Stitch Fix Inc. is something of an anomaly in fashion.
It’s not exactly a retailer and not exactly an e-commerce site, and not so much a subscription box service — but it has elements of all of those.
But in a generally tough world for fashion merchants, it’s growing rapidly with an entirely different model that stresses both head and heart in equal measure.
The combined effect is powerful.
Revenues for the fiscal year ended Aug. 3 shot up 29 percent to $1.6 billion. The company — which prefers to be described as an online styling service — ended the year with 3.2 million active clients and profits of $36.9 million.
All that is enough to make the rest of fashion pay attention to the firm, which sends its users five items to try on at home and let’s them easily return what they don’t want.
There are many nuances to the company’s story and approach, but Mike Smith, president and chief operating officer, attributed much of its success to founder and chief executive officer Katrina Lake, whom he met in 2012, before becoming the firm’s fourth employee.
“She was super clear on the vision for the company,” he said. “With four full-time people at the company, we were doing jewelry and blouses, but she knew that personalization and the way we were going to use data science in our business were going to change the way retail was done. It’s very rare that you have such clarity.
“On the smart-and-humble Venn diagram [featuring overlapping circles that show different traits] there’s not a lot of crossover in San Francisco, but Katrina had it in spades,” he said.
While smart might be a given, humble could be something like the secret ingredient for Stitch Fix, which flew under the radar for a long time in fashion as it refined its approach and grew quickly.
“Humility, from my perspective, is key to disruption,” Smith said. “If you stay curious and are always trying to understand what you can do better, I think you will ultimately get to better answers, take more risks.”
By way of example, he pointed to Style Shuffle, a feature that lets users better define their style — a key element since style is cited as one of the top reasons people don’t want another fix from the company.
“It’s a game that lets you thumb up and thumb down styles you like,” Smith said, noting the company has received over 3 billion style ratings through the feature since it launched in March 2018.
“That was started by one data scientist who was intellectually curious about how to solve style for us,” he said.
Style Shuffle also illustrates another one of Stitch Fix’s strengths — it has been built from its earliest days to process and get value out of big data sets.
Data is central to how the company serves its users since algorithms go through the company’s inventory to serve up style matches for customers and then human stylists make the final selection. But data is baked in with its general operating approach to operations as well.
Smith said Stitch Fix has 125 data scientists but that only seven or eight of them work on the firm’s styling algorithms.
“They work in all parts of our business,” he said, pointing to the role of data science in demand forecasting, in how goods flow through the company’s warehouses, how inventory should be bought and more.
“We see the implementation and opportunity for data science to work in all parts of our business,” he said. “We don’t have the most data — Facebook has more data than us, Amazon has more data than us. What we have is rich and meaningful, highly actionable data.”
Smith said he’s asked periodically if the company plans to take its horde of data and package it to be sold to other players, but the answer is always no.
Stitch Fix gets its data directly from users, who pour the heart and soul of their style into the company’s system to find a fashion and fit match.
Stitch Fix has shipped over 200 million items through more than 40 million fixes. And when shoppers don’t ultimately want what they get sent and return the look, the company often gets feedback on just where the style went wrong.
“We get checkout data about how people are feeling about style, about price, about fit, about quality and that feedback really helps fuel the learnings that we get,” Smith said.
While he said Stitch Fix gets “great fit ratings” on its men’s styles from extra small to large, he said extended sizes, particularly 2XL, were a challenge. Having that granular data on just how a particular size of a style wasn’t fitting and then adjusting, sales increased in the area dramatically.
“We have built trust with the customer that, if they take time on the style profile and they give us feedback at checkout that we will give them a great fix,” he said.
That’s the other part of Stitch Fix — the part that relies on the human in the equation.
“We believe strongly that you have both” the hands-on touch of people and insights derived from data science, Smith said. “Data makes you a better merchandiser, data makes you a better planer. But data doesn’t tell you all the story.”
He pinned much of Stitch Fix’s dramatic growth over the past eight years on the “power of empathy.”
When users sign up for Stitch Fix, they spend 15 minutes filling out a questionnaire that teases out their style do’s and don’ts, their spending preferences, favorite brands and, sometimes, even more personal information, like just what look they need for a big first date.
“That is the power of a brand that is about sharing and is about empathy,” said Smith, noting the brand is not built on transactions, but relationships.
And that relationship with the shoppers — and all their data — helps Stitch Fix serve up a unique offering to each user.
“We believe strongly that the future of retail is personalization,” Smith said. “Personalization has been around for a long time, we’re just the first to play it at scale in retail.”
It’s been working.
“We’ve been cash flow positive and profitable since 2014, which is very rare, out in California, in particular,” Smith said, referring to the Silicon Valley culture that can favor growth over the bottom line.
Whatever it is that works for Stitch Fix, the company stands out as a relatively new name that is both performing well and growing as many brick-and-mortar retailers retreat.
“We’re a full-price retailer,” Smith noted. “It’s important that we deliver a full-price experience to our clients.”
Stitch Fix has to prove itself “worthy” of a full-price model, he said.
The company makes some of its own brands — and of course takes a very atypical approach to the matter.
For instance, Stitch Fix has what it describes as “hybrid designs,” made by genetic algorithms and artificial intelligence that craft looks that shoppers are looking for, but aren’t produced by brands.
It’s a small, high-tech part of the private label business, but indicative of how Stitch Fix approaches problems.
The company also offers many outside brands and Smith said both are needed to satisfy shoppers.
Sometimes brands that sell to Stitch Fix find traffic on their own web sites rising as new shoppers become acquainted with them, he said.
That’s also good for Stitch Fix.
“We want to be a platform that ultimately helps brands get discovered,” he said. “We share our data in an aggregated way with our brands.” That helps brands selling to Stitch Fix get a new read on their consumer and a look at a new approach to fashion.
“What we’re trying to do is bring the best of both worlds to bear,” Smith said.
That means being a company that manages its inventory really well, while also providing the kind of support that a really good sales associate can in-store.
“We don’t have any plans to do brick-and-mortar and we don’t have any plans to do [straightforward] e-commerce because Stitch Fix is the best of both worlds,” Smith said.