Buying Shoes? Digital Stylists Use A.I. to Suggest Clothing to Match

Xconomy National — 

The black skirt had multiple personas. The fact that it was made of leather gave it an edgy vibe, but its A-line fit-and-flair cut was more flirty and feminine. “I fell in love with it,” says Michelle Bacharach, co-founder of FindMine. “But as soon as I brought it home, I wondered, ‘how do I wear this?’ ”

That question, along with a realization that she found herself asking it repeatedly over a variety of her purchases—from clothing items to housewares—led her to found FindMine, an e-commerce startup that uses machine learning and artificial intelligence to help shoppers find clothing items that can complete an outfit.

“I’m constantly having to do all this work after the purchase, or I end up not buying things I like because I don’t know how to wear it beyond one outfit,” Bacharach says.

The idea behind FindMine is to give shoppers recommendations—this blouse or pair of shoes would pair well with the black leather skirt—at the point of sale, when a customer could purchase multiple items, Bacharach says. “Once it’s in your closet it’s too late,” she adds.

For brands and retailers, getting these outfit suggestions to shoppers in the store can result in more spending than if a shopper were to buy only one item. For example, the e-commerce website for men’s fashion house John Varvatos, a FindMine customer, displays a dark gray suede jacket cut in a “military racer” style that sells for $798. Beneath it is a “complete this look” section which advises shoppers to pair the jacket with an azure button-down with subtle stripes, tuxedo-style pants, and a pair of chocolate-brown “Brooklyn lug lace” boots—all made by Varvatos.

Other customers include Adidas and American Eagle, which uses Facebook messenger to deliver FindMine’s clothing suggestions to customers, Bacharach says. She estimates that FindMine puts together 11 million outfits each day.

Stylist services make up a growing niche in e-commerce. Some startups such as Fitz focus on human stylists who come into a customer’s home to help analyze what you wear and why, and then reorganize the clothing and accessories that make the cut.

But others have concentrated on using technology to offer these services to a greater number of people more quickly. Stylistics has ClosetSpace , an app that is digitizing that process, giving customers outfit suggestions from their virtual closets and also from retailers’ items that they might want to buy.

Even e-commerce giant Amazon (NASDAQ: AMZN) is getting into the game with services like Spark, which is designed to help shoppers find compatible new products—from clothes to home decor to beauty and grooming.

“We all use these tools to help manage our lives: traveling, mobile banking,” says Brooklyn Decker, co-founder of stylist app Finery. “When you look at how women will spend more money on their clothing than on their education, there should be something to manage their clothing.”

Decker, an actress, started Finery last year with Whitney Casey, a former television anchor, to create a “wardrobe operating system.”

Once a person sets up an account with Finery, the startup essentially scrapes the online store accounts connected with that e-mail address for the last decade or so, and mines the purchases from those vendors. That information is then used to create a digital closet on the Finery app, which uses a Pinterest-like display. Casey says Finery has 700 stores and brands on its app—and it can add small boutiques if a few customers mention that they shop there.

“We want your data to be working for you,” Casey adds, “to help you make smarter decisions about what to wear and helping you to shop more strategically.”

Finery will also note and keep track of things like return policies, and send automated alerts when the window is closing. So far, the company says, it has about 100,000 users.

I decided to give Finery a spin, and found that—at least for a customer like me—it won’t be easy to create my virtual closet by mining my e-mail messages. I tend to strive for a clutter-free inbox, so I delete my order messages from Amazon or Zappo’s as soon as the deliveries are made. So my Finery Wardrobe only consists of three items.

Oddly, one of them is a purse I don’t own. Another surprise is that while Finery found a pair of black heels I ordered from Zappo’s last year, it did not locate a second pair of shoes in that same order.

Catherine Clark, Finery’s manager of partnerships and business development, says if you don’t have certain order e-mails any more, you can log onto those retail accounts from the Finery website. For me, that means tracking down my passwords for Nordstrom or Neiman Marcus, something I will have to carve time out for later. That kind of time commitment that might turn off some users.

Last month, Finery announced it had raised $5 million in a seed round led by NEA and investors such as BBG Ventures, Farfetch, and RetailMeNot founder and CEO Cotter Cunningham, among others. The funding will be used to add developers and data scientists. Casey says they are working on a beta to use artificial intelligence tools to make styling recommendations.

(FindMine has raised $1.7 million from XSeed Capital, Kiwi Ventures, RevTech, and XRC Labs, as well as a handful of angel investors, Bacharach says.)

The idea for FindMine first occurred to Bacharach eight years ago, and she explored it at a business plan competition while attending New York University’s Stern School of Business. But it wasn’t until recently when the necessary tech innovations—combined with retailers’ search for more personalized ways to target customers—were in place to provide the kind of service that Bacharach envisioned with that leather skirt. “After a lot of false starts, in August 2015, we found our first retail customer,” she says. “Then the growth has been dramatic.”