Fashion Startups: Making Clothes Fit Without Trying Them On

Kira M. Newman Aug 15, 2012 Tweet

The last obstacle for our mass conversion to online shopping, according to its disciples, is a baggy shirt.

In other words: fit. The iPad has already made fashion look beautiful online, and the search bar has made them find-able, but no widespread technology can ensure that mail-order clothes don’t hang off our frame or suction-cup our thighs. That’s why a new wave of fashion startups are talking about body scanners, algorithms, and the science of fit.

The most sci-fi of these, body scanning, is being pioneered by a few startups. Styku and Bodymetrics both use Xbox Kinect technology to create a 3D model of your body, which can “try on” clothes. Both are doing pilot testing in stores; Styku will eventually work on your Xbox or PC, and Bodymetrics works on your TV.

A bit further along in development, Berlin-based UPcload is already available in online stores like Otto and several other German sites. When you’re about to check out, you click on the UPcload button and strike 4 poses in front of your webcam, and UPcload predicts whether the clothes will be tight, loose, or baggy. They claim that measurements are accurate within 3 cm.

UPcload at work. The CD she’s holding helps determine scale.

If body scanners remind you too much of the TSA, many fashion startups use a minimally invasive questionnaire. For example, e-stores like Bonobos and Frank and Oak have added a tool called Clothes Horse. Clothes Horse asks you a few questions about your body and a well-fitting garment you own, then recommends a size and predicts how it will fit around your chest, arms, and collar. Their patent-pending algorithm also learns to make better predictions as it observes which clothes you return and which clothes you keep.

Blank Label, specializing in men’s dress shirts, starts off with a few questions about your body type, shirts you buy, and your collar size. They offer a free remake if the shirt doesn’t fit, which happens 10-15 percent of the time, says founder Fan Bi. But only 1 percent of their 20,000+ customers end up giving up and doing a full return. Zipfit.me helps men, who have much better things to do than shop, find and order jeans by rating the size of their torso, thighs, calves, and “arse” (small to large) and whether they want snug or loose pants.

Whatever their approach, most startups agree that asking people to take their own measurements is a big turnoff. Most of us don’t own measuring tape, and we certainly don’t know where our waist ends and our hip begins. “All the research we did pointed toward the idea that people don’t really want to do that much work,” says Vik Venkatraman, cofounder of Clothes Horse. “At our core, as consumers, we’re kind of lazy.”

Blank Label used to ask customers to measure their bodies, which required help from a friend, but they decided it was too complicated. “It created a lot of user friction and it wasn’t a great user experience,” says Bi. Now, the questionnaire is their main way of recommending sizes, but shoppers also have the option to measure one of their shirts or send one in for the Blank Label team to measure.

Despite these protestations, a company from Estonia called Fits.me has had some success with manual measuring. Like Styku, it features a “virtual fitting room” where your avatar can try on clothes. But rather than scan your body, it asks you to print out measuring tape and input your numbers. You’ll find the Fits.me tool on sites like the UK’s Barbour by Mail.

And Made to Fit Me tries to make the whole measurement process feel high-end and personal. If you don’t know your sizes, the Shanghai-based company will mail you a tape measure nestled inside a sheet of silky fabric samples.

Once upon a time, the only guarantee against ill-fitting clothing purchases online was a good return policy. But with a little help from engineers and programmers, fit is being demystified.