a decent answer, instantly, free.
Ask it whether the thing will fit across your shoulders and it has nothing to work with, because nobody has ever measured you, and no size chart in the conversation describes your body.
That is the layer we are building.
Sources: ASOS plc release and Bambuser announcement, 20 May 2026. Stitch Fix investor releases, Q2 FY26 (11 Mar 2026) and Q3 FY26 (10 Jun 2026). Stitch Fix UK closure, Aug and Oct 2023. Retail Dive on the 2024 styling restructure. Zalando corporate post, 17 Oct 2024. Nordstrom, Zalando and Amazon closure dates as announced. Logg, Minson and Moore (2019), Organizational Behavior and Human Decision Processes 151, pp. 90-103. Lee and Hosanagar, Information Systems Research, DOI 10.1287/isre.2018.0800. Fleder and Hosanagar (2009), Management Science 55(5).
