top of page

Fashion Tech Files: Can AI get us perfectly fitting shoes and clothes?

The perfect fit... whether it is clothes or shoes, we have grown accustomed to the fact that what we wear doesn't fit to the T. Finding the perfect fit is a challenge consumers and brands share. It is in the best interest of a brand to get their customers the right fit, ideally on the first try. The customer will look and feel great in the product, and brands can avoid costly returns and logistical hurdles.


At the same time, as an industry, fashion has moved far away from made-to-measure or custom fit. The majority of us don't own anything that was made or tailored specifically for our body, wedding garments, and those of us who make clothes for ourselves aside. We buy clothes and shoes in "our size". A concept that really only suits industrial production as it allows the manufacturing of large quantities of similar products. Take shoes, for example. From a mathematical perspective, our feet are two three-dimensional volumes. Many people think their feet are mirrored images of one another, but even that is not true. A couple of decades of standing on them daily will make your left foot different from your right foot, even if you can't see it visually. These two 3D volumes are reduced to one number vaguely related to the length of your foot: your shoe size. It is only normal that this is not an effective strategy to get you shoes that fit your specific feet.


One size fits all


In fact, the easiest way to get shoes that always 'fit', is to get shoes with no fit at all. If a shoe is reduced to a larger clunky object, like crocs, a heavily cushioned structure, or a piece of fabric as an upper, like many sneaker designs, the exact shape of your foot doesn't matter anymore. The downside is that it is usually not the most elegant footwear. The same is true for clothing. The one-size-fits-all trend is very interesting for brands and eases buying stress for consumers. Only, most of those clothes really don't fit anybody well.


Consumers are often frustrated with the fact that sizes are not the same for every brand. For clothing, there isn't a fixed measurement per size, making a size EU36 or US 6 not the same in every brand. However, shoe sizes are based on exact mathematical measurements. That doesn't help that much though, since your foot and shoe length is only a small part of the fit. The shape of your toes vs the shape of the toebox can make a big difference in the apparent size.


Technology to the rescue!


There are already a multitude of tools out there that can help customers and brands buy their size. Body or foot scanners turn your measurements into data points that you can compare to the exact shape of a garment or shoe. So far, it hasn't become mainstream though. The success of a technology like that is contingent on consumer willingness to get their body scanned and enough big brands paying for and integrating those tools in their online store. It is hard to get that level of adoption.


Another tech tool is based on comparison. Based on data from lots of customers and lots of clothing items/shoes, the application looks for similarities between the sizing of popular brands you already own versus the brand you are planning to buy. Based on a short questionnaire, the tool will tell you what size you are for that particular design. This tool has been gaining some traction, especially for online multibrand platforms. However, it works best for more simple constructions of clothes and shoes. Not every shoe or garment is a good fit for every foot or body, even if 1 or 2 of the measurements overlap. A jacket that fits nicely around your shoulders and waist but with sleeves that are too short, will still not be a good fit.


Data data data


Left and right, brands are trying to revive made-to-measure, which once upon a time was all that fashion was. However, since production facilities are not designed for this way of retail, it is often still not made-to-measure but rather an approximation of your measurements. Also, to simplify the process of taking measurements since the customer is usually responsible for doing it themselves at home, mistakes happen and measurements that could be essential to fit are omitted.


3D printing is quickly gaining traction as a tool to combine all of that data to create truly custom footwear. For now, it is still limited in what materials and designs are possible, with most of them looking either clunky or more fabric-like.


Artificial Intelligence


The latest hope for fashion tech to solve the sizing challenge is artificial intelligence (AI). As databases full of measurements of human bodies from all over the world are merged with databases of clothing and shoes, we are getting closer to matching bodies with fashion.


However, the problem each of the above-described tech solutions - including artificial intelligence - share is that fit is personal. I have clothes from size XS to XL in my closet and it is not because sizes differ by brand. Tight-fitting or oversized are both fashionable looks. Shoes are slightly less disparage in fit, but after selling shoes for ten years, I can tell you there are also strong personal preferences there. Some people prefer their shoes to be very tight, others like it looser, others like it tight around the heel but loose around the toes, etcetera. While technology might be intelligent enough to take that hurdle as well, it would require all of us to know how we like our fit without really ever being able to compare it to someone else's experience.


If you have ever had something custom-made or tailored for you, you might remember that it was a multi-step process. You had to go back for at least two fittings before it was perfect with a professionally trained seamstress adjusting it exactly to your body or feet probably asking you many questions in the process including if it feels comfortable for you. If you like it at that height or length? Etcetera. Today's technology is not capable of doing all that. And it may never be.


At the end of the day, fashion is an important way for each of us to express ourselves. Show who we are. That we are unique. And that includes how something fits. Reducing not just our bodies but also our personal sense of touch and our preferences into data points might never be desirable, even if it is possible.




Recent Posts

See All
bottom of page