At first glance you would not peg Stitch Fix as a company that depends on data analytics to enhance its business and create value. But upon closer inspection it becomes clear that Stitch Fix is about as advanced as they come in regards to using deep data analytics to deliver a differentiated and sticky consumer product.
The Core Offering
Stitch Fix’s core product is a subscription clothing offering. When users sign up they are asked to take a multi part “style quiz” to help Stitch Fix get a better sense for their preferences. Once that’s done users can indicate how frequently they’d like to receive a box of clothes that have been specifically selected for that user based on their indicated preferences. Upon receiving the box users try on items and have the option to keep the ones they like, at which point they’ll be charged. Users can return the items they don’t wish to purchase to Stitch Fix, free of charge.
While the interaction seems easy enough from the user perspective, Stitch Fix is using countless algorithms and analyses in the background to gradually improve the recommendation engine and user experience.
Stitch Fix establishes an initial profile of the user based off of their style quiz, and hones it using feedback from each style box over time. When users return items to Stitch Fix they are asked to rank the article of clothing on a number of style and fit dimensions. Stitch Fix combines that information with price, environment, and an additional 30 to 100 dimensions to paint a clear picture on user preferences.
Back at headquarters the Stitch Fix Data Science team, composed of over 80 scientists with PHDs in quantitative fields such as math, neuroscience, statistics and astrophysics, uses those data points to develop the most complete map of a user’s stylistic preferences on the web. As an indication of how important Data Science is to the business, CEO Katrina Lake has the function report directly into her.
The Stitch Fix FlyWheel
Like many data intensive businesses, Stitch Fix gets better at meeting the needs of all their users as the number of users on the platform grows. The more boxes a user receives, and the more articles of clothing they buy, the better the algorithm gets at predicting preferences. This in turn improves the shopping experience resulting in more user purchases and an ever continuing flywheel. The benefits of this flywheel are not contained to an individual user as Stitch Fix is able to get a better read on which of the up to 100 clothing dimensions they measure is most important to the average user, enabling them to serve more relevant offers to new users quicker.
In addition to improving the user experience the flywheel also improves Stitch Fix’s value capture prospects. Stitch Fix has begun using the extensive data set they have on their 2m plus customers to inform investments in proprietary first party brands. By building out a suite of first party brands Stitch Fix is effectively vertically integrating, allowing them to expand margins and capture a greater share of the purchases they drive.
Beyond first party brands Stitch Fix has also launched “Direct Buy” where users can browse a curated selection of items for direct purchase in addition to or in lieu of their “style box”. The direct buy business allows Stitch Fix to expand their customer wallet share, while also establishing themselves as the go to spot for all customer clothing purchases. Over time, as Stitch Fix directs an increasing portion of customers’ clothing spend, it should be able to extract rents from 3rd party brands in the form of ad revenues.
Stitch Fix has built a unique and loved company by relying on complex and advanced data analytics. At a time where it has become nearly impossible to compete with Amazon in the realm of e-commerce, Stitch Fix has a plausible path towards out maneuvering them in the realm of high dollar retail.