If you’ve been following PostFunnel, you know we like to talk about personalization and how data plays an integral role in eliminating spray and pray. Today, we’ll examine how Stitch Fix uses its troves of customer data to provide individualized service to each customer.
More from PostFunnel on Personalization:
Winning At Real-Time Personalization
5 Festive Tips to Celebrate Your Customers’ Birthdays
“Data science enables a personalized experience without being creepy”
What is Stitch Fix?
The fashion subscription box that provides personal stylists for customers has over 2.7 million active clients, an audience boasting an 86% return rate, and is worth over $1.6 billion. For Stitch Fix customers, the process is straightforward:
- The customer fills out a survey regarding their build, style, price range, and other personal information
- They receive a delivery of hand-picked clothing items to try on
- They pay for what they like and send back the rest
- Rinse and repeat at the customer’s leisure, with each new shipment becoming increasingly personalized as time goes on
On the company’s end, the process is more complex, as it involves both people and machines crunching numbers and digging into data to determine the best items for each “Fix shipment.”
Personalization is essentially Stitch Fix’s modus operandi. In the words of CEO and founder Katrina Lake, “Data science isn’t woven into our culture; it is our culture.”
Let’s take a look at how Lake and company have created a successful fashion empire centered around customer data.
How Stitch Fix Uses Data
Stitch Fix simply couldn’t exist and operate as it does without technology, so they use it openly and to their advantage. The subscription brand is always collecting information about its customers whenever they engage with the company. The initial survey allows customers to provide preliminary information about themselves, and as a customer further engages and makes purchases or returns, the company collects data on their tastes and style. Any customer feedback is also collected and stored in the company’s CRM.
Stitch Fix’s machines don’t make generalized product suggestions based on surface-level data. Rather, thanks to the dozens of algorithms created by the company’s data science team (more on them later), the machines can dig deeper into the customer’s true desires.
The machines use their information on the individual (via first-hand engagements, feedback, and third-party engagements such as Pinterest likes) to provide a “match score” on a given piece of clothing.
These match scores consider all available information, but in addition to this data, they’re also incredibly granular with their product descriptions. This allows the machines to score items based on a variety of factors, such as style, colors, and patterns, and how they run in terms of size and fit. These pieces of information aren’t considered as individual facts, but as part of the entire picture. The result is an ultra-specific idea of each individual consumer and what kind of items they’re searching for.
The company’s powerful machines also play an integral role in enabling the team to develop new items. As Chief Algorithms Officer Eric Colson explains:
“We’re… combining elements from several parents to create something new. Say we take a silhouette from one garment and the sleeves from another, maybe a collar from a third and a pattern from a fourth. And then we recombine them to create something that has never existed.”
This isn’t done at random. The decision to move forward with the design of a new fashion item is based entirely on data. Here’s Lake’s explanation:
“Many female clients in their mid-40s were asking for capped-sleeve blouses, but that style was missing from our current inventory set. Fast-forward a year, and we have 29 apparel items for women and plus sizes that were designed by computer and meet some specific, previously unfilled needs our clients have.”
Stitch Fix has developed an algorithm that predicts the likelihood of an individual appreciating an item that doesn’t even exist yet. Essentially, this means Stitch Fix can now validate the creation of new items without first needing to test them.
None of this would be possible if the team didn’t place data at the center of its operations.
Other Uses of Automation and Data at Stitch Fix
The team uses data and machine learning in a variety of other ways, such as determining which stylists to match with which users.
A similar process is used to determine the optimal shipping location for orders:
Finally, the team uses complex algorithms and formulas to keep its inventory stocked and anticipate future demand:
Whether looking to provide more highly personalized recommendations, better streamline fulfillment processes, or keep costs to a minimum, the Stitch Fix team first turns to data.
The Human Side of the Equation
Of course, while technology is integral to the way Stitch Fix leverages data, it’s still the humans who truly enable it to operate so seamlessly. Stitch Fix’s team of data scientists are constantly working on creating new algorithms and improving current ones.
Says Lake, “We’ve developed dozens of algorithms that no one ever asked for, because we allow our data science team to create new solutions and determine whether they have potential. No one explicitly asked the team to develop algorithms to do rebuy recommendations, for example.”
This autonomy transfers to Stitch Fix’s stylists, as well.
As mentioned earlier, though algorithms are responsible for providing initial product recommendations (based on match score), each Fix shipment truly is hand-picked by an actual stylist within the company. During the selection process, stylists not only have access to the data, but also rely on their personal knowledge of the customer. This allows the stylist to fine-tune recommendations as necessary, providing the customer with the perfect box.
Lead stylist Layla Katz explains, “It helps to not have to worry about the broad strokes of what a client does not want. Then we can make creative decisions about what will fit her body and her lifestyle.”
Stylists are often able to collect information that computers cannot. Typically, this involves the more qualitative comments or suggestions customers may make “on the fly.”
“When a client decides which items to keep or send back, she can go through her profile and let us know item by item if she liked the fit, the price, the quality. That goes into the algorithm and helps it suggest more for the next time,” Katz added.
The stylists help “teach” the machines and improve the algorithms, and in turn, the machines provide laser-focused and accurate product suggestions.
As much as this may sound like something out of a science fiction novel, the humans and computers over at Stitch Fix have begun operating in a sort of symbiotic harmony.
As Lake explains, “A good person plus a good algorithm is far superior to the best person or the best algorithm alone… We need them to work together.”
As both humans and machines learn more and more about their customers — and figure out more advanced ways to use this data — it will become easier for Stitch Fix to cater to the needs of its individual customers. Will you follow their lead?
The post How Stitch Fix Uses Data to Increase Sales and Engage Customers appeared first on Post Funnel.
Latest posts by tomas (see all)
- NEWS // New Representation: Tea & Water Pictures - 22.9.2019
- Value Stream Management - 22.9.2019
- IIBA Podcast Episode 9: Data analytics: Transforming data into insight and action - 22.9.2019