Lingerie and algorithms: Big data comes to fashion

By Carmel DeAmicis , written on October 23, 2013

From The News Desk

Does your least favorite shirt fit you too tightly in the shoulders but too baggy on the belly? That's data. Do your breasts pop out of your bra and the straps dig in too tight? That's data. Are sleeves strangely loose on you? That's data.

The data trend that has overtaken everything from job recruiting to sales leads has finally made its way into an unexpected place: the world of Anna Wintour. That's right. Big data has arrived in Fashionland.

There's multiple already funded Kickstarters underway right now, all that use data to produce perfectly fitting men's t-shirts, in 24-50 unique sizes for all fellas of all shapes and sizes.

Furthermore, yesterday one of the original data driven clothing startups, True&Co, announced its own line of lingerie.  The company helps women pick bras that will fit them well by asking questions like "is your left boob bigger or your right boob?" It then factors the information into a proprietary 'algorithm' that tells women the best bra brands and sizes to wear.

Starting yesterday, in additional to its math-fashion services it will offer its own bra line too. True&Co has named the line "She Walks in Beauty (+Light)" which is quite possibly the worst fashion name ever. It manages to simultaneously evoke religious imagery and my grandmother in a nightie.

The company started in 2011 when the founder Michelle Lam got fed up with how hard it was to find bras that fit. As she tells it, she bought 500 bras and used champagne and sushi to lure friends over for test runs. They filled out surveys about how different bras fit, their likes and dislikes.

Then, Lam gathered more and more data eventually surveying 200,000 women to feed an algorithm that would recommend the best brands for a woman to wear depending on the state of her chicken wings, whether her boobs hang low, wobble to and fro, and what bras have fit her well in the past.

Bras are much harder to fit than other clothes, and True&Co is the only data-fashion company I found targeting women's clothing fit. There's Stitchfix of course, but that uses an algorithm to pick types of clothes a woman might like, not necessarily the clothes that fit her best.

The other data-fashion startups, surprisingly, focus on men's fashion. Or perhaps not surprisingly, given that it's easier to create a clothing-fit algorithm for men's bodies. Boobs and hips and jiggly bits would make the math a bit more complicated for women's clothing.

I wrote about Threadmason last week, which launched a Kickstarter to produce its perfect t-shirts for dudes. Threadmason's co-founders surveyed men about their pain points when it came to shirts, took the data to fashion designers, and developed 24 distinct sizes to fit men of varying torso, chest, shoulder, and arm sizes. Voila! A practically tailored shirt that you can order off the Internet. Eight days into its Kickstarter, Threadmason has surpassed its initial goal and has raised $23,702 out of the $20,000 target, with 363 orders pending.

But Threadmason is two steps behind a different successful Kickstarter claiming to have mastered the art of the perfect man shirt -- Stantt. Stantt's Kickstarter launched exactly one week before Threadmason's did.

Stantt used body scans of men to build 50 different sized T-Shirts. Its Kickstarter has raised $69,364 of $15,000 amount and still has four weeks left. Silly as it sounds, there's clearly a market among men for the perfectly tailored tee.

These companies have launched in the wake of other startups with a different approach to the tailored outfit. Startup Trumaker does actual tailoring and sends 'outfitters' to guys' homes to take their measurements. With the measurements, Trumaker produces perfectly fitting button down shirts.

In comparison, e-commerce YC grad Vastrm helps men buy tailored polo shirts. Customers receive a few sample shirts in the mail, and then they can use them as a basis to make measurement adjustments online -- like choosing a longer sleeve length or different pockets. Once they've picked their ideal measurement a fresh, fitting batch gets shipped back to them.

The businesses that focus on algorithms and data to help people find a perfect fit will probably perform better over time than the ones that rely on people. Both Trumaker and Vastrm add a layer of complication with the human element. With Trumaker, Outfitters have to go to someone's home to take a person's measurements. With Vastrm, the purchaser has to do the leg work to figure out his own perfect fit.

It would be easier to scale a company with set formulas geared to find sizes that fit most people, as True&Co, Threadmason, and Stantt are doing.

The data approach to fashion is compelling because if the trend expands it will allow people to get tailored versions of all their basics, without having to shell out the cash.  Some day in the future, you may not have to compromise on clothing just because you've got a J-Lo booty with weirdly skinny calves and can't find jeans to fit you.

As I mentioned in the Threadmason story, it's a decidedly first world problem but hey…there's money to be made in people's vanity.

[Image via: Thinkstock]