Fifteen floors above San Francisco’s historic shopping district, Stitch Fix CEO Katrina Lake is scrolling through her client queue, selecting clothes for a 47-year-old mother in Southern California. Lake is in the midst of a busy month. In addition to running the 5,800-person company she started out of her apartment seven years ago, work is taking her to Berlin and New York City. But she is still taking time to style “fixes,” the personalized boxes of clothes that her company has shipped to more than 2.5 million people in the past year.
“What makes this company special is undoubtedly technology,” Lake says, glancing over the “match scores” that algorithms have assigned to rows of shirts and skirts on her screen. Clients can choose to buy the items in their box or send them back, and these data-driven predictions help stylists pick clothing that clients are more likely to keep. “That being said,” Lake goes on between sips of water with lemon, “I just don’t believe in the tech tag. At the end of the day, I think Uber is a transportation company. I think Airbnb is a hospitality company. I think Facebook is a marketing company. And we are a retail company.”
Lake, 35, stands out among her fellow tech executives. She’s not only a female CEO at a time when about 4.4% of venture-capital deals go to female-founded companies, according to database PitchBook. She’s also the youngest woman to ever lead a business to an initial public offering. Although the Bay Area native used to resist people’s tendency to focus on her gender, that has changed as Stitch Fix has grown. “The biggest thing that we could do is to be an example of a company that is founded by a woman, that is in a women’s industry and goes public,” the CEO says. “Because if nothing else, it wakes up the capitalists.”
Stitch Fix is on track to bring in $1.2 billion in revenue this year, surpassing estimates for more established startups like Yelp and Zynga. While the company is a rare example of a tech firm that deals in apparel, its business model is built on the same currency as many others: vast troves of information about its users. Only by getting personal, Lake says, can the company transform the way America shops, sending people clothing they’ll like rather than leaving them to search for it. To do that, the company now employs more than 3,000 stylists and more than 75 data scientists (led by a former VP of data science and engineering from Netflix). “What we do right now is crazy,” Lake says of digging through racks of clothing at the mall or trying to find a perfect pair of pants by visiting dozens of stores online. “I just don’t believe that could possibly be the future.”
Lake came up with the idea for Stitch Fix while attending Harvard Business School. The premise is simple: sign people up to receive shipments of apparel that are chosen for them through a combination of personal styling and data analysis. The complex part is trying to increase what the company calls its “keep rate.” Algorithms help curators choose five items for each shipment. Ideally, customers will love all of them. Often they don’t. The company won’t say how many items customers keep on average, but a spokesperson says the figure has been going up each year.
Lake explained how the process works during a lunchtime tour of one of the company’s five warehouses, in South San Francisco. Strolling by racks of clothes, her petite frame wrapped in a sky blue coat, she talked over the rattle of conveyor belts and the oozy rap of Post Malone, a soundtrack for workers preparing boxes to send to customers across the Northwest. When a stylist opens up the company’s proprietary platform, the first options they see have been automatically culled from the vast array of inventory in warehouses like this one. Those algorithmic recommendations are based on factors ranging from the client’s body shape and age to the apparel’s sleeve silhouette and price. But the final selection is also a product of the stylist’s discretion, shaped by feedback that the company solicits from users every time a box is shipped.
As stylists move through the queue, they also see personal information about each client: their name, their weight, their location (and the weather there), whether they are a mom or dad. The more the company knows about subscribers–especially what apparel they like and dislike–the better those match scores can get. Another result is that Stitch Fix has an enormous amount of data about consumer habits and opinions on the fit, price and style of clothes. That information is used not just to optimize fixes but also the clothing itself. The company works with hundreds of brands, and some are changing their specs because of the aggregated feedback that Stitch Fix shares. “It’s infinitely better than data [vendors] are getting through any other channel,” says Lake, who emphasizes that the company would never divulge information about individual customers.
Although Stitch Fix does this for free, it’s natural to wonder whether vendors would be willing to pay for that kind of insight. “They may,” Lake says, “but our model is not one that sells data.” Monetizing user info has put other firms in hot water. “There are a lot of companies that haven’t been really transparent around what data they’re collecting, what they’re doing with it,” Lake says. “And I have a little bit of fear that there’s been a degradation of trust.” She says Stitch Fix is clear about its bargain with users. “If you share with us that the dress is too big or too small,” she says, “next time we’re going to get you a dress that is going to fit you better.”
Lake is also having a busy month because she is the mother of a toddler, but she has mixed feelings about people asking about that part of her life. Just 17% of venture-backed companies have at least one female founder, according to Crunchbase, and Lake is a model of what is possible for women who are trying to make it in the startup world. “Women genuinely want advice. It’s hard, and I think people want to hear validation that it’s hard,” she says. “Everybody should also be asking men.”
Like other industries, tech has come under fire for corporate cultures that can be hostile to women. Lake won’t speak about it, but news reports suggested that she once asked for a VC to be removed from her company’s board after his behavior made her uncomfortable. She says that she knows it’s difficult for women to speak up when they encounter bad behavior and feel outnumbered. “You are reliant on a venture population which is [overwhelmingly] male in order to keep your company alive and your employees paid,” she says. “It’s a really impossible choice to make.”
Female CEOs sometimes complain about being tokenized, like being included on a panel to provide a patina of gender diversity. Lake says this kind of thing happens to her “all the time,” though at this point it might take the form of her being asked to invest in a company. If change is happening for reasons of appearance, she says, that’s still better than nothing. But she also believes that decisionmakers are realizing that being more inclusive of women, on every level, is simply good business. Consider the male investors who chose not to fund her female-focused startup. “Even if they are not going to get excited about fashion, if they feel like they missed out on that one,” she says, “that creates a natural and capitalist reason for people to pay more attention.”
This appears in the May 14, 2018 issue of TIME.