Stitch Fix: Applying Data Science to the Art of Personal Style

By Alex Batty, MHI Marketing Communications Coordinator |@mhi_alex

At this point, I’m guessing most Americans have heard of personal styling service Stitch Fix.

For those of you who don’t know, Stitch Fix is “a personal styling service that sends individually picked clothing and accessories items for a one-time styling fee. Customers fill out a survey online about their style preferences. A stylist at the company picks five items to send to the customer… Once the shipment is received, the customer has three days to either keep the items or return them. If they keep an item, the initial styling fee applies towards the cost of the item.”

Founded in 2011 by former a J. Crew buyer and her sister, the company wasn’t profitable until 2014, but has really started to boom in the past four years. But while I was researching, the line that was interesting was this one: “[Stitch Fix] employs more than 3,000 stylists and more than 75 data scientists (led by a former VP of data science and engineering from Netflix).”

Data scientists?

Why does a clothing company need data scientists?

And it’s actually really cool. It ties into NextGen supply chains and the how Stitch Fix has managed to deliver personalization at a large scale.

This article from eTail explains more. From a customer perspective, the service is relatively simple: They send you stuff, you send back stuff you don’t want. But the secret to getting the personalized recommendations is backed by a whole lot of data science and machine learning algorithms.

However, as eTail points out, “no matter how many magic formulas and algorithms are at work, they will only ever be as good as the data that drives them – and data is where Stitch Fix really excels.”

One dataset is provided by customers in the onboarding process. Customers provide fundamentals like height and weight, but also give style preferences, personal traits, and lifestyle.

On the inside of the process, Stitch Fix tags each clothing item with “”match scores” derived from client preferences, and then ranked. Even more data is collected from customer feedback, which is solicited every time a customer receives a curated box,” and gains insights into customer preferences outside of self-reported data.

However, while CEO Katrina Lake says that “Stitch Fix wouldn’t exist without data science,” they do include the human element in combination with the data science. The human stylists can override the product assortment the algorithm recommends in response to specific customer needs. Stitch Fix CTO Cathy Polinsky says, “By combining the art and science of styling we’re able to create a far better client experience.”

Stitch Fix is yet another example of the NextGen, digital, always-on supply chain disrupting traditional retail. With personalized clothing recommendations showing up right at your door, customers no longer feel the need to shop in-store any longer. Amazon has introduced Prime Wardrobe as a competing (and slightly different) service to mange clothing returns in their supply chain. Brick and mortar clothing retailer Nordstrom has introduced Trunk Club, a similar service to Stitch Fix but exclusive to Nordstrom brands.

There has been a lot of fear lately surrounding the idea that machines and algorithms and robots are going to take human jobs. But really, adding machines and machine learning is about enhancing supply chain processes. When robots and algorithms take over the menial data-processing tasks they’re so good at, it frees up human capital to spend time on the human necessary tasks – providing that special touch. As Lake puts it, “It’s simple: A good person plus a good algorithm is far superior to the best person or the best algorithm alone.”