The Clothing Industry Is Spending Millions to Crack the Code of Virtual Fitting

Amidst the stay-at-home monotony, I’m trying to figure out if the hem on the black, pima cotton “classic” T-shirt, size small, from Buck Mason’s web store, will be too long on me. Maybe the extra small would be better? But then, historically, extra small T-shirts have always felt a bit too… accentuating. My brain is turning into applesauce, but I spend nearly 30 minutes on Memorial Day evening toggling back and forth between the small and extra small T-shirt, torturing myself in this digital retail wasteland before abandoning the purchase entirely. 

Why? 

Because I know in my bones (which are clearly smaller than most) that despite the “size guide” — which makes no mention of the length — and despite the ostensibly helpful information that the model in the pictures is “6-foot-2, weighs 175 pounds and wears a medium,” there’s just no good way for me to know which size of this unremarkable $35 T-shirt will fit me. It’s the all-too-familiar dance with the online retail machine that gives you every imaginable clothing option, but leaves figuring out your size to your imagination. Will the size 32 jeans from Levi’s fit the same way as the size 32 pants from Banana Republic? Probably not, but there’s no way to know until you’ve tried them on.

“Our chief analytics officer has a PhD in physics from MIT,” says Romney Evans, co-founder and chief product and marketing officer of True Fit, a data-driven platform that helps retailers decode fit and size for their customers. “He’s an astrophysicist. And he talks about how this [online sizing] problem is one of the hardest problems he’s ever worked on.”

In some respects, this sizing problem is the byproduct of a different sizing solution. The reason there’s no standardized sizing really just boils down to the huge variability in human bodies across the population. “Probably the best way to solve it would be to have tons and tons of different sizes that are both length and curviness and width dimensions,” Evans says. “And just create tons of permutations of sizing.” 

But since that’s not practical from a design and manufacturing perspective, brands have had to create products for target segments of consumers who they believe their products will resonate with. “Every brand caters their sizing system to who they believe are their core customers,” says Evans. “As such, their sizing system becomes a source of competitive advantage.” An extreme example would be to take “a size four at Abercrombie and Fitch that’s catering to a 17-year-old and an 18-year-old, and compare that to a size four at Chico’s.” “You’ll see that the waist measurement could be inches different,” he continues. “The reality is, and we’ve done studies on this ourselves, we can see that the average waist measurement of an individual as they get older increases. So if you think of the size distribution of a population of 45- to 65-year-olds, it would look very different than, say, the size distribution of the same number of 14- to 23-year-olds.”

Buck Mason, the company whose T-shirt I spent my holiday weekend struggling with, claims their core customers are men between the ages of 18 and 65, which, clearly, indicates a wide swathe of body types. Because of this, they offer a “virtual fitting,” which sounds like the future, but is only so in name. In reality, it’s a video conference call with one of their style experts, who, with remarkable grace and patience, gives me “virtual” — albeit two-dimensional — access to their brick and mortar store. 

“I’m here for any questions that you may have about sizing or color or whatever,” says Jose, the style expert. “I’m here at the store, so I have access pretty much to the entire collection. If you have questions — like, maybe you have a pair of jeans and you want to know what to wear them with — I can also help you with that or how to build an outfit.” The idea behind the virtual fitting process, he says, is to replicate the in-store experience as extensively as the two-dimensional camera will allow. But again, as far as sizing goes, even Jose admits that it’s more difficult if you’re unfamiliar with the sizing scale of their brand.  

As you can imagine, this problem, in a world where brands are relying on us to make our purchases from home, is the online retail industry’s equivalent of the Enigma machine. “We’re monitoring COVID and we’re seeing online traffic at this exact time, now versus last year, is up about 88 percent,” says Evans. Figuring out a way to integrate a sizing solution for buying clothes online, then, is truly the final frontier for digital retail.

But solving this equation is hardly a new endeavor. In 2017, TechCrunch reported that Amazon had “acquired Body Labs, a company with a stated aim of creating true-to-life 3D body models to support various B2B software applications — such as virtually trying on clothes or photorealistic avatars for gaming.” “One source suggested the price-tag Amazon paid for Body Labs could be $100M+,” per their report.

That’s only one of the more recent virtual solutions re-envisioning the retail experience. According to Matthew Szymczyk, CEO of Zugara, a company that develops and “licenses augmented reality software and creates natural user interface experiences for brands,” this quest for the perfect sizing solution has been underway for over a decade. “Let’s see, it’s 2020. So for the last 12 years, we’ve been doing virtual dressing room technology,” says a noticeably exasperated Szymczyk.

Szymczyk tells me that they started by doing e-commerce versions using just a two-dimensional camera where they could step back and use their webcam, to see what a garment would look like on themselves, “like, if their skin tone looks good with that certain color,” he says. “Initially, it wasn’t about the fit.” But the issue for the entire augmented retail reality space is that people have increasingly stopped using their computers to shop. “Most people moved to m-commerce — mobile devices,” he says. “You can’t hold your phone six feet away from you. So that was kind of the issue, and there are other technical issues with Flash dying out.”

The technology was repurposed in 2012 with three-dimensional cameras and used “in in-store event-type kiosks.” “That was really driven by demand, so we shifted in that direction,” says Szymczyk. “So, the last eight years we’ve been focusing more on that.”

Still, Szymczyk believes that while there’s a lot of challenges with virtual fitting room technology right now, it really boils down to this: The camera-based solutions aren’t going to provide fit for the customer. “It’s just not possible yet,” he admits. “The other issue is, for these solutions to work, you have to take off all your clothes. It’s not like an X-ray. It’s not going to go through and be able to tell exactly if you’re wearing loose boxers, it’s not going to be able to tell where the thigh is inside the boxers.”

Historically, yet another problem with a virtual fit approach to sizing is that consumers are very discerning, says Evans. “[The technology] has to look really good, and it just doesn’t yet.” Another challenge, he continues, “is with product life-cycle challenges, based on the way people design garments and the fact that a huge percentage of garments sell super quickly and they’re never seen again. So the process of creating digital assets that are compatible with a virtual try-on experience, there are a lot of technologies that all have to come together at the same time.” 

Which is to say that, even though the virtual fit technology has developed some pretty realistic-looking garments to appear on a body, there are other issues with being able to scale that process.

Complicating things further is the fact that sizing is still based on what Evans calls “aspirational shopping.” “Seeing something on a form that you aspire to look like versus something that’s realistic is another issue,” he says. “There’s competing tensions to get all of those things just right in a way that a consumer will accept and that the industry can scale, which is complicated.”

And even if the kinks were ironed out, Evans still isn’t convinced that the hyper-realistic virtual approach to sizing would be the best way forward. “There are all these kinds of competing forces,” he says. “On one end, some consumers want it to look really realistic so that they can understand how well an item’s really going to fit them.” On the other, he says, “the closer you get to realism without actually achieving it, the more disconcerting it is to us as a viewer. Some psychologists describe it as being an instinct that we have that helps us avoid death — things that look lifeless.” 

In other words, to some degree, the future of fashion and virtual reality came and went in 2012 with the introduction of in-store virtual kiosks, and we as consumers decided it wasn’t what we wanted. 

For those reasons, Evans tells me that True Fit — which in 2018 was valued at $500 million — took a different approach. “What True Fit does is very similar to Spotify or Pandora for music,” he says. “We basically allow a fashion shopper to create a little profile where they tell us some things they currently wear and what size they wear on those items, which helps give us a sense of their personal style. It also gives us a sense of what currently fits them.” 

Taking that information, True Fit has done the painstaking work of mapping what they call the fashion genome, “which essentially just maps all of the important attributes about a garment or a shoe to the buying behavior of millions of consumers.” “Based on what you tell us that you like, we can translate what other items will likely suit your style, your fit and your size as you’re shopping,” Evans explains.

The questions True Fit asks are mainly about a person’s body type. “Things like height, weight and age,” says Evans. “Then, if they’re shopping for a pair of jeans, we might ask you, what’s a pair of jeans you currently wear, what brand is it, what size do you wear, and you can tell us about multiple items if you want, but one is enough.” 

The result is that, in real-time, True Fit gives a size recommendation on every item that you see. “There are other contexts in which we can give you style recommendations,” Evans adds. “We can actually help find the items that best suit your style, fit and size, they kind of float to the top of the assortment.” 

Currently, more than 250 sites around the globe — from clothing retailers like Macy’s and Levi’s to Under Armour — use the True Fit system for helping consumers find their perfect size in a given garment. “If you go shopping on Gap’s website, you go to the product detail page and right near the size selection area for a product, it offers to enter your information,” says Evans. “You do that once, and then you should see your [size] recommendations on all the pages that you visit.” Essentially, Evans says, they input whatever the customer’s sizes are in apparel they currently own and “translate their experience to this new brand.”

Szymczyk — who, again, has been developing virtual software in this space for the past 12 years — agrees that “True Fit is on the right path.” “There are some people trying to do avatars where you create an avatar yourself, but that’s never really taken off,” he says. “People have been trying that for the last 10 years.” But Szymczyk reasserts that until a virtual experience can replicate the tactile feel of something as it fits on you, it won’t ever be a good enough solution. “You’re always going to run into that issue — no matter how well you can represent it virtually, it’s not going to ever override that feeling of where it’s bunching up under your arm, around your shoulder,” he tells me. “You can’t really replicate that.”

Intrigued at how this all compares to my Buck Mason experience (which, like so many smaller stores, doesn’t have the True Fit integration), I reach out to Fiona Leahy, Buck Mason’s customer experience expert. She tells me that, initially, “it wasn’t an idea for a virtual fitting room, but an idea to pick out what you wanted online and have it ready to try on or have ready to pick-up in-store.” “[It] was a back-burner project that we decided to put forward during this time with our curbside pick-up,” she continues. It is, then, less a virtual fitting room than a “one-on-one customer experience for those who have never been to our stores before, and may have only shopped online.”

It’s a smart move, and despite not really living up to its “virtual fitting” promise, for them and all the other retailers struggling to mitigate their customers’ fears around going into their actual stores, it’s one that will have to suffice. At least, until the virtual technology catches up.

---

Andrew Fiouzi

Andrew Fiouzi is a staff writer at MEL.