The company that turned "I hate shopping" into a data problem - and shipped the answer to your door.
STITCH FIX // San Francisco, California.
An online personal styling service where an algorithm narrows the rack and a human makes the final pick.
Most retailers ask you to shop. Stitch Fix asks you five questions and then mails a box. Inside are a handful of items - a shirt, a pair of jeans, maybe a jacket you would not have found on your own - selected for you specifically. You keep what you like, send back the rest in a prepaid envelope, and tell the company what worked. That loop, repeated across millions of clients, is the whole idea: a fashion company built less like a store and more like a recommendation engine with a loading dock.
Founded in 2011 and public on the Nasdaq since 2017 under the ticker SFIX, Stitch Fix has spent more than a decade arguing that the hardest part of buying clothes is not payment or delivery - it is choosing. Its answer pairs machine learning with professional stylists, and it closed fiscal 2025 with roughly $1.27 billion in net revenue.
At its core, Stitch Fix is a personal styling service delivered by mail. A new client fills out a Style Profile - sizes, fit preferences, budget, the occasions they dress for, styles they like and avoid. From there, the company's recommendation system sifts an inventory drawn from more than a thousand national brands and Stitch Fix's own private labels, narrowing thousands of options to a short list. A human stylist reviews that list and selects the final items that ship in a "Fix."
The client tries everything on at home, with no store trip and no crowded fitting room. Whatever does not fit or does not feel right goes back. A styling fee applies to each Fix and is credited toward anything the client keeps, so the curation is paid for whether or not a purchase happens - but the fee disappears the moment you buy.
The service spans women's, men's and kids' apparel, along with shoes and accessories, and covers ranges that mainstream retail often treats as afterthoughts: petite, plus, maternity and tall. For clients who would rather skip the box entirely, Freestyle offers an on-demand storefront of personalized picks that can be bought directly.
Answer a style quiz on sizing, fit, budget and taste.
Algorithms shortlist; a stylist picks the final items.
A Fix arrives home. Keep what fits, return the rest.
Your feedback trains the next, better recommendation.
Stitch Fix serves individual consumers - women, men and kids - primarily in the United States and the United Kingdom. The through-line among them is a preference to be dressed rather than to browse: busy professionals, new parents, people returning to the office, anyone whose relationship with a shopping cart is more chore than pastime. Net revenue per active client sat around $559 in the first quarter of fiscal 2026, a figure the company has worked to lift by deepening each relationship rather than only chasing new sign-ups.
Because the model captures explicit feedback on every item, Stitch Fix accumulates an unusually granular picture of fit and taste. That is useful for a shopper who wants to be understood, and it is the raw material for the company's buying, allocation and design decisions.
E-commerce offers near-infinite inventory and near-infinite scrolling. Stitch Fix collapses it to a handful of items chosen for you.
Sizing varies wildly across brands. A profile plus feedback history narrows the guesswork before anything ships.
No store trip, no fitting room. Try on at home on your schedule and return the rest for free.
A stylist introduces brands and pieces a client would not have searched for - discovery without the effort.
Bars scaled for illustration. Quarterly figures are not additive to the annual bar. Sources: Stitch Fix investor releases, 2025-2026.
The obvious comparison is other styling boxes and try-before-you-buy services, and Stitch Fix has plenty of company there - Trunk Club (bought by Nordstrom), Amazon's personal-shopping features, Nuuly, Wantable and a field of niche subscriptions. What separates Stitch Fix is the depth of the data layer underneath the human service.
Most retailers treat returns as a cost to minimize. Stitch Fix treats every keep and every return as a labeled training example, feeding a system that gets more accurate with each cycle. The company built one of retail's largest applied data-science organizations and even runs a public engineering blog, MultiThreaded, that documents the work.
The other distinction is philosophical. Stitch Fix does not frame this as machines replacing stylists. The algorithm narrows the rack; the stylist makes the call. The interesting part of the company lives in that seam - taste that resists spreadsheets, sitting next to math that makes taste scale.
The algorithm narrows the options. The human makes the final pick. Neither works as well alone.
A curated shipment of ~5 items chosen by stylist and algorithm, tried on at home. Keep what you love, return the rest.
On-demand storefront of personalized recommendations you can buy directly - no scheduled box required.
The data system that captures size, fit, budget and taste, matching clients across 1,000+ brands.
AI-assisted tools to visualize and build outfits; credited with lifting Freestyle spend.
Exclusive labels designed from demand and fit data across women's, men's, kids, plus, petite & maternity.
Order Fixes on your own cadence or shop Freestyle. There's no required recurring plan.
Stitch Fix is a direct-to-consumer retailer, but with an unusual engine. It earns revenue primarily from the merchandise clients keep, plus styling fees that are credited toward purchases. The company carries its own inventory and uses machine learning not just to recommend but to buy and allocate product - deciding what to stock, in what sizes, for which clients - which helps hold down the returns and markdowns that punish traditional apparel.
Freestyle broadened the model from a scheduled box into an always-on storefront, giving clients a second way to spend and the company a second way to monetize the same profile data. The strategic thread running through recent years is depth: more revenue per client, more categories such as activewear, footwear and accessories, and AI tools that make each interaction more useful.
Stitch Fix's expertise sits at an unusual intersection: apparel merchandising, human styling and large-scale applied machine learning. The company hires data scientists and professional stylists in roughly equal spirit, and its technology stack reflects a serious data operation - cloud infrastructure, streaming and analytics pipelines, and a long-running culture of publishing its methods.
In the broader market, Stitch Fix occupies a specific niche within US apparel: personalization at scale, positioned between generic e-commerce on one side and traditional department-store service on the other. It is not the largest apparel seller, and the public markets have been volatile about the name since its 2017 debut. But few competitors combine a decade of proprietary fit-and-taste data with a human styling network of comparable size.
The current chapter, under CEO Matt Baer, has been a turnaround story: trimming, focusing on the client relationship, and leaning into AI. Fiscal 2025 delivered a return to adjusted year-over-year growth, and fiscal 2026 opened with revenue up 7.3%.
Katrina Lake launches the styling service from her Cambridge apartment while at Harvard Business School.
Benchmark Capital leads a $12M Series B as the data-plus-stylist approach gains traction.
The company deepens its analytics organization and expands merchandising as the client base grows.
Stitch Fix goes public as SFIX, raising about $120M near $1 billion in annual sales.
An on-demand storefront lets clients shop personalized picks without a scheduled box.
Fiscal 2025 closes near $1.27B with adjusted YoY growth and new AI styling tools.
Fiscal 2026 opens with 7.3% revenue growth as Vision AI tools lift Freestyle engagement.
The company began under a different name, run out of Katrina Lake's Cambridge apartment.
Early clients paid the styling fee by physical check, and the first Style Profiles ran on SurveyMonkey.
The two disciplines sit side by side - the seam where the company's edge actually lives.
MultiThreaded, the company's engineering and data-science blog, documents how the machine learns.
You complete an online Style Profile, then a stylist and algorithm curate a shipment (a Fix) of clothing and accessories. You try them on at home, keep what you like, and return the rest. A styling fee applies and is credited toward anything you keep.
No. You can schedule Fixes on your own cadence or shop on demand through Freestyle. There is no required recurring subscription.
Katrina Lake founded it in 2011 (originally "Rack Habit"), with Erin Morrison Flynn involved on the early merchandising side. The company went public in 2017.
Freestyle is Stitch Fix's on-demand storefront that shows personalized recommendations you can buy directly, without waiting for a curated box.
It uses machine learning to match clients to items across 1,000+ brands, guide buying and inventory decisions, and support stylists - with each keep or return improving future recommendations. Newer Vision AI tools help clients visualize outfits.