Need Some Fashion Advice? Just Ask the Algorithm

Wired | 9/12/2019 | Staff
adele2234 (Posted by) Level 3
Click For Photo:,c_limit/gear_stitchfix_1149444120.jpg

These algorithms supplement the human touch of 3,000 Stitch Fix stylists, who use the data to curate each client's box of clothes. The company has become pretty good at predicting what people want: Nearly 90 percent of its clients are repeat buyers and, anecdotally, the company seems to understand the particulars of sizing and personal style that can make shopping so difficult.

"We're getting really good at the item problem. But it's so much more complex when you start to say which items go together," says Hilary Parker, a data scientist on Stitch Fix's recommendations team, who developed the Shop Your Looks algorithm.

Stitch - Fix - Clients - Parker - Shopping

Like many of Stitch Fix's clients, Parker says she felt confident with "object shopping"—finding individual items that she liked—but struggled to put those pieces together. "I'd buy them and hang them in my closet and admire them, like a sculpture," she says. "It seemed like a good idea to explore that more algorithmically."

First, Parker needed to find data about what kinds of outfits "worked." Stitch Fix has an inventory of millions of items scattered across its six warehouses. Modeling every single possible outfit set would be impossible, and even if it were, most of those outfits wouldn't make sense. Instead, Parker enlisted the human stylists.

Stylists - Card - Item - Client - Box

These stylists already create an "outfit card" around each item that goes out in a client's box. It shows how a singular piece could be styled into a complete outfit, drawing in data about a client's personal preferences. For Parker, it set off a light bulb. "That's data that we hadn't been leveraging," she says.

The trick was figuring out how to scale the one-off outfit cards, which are hand-curated by each stylist, into a database of...
(Excerpt) Read more at: Wired
Wake Up To Breaking News!
Sign In or Register to comment.

Welcome to Long Room!

Where The World Finds Its News!