clock menu more-arrow no yes mobile
An illustration of a woman in a changing room looking at herself in the mirror with dresses hanging behind her.
Clothing sizes are all over the map.
Sarah Lawrence for Vox

Filed under:

Anyone with a body knows that clothing sizes are flawed. Could there be a fix?

What does it even mean to “be a size 8”?

If you needed to explain how human beings buy clothes to, say, an audience of aliens, you would sound fairly ridiculous. Designers turn flat fabric into 3D body coverings in a limited number of sizes and even more limited number of shapes, and then we purchase what we can, hoping for the best, becoming angry at our bodies for not corresponding to the coverings instead of the other way around.

Buying clothes online has only complicated matters: The fraught process of finding an appropriate and comfortable fit isn’t exactly easier when it’s undertaken as an information-deprived guessing game while drinking wine at home. Failing to solve the problem of fit isn’t an option. Returned products eat into companies’ bottom lines as unwearable clothes contribute to environmental waste.

It seems like these problems should be quite solvable, and yet they persist. A theoretically logical solution to the confusion around clothing sizes has been proposed many times over: standard, universal sizes that don’t vary between brands. The idea is that some entity would step in and create The One True Size Chart, and brands would label their clothes with the accepted official sizes, solving all confusion around buying clothes forever. It’s never going to work in the US — and for very good reasons.

Size universality is an alluring idea that appeals to our need for order, but order is a false god that would actually make it harder to buy clothes by limiting the shapes and dimensions available for our diverse bodies. Instead, we need something to help us navigate the existing size variation. That helper is new tech about to transform how we shop for clothes online. Recommendation widgets that guide us to the right sizes, apps that use a cellphone to measure the body, and brands offering affordable garment customization are about to make it easier to buy clothes that fit without ever visiting a dressing room.

What if sizing variation between brands is actually chaotic good?

While clothing sizes for men’s bodies grew out of the Napoleonic wars and the need to clothe soldiers, it was the approach of WWII that necessitated a grand codification of the bodies of women, who wore uniforms while working in factories to make war supplies. In 1941, Ruth O’Brien and William Shelton published Women’s measurements for garment and pattern construction, a report based on their Work Projects Administration study. The project recorded the weight and 58 measurements of primarily middle-class, young, white women.

“The 1941 study was the basis of all women’s clothing sizes” in the postwar era, says Lynn Boorady, department head and professor at Oklahoma State University. It “had additional studies” added “over the years and the data was massaged but it never strayed far from the standard ‘hourglass’ shape.” Despite the fact that just 8 percent of women in the US have true hourglass shapes — and that women of color need clothes — postwar mainstream clothing was dominated by a silhouette made for young white women wearing girdles. In the last two decades, new sizing standards put out by ASTM International and Size USA updated measurements and shapes. “We saw the new sizing for ‘curvy’ or ‘straight’ figures right after the Size USA data was available” in 2003, says Boorady.

We now have better and more diverse data about what bodies in the US are like, but that doesn’t mean that the next step should be a universal size chart that creates standard sizes used by every brand.

If we impose a universal size chart, “a lot of people are going to go without clothes,” says Kathleen Fasanella, an apparel industry pattern maker with 40 years of experience in the industry. The reason comes down to anthropometric variation. Anthropometry is the study of the measurements and proportions of the human body. Sometimes it’s used for racist pseudoscience (example: phrenology) and sometimes it’s used to understand the range of foot sizes in a given country. The US has a lot of body dimension and shape variation because we have a lot of genetic variation. A universal size chart in the US would be a disaster — and it would likely center on the dimensions and shapes of white bodies, as the 1941 O’Brien and Shelton study did. “We’re not going to be able to standardize sizes until people’s bodies are standard, and they’re not,” says Fasanella. “Sizes are a social construct. What’s a size 10? Nobody can agree.”

Brands don’t want universal size charts. Think of a brand’s clothing shape, the dimensions of each size, and the amount of ease in their clothes as a form of intellectual property; it’s the secret recipe that makes each brand special and keeps customers coming back. In fact, brands take their sizing and fit so seriously that they create clothes that fit their customers, veering away from published size standards as they gain more data. “Size charts are the beginning of the process; design and fit must also be taken into account,” says Boorady. According to Fasanella, “Everybody sizes to their market, to their customer base.” Even if every brand were working with the same size chart, that doesn’t mean they would produce clothing of uniform dimensions.

This tailoring of the brand size chart and fit to their customers’ bodies and preferences has been called “vanity sizing,” often represented as a craven and confusing practice that preys on womens’ desire to be either smaller or larger depending on the garment type. In reality, brands are re-centering and adjusting their clothing dimensions so they can make clothes for their customers. The dimensions of sizes may have gotten bigger over time for some brands, but a whole lot of that is related to wonderful things such as the fact that we’re not usually wearing girdles and clothes are being made to fit the bodies of women over 30.

Why shopping online for clothes is so damn hard

If you’re uncertain about your size in a particular brand and cannot try things on in a retail store, the logical thing to do is hunt down the brand’s size chart on their website. For US women’s brands, the size chart usually lists the bust, waist, and hip measurements in inches for each size sold. It’s a great resource, if often ferreted away and hard to access on mobile.

A woman’s sweater size chart from Patagonia.
Screenshot of

Here’s the problem: When I surveyed my Instagram followers, a highly unscientific sample of over 1,000 respondents indicated that only 75 percent own a measuring tape meant for bodies (as opposed to a more rigid metal measuring tape meant to aid home improvement projects). If we do happen to have a measuring tape at home, one study indicates that we’re not always great at knowing how to use it to take clothing measurements: On average, errors ranged from -4.54 cm (1.79 in) to +6.15 cm (2.42 in). With an error rate like that, a measured customer could easily buy clothes at least one size off.

When you do manage to take decent measurements, then there’s the harrowing decision of how to wedge your body into a size chart that might not line up with your shape. If the measurements and size line up, that doesn’t mean that the suggested size will align with how you see yourself; many people identify with a particular clothing size and a change can be jarring. Finally, even if your perfectly measured body matches very neatly with an accepted size, your fit preference might necessitate a different size altogether.

Fashion’s surprisingly awesome sizing anarchy requires a matchmaker operating between brands and customers. That’s where tech currently delivers for fashion e-commerce.

Fit prediction: a guide through sizing confusion

At this moment, on certain online shopping sites, you can type in some data about yourself and get a personal size prediction for that brand. Brands hire firms such as Fit Analytics and True Fit to advise their e-commerce customers on the size they should buy (neither responded to requests for comment for this story). Online-only and plus-size retailers such as ASOS, Lane Bryant, and City Chic have invested in customer-facing size adviser widgets, but so have brands with strong brick-and-mortar footprints such as Macy’s and Kohl’s.

Upon clicking either company’s sizing helper link on a product page, a window pops up and asks about things like the customer’s height, weight, age, hip and midsection shapes, bra size, preferred garment fit, and favorite size in competing brands. Not all questions are required, but answering more questions improves the quality of the prediction. Depending on the company, one or multiple size options might be recommended.

Learning your size without taking off your clothes might sound like a totally new invention, but the concept actually has long roots. O’Brien and Shelton, authors of the 1941 sizing study, determined that height and weight were actually good at predicting one’s body dimensions. The problem was that people didn’t want to walk into a store and disclose their weight — and they may not have even known their current weight since home scales were less common. More recently, attempts to predict size have failed because customers didn’t understand why certain data was collected. “J.C. Penney tried doing this about 12–15 years ago with pants; they would ask questions about bra size and it completely confused people. But it can be a valid question,” says Boorady, since “if you have a fuller bust, you are likely to have a fuller hip.”

Fit prediction technology has some challenges. The links to fit widgets tend to be so unobtrusive on a desktop browser that some customers don’t even realize they’re available. Many brands cut the link to size prediction for their mobile sites or insert it inconsistently even on desktop sites. Since size prediction hasn’t been discussed much in the popular press, customers may wonder why they need to share information such as their age with a company trying to sell them pants (answer: age impacts how weight is distributed) or have concerns about how their data might be used. For customers who have deliberately stepped away from using a scale, prediction based on height and weight isn’t going to work.

Despite the current limitations, fit prediction widgets are the fashion tech that’s already more or less working for e-commerce. At this point, the main hurdle for fit prediction seems to be a lack of public awareness, since most attention has gone to high-profile projects with far shakier track records.

Can body measurement projects and mass customization deliver a perfect fit?

The most visible fashion/tech collaborations to date have produced uneven results ranging from brand-specific measurement projects to avatars that creep us out to customization that doesn’t always fit us. At some point soon, these technologies will combine into a truly useful system for buying clothes that fit us in every sense.

Japan’s Zozo launched an ambitious custom clothing project in 72 countries starting in 2017. The Zozosuit, named one of Time’s best inventions of 2018, was a social media hit, utilizing fairly cute spotted stretchy body tights that worked with a branded app and cellphone camera to capture lots of measurements without a measuring tape. The capturing process took some troubleshooting, but the measurements themselves were shockingly good: The Zozosuit even managed to measure inseam and torso length without the need for a helper. The problem came when doing more with those measurements other than wondering how you came to be so asymmetrical; customers said that the “custom” Zozo clothing didn’t fit and took forever to ship. Profits for the brand fell when sending free fit suits didn’t translate into sales, and Zozo pulled out of the US and Europe in April 2019.

Excitement around 3D renderings of clothes on customers’ avatars similarly collapsed in recent years. Fashion media and academics predicted that Clueless-style avatars would help sell customers on style and fit when shopping online. Alessandra Vecchi, senior research fellow at the London College of Fashion and principal investigator for the e-Size project, says that her team found that using an online shopping avatar or a rendering of the customer’s own body would frequently discourage purchases and that an avatar “often acts as a turnoff to the user.”

More promising is mass customization. Brands like Indochino and eShakti offer “custom” or “bespoke” clothing that allows for the garment to be made for each individual based on their measurements, with room for style changes such as longer sleeves or a wider neckline. “This is not someone drafting a garment for each individual, it’s selecting pre-made [pattern] pieces based on the measurement to customize [the garment],” says Fasanella, which helps to explain why such clothes are often priced only a bit higher than ready-made garments.

Uneven reviews of eShakti and Indochino suggest that the pattern database technology has promise but hasn’t always been successful in delivering clothes that fit well. As customization technology improves and measuring technology along the lines of the Zozosuit emerges to help customers capture accurate measurements, experts predict that mass customization will help reduce waste and returns while boosting customer satisfaction and profits. For now, ordering made-to-measure clothes online requires a leap of faith.

Despite technology’s current solutions to the fit problem flopping or offering modest improvements on the status quo, the fashion industry is betting on tech solving the problem in the future. Amazon Body Labs is currently offering $25 gift cards to customers who consent to be scanned at their NYC project office. A workshop in Copenhagen last month dove deep into how fashion recommender systems that draw on everything from preferred fit to personal style can be refined and expanded.

Vecchi says, “I think [fit recommendation systems] will become the next big thing” together with “actual measures with the aid of your mobile” and algorithms that “take into account other ‘softer variables’ like fit, any possible preference for the degree of softness of the garment fabric to the touch, smell of the fabric, and so on.” The solutions to fashion consumers’ fit problems are slowly coming together and there’s reason to hope that they’ll link up soon to create brand and size matchmakers that will actually help us find clothes that fit — even if we’re shopping while drinking wine.

Sign up for The Goods’ newsletter. Twice a week, we’ll send you the best Goods stories exploring what we buy, why we buy it, and why it matters.

Donald Trump

Trump is suddenly in need of a lot of cash. That’s everyone’s problem.

Supreme Court

The Supreme Court will decide if the government can seize control of YouTube and Twitter


Why Lyft and Uber drivers did their largest strike ever

View all stories in Money

Sign up for the newsletter Today, Explained

Understand the world with a daily explainer plus the most compelling stories of the day.