The Body-Shaped Hole in Your Digital Transformation
Why bad fit is endemic to fashion and how body data can help create more engaging and frictionless shopping journeys for the customers when it comes to sizing and fit.
Meet Robin. She’s right in the crosshairs of your target demographic, and she wears a Medium. She follows you on Instagram. She’s bought a few pieces from you over the years. But not as many as she might have purchased, because your clothes might fit someone like her… but they don’t fit her. At least not the way she wants them to. This is why she returned the occasion dress she bought for a friend’s wedding last summer, and why she doesn’t really wear the roll-neck sweater she picked up the winter before.
Robin isn’t alone in feeling like the clothes she wants to love were designed with someone else in mind. In every US state, data suggests that real customer bodies diverge from the idealized measurements used for design and technical development by between 1 and 8 inches at every critical point of measure. And there’s every indication that this is a trend we can extrapolate worldwide.
Robin is also a prototypical digital-native shopper. She can be fiercely loyal if you give her a reason to be; or she can take that loyalty to any number of competitors if you don’t.
This is why the distinction between designing for Robin, or designing for someone like Robin matters to her. If your products are always a compromise on fit, she’s likely to come away feeling like just another number. If they make her feel good, she could become your biggest brand advocate.
But misaligned fit is not just a problem for Robin and other shoppers. If the fit model you used to design Robin’s Medium was the basis for the other sizes in your range – scaling up and down with linear or non-linear grading rules – then your Small, Large, and plus sizes are also not going to fit the customers they are aimed at.
That’s a problem for you. Of $3 trillion in retail sales, around $550 billion gets returned, and it isn’t uncommon for a large brand or retailer to find itself saddled with billions of dollars’ worth of unsold stock in a single year. Bad fit is a huge contributor to both of these statistics.
But it’s not a problem exclusive to you. Because if Robin does turn to another brand, she’s likely to find the same issue. To be blunt: bad or miscalibrated fit is endemic to the entire fashion industry, because the physical and digital dress forms that brands design to are based on abstract, aggregated data at best, and plain assumption at worst.
Fashion has tried to solve this issue before, using size surveys. These were broad-brush initiatives that assessed key measurements of a particular cohort – usually a regional one. Height, bust, waist, low hips, and a few other foundational points of measure were taken for men, women, children and other groupings, and a median value was established for each. The methods used for these surveys improved over time – from tape measure to stand-up scanning booths – but the principles remained the same.
These broad segments are the foundations for most physical dress forms and off-the-shelf digital avatars, with designers and garment technicians making further tweaks for their brands, based on intuition.
But the misalignment between these broad segments and a brand’s real customers can be vast. In one body data project completed by 3DLOOK, it emerged that a majority of the brand’s real female customers were shorter than the fit model by almost half a foot, and all were wider at the waist by between 3 and 8 inches. And this is before the distribution of different body shapes – hourglass, straight, triangle, inverted triangle as so on – were taken into account.
These are not minor variances. And the results of carrying incorrect sizing data through, from design into technical development and manufacturing, are significant.
Meet Robyn. She’s Robin’s digital counterpart: a set of more than 65 different points of measure, all unique to Robin herself. Robyn started life as an optional step on an ecommerce storefront, when Robin elected to provide two photos – front and profile, taken with her smartphone – from which a full 3D model of her body was then generated. That model represents Robin to a very finite degree, with no approximation or aggregation, and Robyn can be quickly matched to the correct size that’s currently being held in inventory.
But the process of creating Robyn is not where her value really lies. Body scanning has been an essentially solved problem since stand-up body scanners were used in size surveys. The replacement of that cumbersome machinery with a smartphone camera or webcam is more convenient, but it’s not world-changing.
Robyn’s revolutionary potential lies in the other value she can unlock – for Robin and for you, upstream and downstream. As a slice of body data that’s individual, accurate, and, after being stored securely and anonymously in the cloud, accessible to other enterprise platforms, Robyn has a lifespan far beyond a single virtual try-on. Because everywhere the limitations of idealized initial fit data became compounded, the value of accurate, personalized fit data can be enhanced.
Unlike traditional production models – where overstocks and markdowns are inevitable – selling more well-fitting clothes can be done without loading up on excess inventory, helping you to cut waste and improve your sustainability credentials.
Take Robin – the physical one. Following a link from your Instagram bio, she’s been browsing your eCommerce storefront, and wondering whether to take another chance on a Medium in a new style she really likes. Presented with an option to get an accurate fit recommendation to help her decide, she snaps her photos and Robyn is born, along with the suggestion that Robin will better fit a Large in some of your styles. This eliminates the need for at least one return.
But while Robin comes away happier, she’s still not 100% satisfied when her order arrives. Because where the Medium pinched, the Large bunches up. Not enough to return, but enough to bother her. Luckily, your designers are now working from a better understanding of their target market’s dimensions – because at the same time Robyn was being created, other customers across your size range were doing the same. Now their combined data points and the insights they reveal will inform product design, fit sessions, and size labelling for the next cycle of development – and those garments will, in turn, be tested by the market and refined further.
Suddenly that endless cycle starts to look like a positive thing – a loop of design, development, sale, and improvement with real customer body data at the center.
That same data can also be integrated with other core product information and housed in a centralized design and development platform – ideally PLM. Body data can then be used to make steady, iterative progress towards other fashion retail holy grails: on-demand manufacturing, augmented reality try-on experiences with simulated fit, and even personalized marketing.
Until your other operations can support these longer-term goals, though, a library of accurate, trustworthy body data can power that ongoing process of refinement – allowing you to make better-fitting products that Robin and shoppers like her can buy with confidence, and know they’ll want to keep. And with eCommerce conversion rates improving by up to 3X, there’s a good chance you could sell more of them, too, making a material difference to your bottom line. But unlike traditional production models – where overstocks and markdowns are inevitable – selling more well-fitting clothes can be done without loading up on excess inventory, helping you to cut waste and improve your sustainability credentials.
In a world where brands and retailers are faced with little choice other than to go digital if they want to compete, the era of relying on historical size surveys and approximate segmentation is over. Because the body-shaped hole in your digital transformation is one your customers could easily walk out through.
This article was originally published on The Interline