All this extra clothing being bought remotely is resulting in a mountain of product returns. In the UK, levels of e-commerce fashion sales that ended up being sent back reached £11.4bn in 2020
. At least half of returns (and, in the growing plus size sector,possibly considerably more) are reported as being sent back due 'fit problems' (Bizrate Insights
survey of 1,052 consumers in June 2019, for example, found 55% of consumers said size was the top reason why they returned an online purchase). So, looking to the future, it’s clear that e-commerce apparel size and fit is a huge, ongoing (and indeed, growing) problem.
With garments that are sold 'sight unseen', the fundamental feedback loop is a binary one. Either a dress is kept, or it is sent back. It fits or it does not. On or off. Unless the retailer puts some effort into obtaining a lot of extra information about either the consumer or the garment, very little is learned about fit each time a product is sent back. And even if the retailer does pursue feedback from their consumers, if it is not exhaustive enough, it’s all too possible to learn the wrong lesson altogether.
Back in the old 'analogue' days (when customers were fitted in bricks-and-mortar stores), it was easy to see with human eye the intricacy involved to fit for apparel the wide range of diverse female body shapes that are found in the population.
Take two women, for example, both of whom have exactly the same height, waist,and hip measurements (although with differing body shapes), trying on identical dresses in the same size. Woman A says that the garment is too tight on the hip, yet B says that hers fits perfectly. How can this be, when both have exactly the same size hip? But A is pear shaped: she is very small on her top half. Larger parts of the body take more fabric to cover (not just widthways, but also lengthways), so this lady’s small top half has taken up less coverage, meaning that the narrowed waist of the dress has drooped down and is actually sitting on the wider part of her hip, causing it to be too tight in that area. If A were at home trying on this dress, she would send it back, and if she was asked, it’s fairly likely that she would tick the box indicating 'too small'. Actually, the dress is a little too large (on the top half). The data point about size gained from this transaction would be at best meaningless, and at worst counterproductive, if a simple 'tick-box' question was asked.
There is nothing 'wrong' with the size or cut of this dress: it is simply not the right shape for this consumer, which illustrates the inadequacy of relying on the measurement/size grids that are often the only resource that consumers are offered when deciding which size to order online. How would A be expected to choose which size of dress, when she is smaller on her top half? The best option for her would actually have been not to have bought this style at all, because it evidently does not fit a pear-shaped woman, but this information is too complex to be made clear. It is for this lack of clarity that many women who have diverse body shapes do not bother with the retailer’s sizing guides, choosing either to rely on brands that are familiar to them, styles that they have had previous success wearing, or customer reviews.
Here’s another instance: a lady returns a blouse and gives the reason for doing so as that it was 'too tight on the arms', which is bewildering, because it turns out the styling of the sleeves is extremely wide, meaning it would be almost impossible for them to be too tight on anybody’s arms. In fact, the raglan-style armhole is cut too deep, meaning that the sleeves are at an impractical angle to the body, restricting movement and making them feel tight. This is a manufacturing fault, but it is not being reported as such by the consumer, who, not surprisingly, doesn’t happen to be a pattern cutting expert. So again, the 'tick' ends up in the wrong box.
There are thousands of other examples of the complexity surrounding the issue of customer fit, which not only show how insufficient the crude on/off binary of 'keep-or-return' is, but also go on to illustrate how a cursory survey undertaken into the fit of returns is in itself likely to be insufficient,especially as the customer herself doesn’t always know what the problem is.
Of course, there is nothing intrinsically wrong with a binary system: provide enough little squares containing either a one or a zero, for example, and it’s famously possible to arrange them in such a way that they depict a fairly convincing black and white image of Marilyn Monroe. But the more complex and detailed the picture, the more data points it is necessary to obtain. And apparel fit is incredibly complex. Clothes are not like other consumer products: one cannot compare, for example, the return of a set of curtains (which happen to be too small) with that of a blouse, which could be said to be 'too small'. There can be no doubt that, using the blunt tools usual in the sector at the moment, it would be completely impractical to expect a customer to give the quality and quantity of information necessary to enable the retailer to reliably understand the reason for the return of the blouse, in the same way as they can, with confidence, simply explain why the curtains were sent back.
It’s easy to conclude, then, that it is pointless trying to obtain any information from consumers about apparel returns, yet I would argue that there is a need to gain information in any and every way possible. It all goes back to the pixelated image of Marilyn: in order to create a detailed picture, it is necessary to obtain as many data points as possible, so it would be unwise to ignore this potential resource.
There can be no doubt that garment fit is something that should be tackled primarily at the 'business end' of the transaction between retailer and consumer: the point of sale, rather than when it is being returned, when it is apparently too late. Before anything is sent out, there should be a thoroughgoing investigation as to the shape of the human being towards whom the garment is being directed. I would argue that, in the future, much (if not all), of sizing and fit information, pre-purchase, will be obtained using some kind of technological imaging of the consumer. Such is the complexity of the human body; it would be unrealistic to expect to gain enough detail any other way. And this automatic analysis of people is not a one-off: in order to cope with the ever-changing body shapes and sizes of people as they move through their lifetimes, this will need to be a continuous process. But the detailed knowledge of the body shapes of consumers would be of limited usefulness if it is not matched, both with a perfect understanding of the measurements, fits and shape of the garments being retailed, but also with a much wider choice of sizing and grading of online brands to offer each consumer.
It would therefore be wasteful to ignore the stream of information that can be gained from the return of garments, particularly when developing an appropriate inventory. When approached with a subtle and thorough system to understand the reason why an item is being sent back, a return survey can be a useful tool to identify faulty pattern cutting or wrongly graded items in general, as well as suggesting body shapes that are being poorly served by the brand. Also, it is one 'extra level' of information about fitting the individual concerned.
Human beings are not just bodies: they are minds and personalities as well. Each subject has a set of 'fit preferences' that govern how they prefer to wear their apparel. Perhaps an individual likes to wear tight clothing all the time. Or only when they are exercising. Maybe this person prefers baggy attire when working out, but close-fitting outfits when they are socialising. These are highly personal preferences that make up a 'fit ID', which can be borne in mind when trading to them in the future.
Some retailers have the luxury of a huge number of sales, creating statistics from which general predictions and trends can be extrapolated. Perhaps certain preferences turn out to be universal, such as, for instance, a ubiquitous choice to prefer a certain style of dress in a smaller size in black and a larger one in white. Consumers should be heavily incentivised (with free postage, gift vouchers, points and special offers) to provide much more detailed information to help retailers understand if some fit choices work, and others don’t, across as wide a spread of the population as possible.
Arguably, the return information presently being obtained by fashion e-commerce is fairly half-hearted: considering the waste involved (both financially and ecologically), much more attention needs to be given to the reasons why apparel is being sent back as unsuitable. Motivating the co-operation of consumers, asking the subtle, incisive questions and expertly analysing the answers is a form of art (and should be a profession in its own right) that can offer a meaningful picture as to why some garment sales end up in our growing mountain of returns.