Engineering The Perfect Big-Data Bra

True & Co. used a Netflix model of big-data mining to create a new bra line. Does this finally mean perfect undergarments for all?


There’s nothing sexy about big data. Then again, for anyone who has ever spent time in a lingerie dressing room, there’s nothing sexy about bra shopping. That very experience of letting a stranger feel you up before spending hours half-naked trying on cup sizes that don’t generally translate to your body is what inspired Michelle Lam, cofounder of True&Co., to reinvent the bra.“I looked at this bra and said ‘I wonder when the last time someone took a fresh look at the bra, and reinvented it,’” she says.


Brassieres, however, are complicated garments to manufacture. “It’s not like T-shirts,” Lam discovered after researching the field with cofounder Dan Dolgin, a boob-garment sales veteran who spent 17 years at Vanity Fair. “The more data you have, the smarter decisions you can make,” she added. Unfortunately, Lam had no data.

Creative director Nikki Dekker and Michelle Lam

In 2012, Lam and Dolgin founded True&Co., a shopping startup that uses algorithms to improve the bra-buying experience. Bra shoppers head to the site, where they first take a fit quiz to determine everything from shoulder-strap slippage to breast shape. (Are you shallow or full?) The site then suggests styles and sizes based on those answers. Up until now, shoppers would choose from popular brands like Calvin Klein and Natori. Then a box full of those selections shows up in the mail for the customer to try on and buy or send back, an experience much like that of eyeglass startup Warby Parker.

Starting today, True&Co. offers its own brand of data-engineered garments, the She Walks in Beauty (+ Light) collection, alongside traditional bra brands. With information collected from more than 200,000 women who have taken the fit quiz, in addition to their personal responses after the at-home try-on and purchasing process, Lam believes she has created a line of bras that have a “really good shot” of fitting. “It’s a collection true to our philosophy: There’s no one bra that fits all women,” Lam says.

Some critics argue that cold, hard technology alone can’t address age-old problems like quadriboob or collapsing cups. “My boobs don’t need an algorithm,” wrote Sindhya Valloppillil, a blogger at Open Source Fashion. “An algorithm cannot provide you with a better fit just as answering questions online cannot help you find the best pillow for your preferences. Some products need to be touched and tried on.”

Michelle Lam

Lam doesn’t (entirely) disagree with Valloppillil. “Big data is not the answer to everything,” she admits. “But the design process is not just a machine spitting out a spec.” In fact, she argues that her data-driven approach has broader applications than traditional bra-making. Brand names make their undergarments based on a fit-model process, where they craft, say, the perfect 34C based off of a model. That generally means that a certain type of woman is drawn to a certain brand of bra, and that only a few body types are served in the market.


By using information from hundreds of thousands of women, True&Co. has identified more than 6,000 distinct body types. The company’s in-house collection has attempted to create bras that cater to more of those breast genres.

“There are body-type variations that can be grouped together and addressed with one kind of bra design versus another,” explains Lam. “We’ve utilized the data in the quiz to figure out what are the design features that a certain group of body types would appreciate.”

She describes the process as “Netflixian,” a not-surprising analogy since the True&Co. CTO worked at the streaming service, and one of the start-up’s advisors is Netflix’s head statistician. The company looks at the problem category it defines as “busting out,” for example, the same way as Netflix would look at action movies. When a woman self-describes as “busting out,” True&Co. can look at similar customers’ previous shopping experiences and choices to recommend the right bra. And, just as Netflix used viewer data to invest in original programming that should statistically perform well, True&Co. used its shopper data to create a line that addresses statistically relevant bra problems.

For example, if the Made of Stars bra fell under a Netflix-esque, uber-specific category tag, it might read Bras for the Bustier Woman Who Fears Bulges. The style only comes in D or bigger sizes, and was manufactured with material types to reduce underarm flab–which True&Co. affectionately calls chicken arms–a problem a high percentage of “busting out” women identified. The line includes 10 type-specific pieces like that, which range from $22 to $88.

Indeed, it’s a utilitarian and somewhat nosy approach to an intimate garment. But, in the bra startup world, True&Co. isn’t even the most invasive. Brayola and MeCommerce, two other companies in the field, ask users to submit almost nude photos of themselves as part of the fit test. Rather than flash a picture of a big-boobed woman wearing a too-tight bra, as Brayola does, True&Co. attempts to make things as nuanced as possible. Its fit quiz uses drawings, rather than photos. The garments come in a box with tissue paper. And, best of all, you get to try them on at home, away from the mean glow of overhead fluorescents.


Big data for underwear might sound overly calculated. But when it comes to bras, wearers want form to follow function, but not look purely functional. Finding that sweet spot is actually a science, argues Lam. “Lingerie is one of those interesting things where beauty is predicated on something that’s very objective,” explains Lam. Since the data set includes both aesthetic and practical tastes, True&Co. thinks it has a better chance of crafting the objectively perfect model.

Time and data will tell if statistics sell. The in-house line will have to compete with 50-plus established brand names. A She Walks in Beauty (+Light) design has to catch the eye of the browser. (The name alone probably won’t do the trick.) And, of course, if someone tries it on and doesn’t like it, she can always send it back.

[Ed. note: We changed this story after it was originally published to clarify who, exactly, at True&Co. came from Netflix.]

About the author

Rebecca Greenfield is a former Fast Company staff writer. She was previously a staff writer at The Atlantic Wire, where she focused on technology news