Can a computer help you shop for work clothes? Fortune puts algorithm-based Stylit to the test
Today’s working women arguably have more items on their to-do lists—and less time to accomplish those tasks—than ever before. So, it’s no surprise that a host of online services have sprung up to help cross one of those items off the list: shopping for clothes.
But while that solves one problem, it raises a new one: Which company should you choose? As a young professional still building my office wardrobe, I decided to put five of the most popular online styling services to the test, all with a single goal: finding stylish, affordable workwear. What follows is the third installment in our weeklong series. To read previous reviews, click here.
How it works
Stylit is an Israeli startup that brings together human stylists and a machine learning algorithm to help users pick out clothes. Co-CEO Assif Versano says Stylit is “more data company than personal styling service.” As with most online styling services, the process begins with an online profile. Then, users receive weekly curated “looks” on the Stylit website and can order any items that they like.
The online profile is largely image-based. I answer questions about size, fit, and budget preferences, and am asked to select looks that resemble my wardrobe. I run through what I wear on a casual day, for work, when going out, as a little black dress, in autumn, during spring, and at the beach. This is a bit of a frustrating process, as it does not allow me to fully express what I do—and don’t—like. For example, from the looks below, I chose the second from right. My rationale: I wear mostly neutrals, prefer pants to skirts, and shouldn’t wear jeans to the office. However, I don’t really like lace or bows, nor do wear heels, so it’s not a perfect fit.
Versano says that this portion of the survey is actually a method of “clustering” users by preference, assuring me that just because I select a certain look doesn’t mean I am typecast as a single style. He explains that the algorithm combines the data that I share about my preferences with information from similar users to predict what I may like.
Once my profile is complete, I receive a weekly email alerting me that my picks are ready. Each week I receive four new outfits for various occasions: some casual, others glamorous. Each look comes with clothing and shoes, as well as accessories and sometimes body care products or makeup. Versano explains that while the outfits are 90% computer-generated, the algorithm is drawing from a range of items that have been tagged and classified by actual human stylists.
I decide to order the Ivy & Blu wrap dress from the “Lovin’ the Leopard” outfit. The dress was for sale on Amazon.com and I purchase it just as I would had I found the dress on my own.
The dress I received was as I expected. It fit nicely and the quality was in line with the price point ($56).
Because you have no control over the type of looks you get each week (casual, formal, etc.), Stylit is not a good fit for shoppers like me, who are specifically interested in office wear. Versano says that the next iteration of the service, which he expects to launch by the end of 2015, will address this issue. Users will be able to select the specific types of outfits that they are interested in seeing, whether they be clothes for work, going out, or working out. “You’re going to love the new version!” he says.
There’s also no way to give feedback to improve selections other than the five-star rating system, which I find confusing. If I like one item in an outfit, does that mean I should give it one star—or does that suggest I hated the whole thing? Versano says Stylit is teeing up a fix for this, too. The new iteration of the service will let users switch out certain items in an outfit. So, using the example above, if I liked the Ivy & Blu dress, but not the Dooney & Bourke bag, I could swap out the bag for a different style.
There are a few great things about the service as it is currently exists. It’s absolutely free to use, except for any shipping and return costs, which differ based on which of the 600 or so partner retailers you purchase through. And the selection is huge: Versano says Stylit’s database has access to 20,000 brands and 3.5 million items. For those on a tight budget, the service is helpful for finding bargains and sale items. For example, the Dooney & Bourke bag above was originally $228, but Stylit found it marked down to $182. The experience is like going to a discount department store: Most of the things may not be to your taste, but you can find some amazing deals if you have the time and patience.
While an interesting idea with a lot of potential, Stylit would not be my go-to for finding work clothes today. I am, however, excited to see what the company does next.
Up next: Tog + Porter
Subscribe to The Broadsheet, Fortune’s daily newsletter on the world’s most powerful women.