Apparel fit and inclusivity
With consumer apparel purchasing increasingly moving online, the subject of apparel fit is at the heart of fashion e-commerce. In this article, four industry insiders come together to merge their differing viewpoints:
Mark Charlton - Fit & Sizing Thought Leader / Strategist - https://www.linkedin.com/in/mark-charlton-fit-leader/
Jessica Couch -Fit Tech Consultant - Consumer Tech Product Manager - Fit Tech Expert Strategist- Body Positivity Champion #womenintech https://www.linkedin.com/in/jessicavcouch/
Richard Irons -Lead Engineer at Rakuten Fits Mehttps://www.linkedin.com/in/richard-irons/
Emma Hayes -Plus-size Womenswear Sizing and Fit Consultant and Expert linkedin.com/in/emma-hayes-3a58553b
‘Diversity inclusion’ is a term used frequently by corporations intending to ensure everyone has a voice and that there is equal representation of gender, race, religion and other human variations. Equally important is diversity of thought.
So how does this concept relate to the fit of apparel?
Each week brings fresh potential technical solutions to bear on the current apparel fit problem. This is a Good Thing, as the tech geniuses are recognising fit as an area where technology can offer a significant contribution.
It’s our opinion that most of these advances are instigated and developed within the somewhat rarefied environment of the tech industry – employing one very specific way of thinking. We note – not as a criticism, but as an observation – that there is an opportunity to redress any imbalance of reasoning by introducing some art into the science.
This observation is not a novel one: for example, it is supported in principle by The Medici Effect (Harvard Business School Press, 2004), which explores why the most powerful innovation happens at the intersection where ideas and concepts from diverse industries and disciplines collide.
Apparel fit is part art/emotion and part science/tech
Think about the last time you purchased a garment that fitted amazingly… how did it make you feel? Apparel fit speaks to, and stimulates, the senses. It creates an emotional connection greater than the sum of its parts: much more than mere body dimensions and garment measurements.
So what’s raising the age-old problem of apparel fit among the tech solutionists?
E-commerce apparel return rates are eroding brands’ and retailers’ margins and profitability. As e-commerce continues to grow, this erosion can no longer be sustained… or masked.
But as a consumer, what do I care? If I don’t know what size I am, know for certain that I will like a certain product, or that it will suit me, I have the option to order it anyway – perhaps in multiple sizes – hoping to figure out for myself whether it will work.
We all know that so-called ‘free shipping’ and ‘free returns’ are, of course, nothing of the kind. It’s these delivery costs, coupled with the task of processing returned products back into inventory, and attempting to balance stocks when over half of demand is returned, that are causing the margin erosion and higher prices to the consumer.
Reasons for high returns
Apparel e-commerce return rates on average hover around the 50% mark – 70% of which are attributed to poor fit. It’s a cliché, but for such a tiny word, ‘fit’ is a very complex process!
To put it in a nutshell, ‘fit’ is where individual consumers’ body measurements meet brands’ sizing and garment specifications; designers’ fit ideas meet consumers’ fit preferences; real-life material properties meet consumers’ fabric expectations; and designers’ styling decisions meet the pace at which consumers are willing to adopt trends.
Many of us are aware that in future we will be able to take 3D scans of ourselves from our mobile phones or similar devices. These will generate accurate avatars of our bodies, complete with all our measurements, upon which we will be able to virtually ‘try on’ potential purchases – checking our images on-screen in three dimensions for how good the fit is, and whether the style suits us.
At the time of writing, all over the world, many apps, devices and methods are being developed that are advancing rapidly towards this dream. For example, there is an app on which you can see a three-dimensional avatar of your body – complete with measurements – after simply taking front and side view photographs on your phone. Another app allows you to upload pictures, and your virtual-reality self will then try on the clothing of your choice – draping naturalistically. There is a clever hand-held device that takes your measurements by scanning you. There are even smart body suits and scanning pods, which offer the promise of the gold standard of human measurement: a perfectly accurate rendition of your entire body in three dimensions. These all exist today at various levels of development.
Such devices are exciting and headline-grabbing, but it’s unlikely that most of the companies selling us apparel online will opt for them quite yet – partly for technical reasons, but also because they need to be integrated into the systems currently employed in the fashion industry. In the early stages, retail companies will need to ‘grow out’ their operation to merge with the technology – and many changes will be required.
Fit tools are clever online algorithms that work out which sizes of apparel need to be ordered, based on ‘inputs’ – and it is these tools that are making the big inroads right now. Inputs are various pieces of customer information – weight, height, age, perhaps body measurements, ordering/returns history, and body shape – which the consumer loads into the tool. In the near future these will also include personal preferences. A vital ingredient of these tools is profound apparel knowledge, allowing them to match the consumer with the optimum garment.
Even at this early stage, this tech is proving to be effective – the best tools boasting a considerable reduction in the number of product returns. They also have the advantage that they are already doing a lot of the heavy lifting required for the digital transformation of the fashion industry. This is what is building the infrastructure that will plug into all the extra data that’s collected.
The human angle
However, like all new technologies there are going to be issues surrounding adoption by the public. Predictably, the tech people may think that the problems are all centred on the technology, but there are considerable social, psychological and emotional difficulties to overcome. As consumers, we have to learn how to travel around this new technology.
Whatever tools we use, we are asked to take some time gathering – and inputting – information. But there are problems with asking people to do this, and they fall into two categories…
The measurement problem
Studies show that our measurements are in a state of constant flux, so measuring will not be a one-off activity. We are being asked to continually monitor our measurements and weight – possibly on a monthly basis – regardless of whether we use a tape measure or scanning device.
There are confidentiality issues to think about. If we are not going to have to keep repeating ourselves with every company we buy from, we will have to develop methods whereby our information can be shared between various organisations.
Our experience is that people only substantially change their behaviour and attitudes when there is something in it for them, and that something often has to be more important to them than a new pair of jeans – even if they fit beautifully.
The phrase ‘conform to new habits’ fills consumer experts with a mixture of dread and concern. Can we consumers really be expected to be ‘educated’ into new habits? In our leisure time (and shopping is supposed to be that) most of us want to undertake enjoyable activities with an instant reward, rather than toiling through worthy chores in the hope that something better will come along later.
We need to create usable, enjoyable tech that will draw everyone in from the inception; ideally, fun tech that we don’t even notice we are using.
The revelation problem
The second problem is revelation. Many people don’t know, don’t want to know, don’t believe and/or would never tell you their accurate measurements.
We need tech that is ‘unconscious’: having given our permission for the data to be collected, we should have the right not to have to have any interface with our body metrics unless we choose to do so.
The future of fit technology
Fit is becoming a buzzword and everyone has an answer to the online returns problem, but the best solutions have two qualities:
1. Ease of use– How simple and convenient the solution is: mobile phones vs. specialist devices for example?
2. Ease of integration– How easy it is for brands to integrate the technology into their current systems?
The best technologies do not try to train users to have habits that are not simple or natural. They allow end-users easily to add technology into their everyday lives. Accuracy is key, and the less effort required the better.
Neither do the best technologies try to do everything. Instead they connect to existing technologies and enhance outcomes.
Many smaller brands find it difficult to integrate fit technology because their current ‘solutions’ are unable to connect to other solutions, and buying an entire suite of IT products is not an affordable option.
Expensive, rigid technologies are out. The best technologies are those which integrate easily with existing platforms and create more efficiency. Because tech has not existed in fashion in the past, many departments are siloed and are not properly integrated for it. Great technology companies have to take this fact into consideration before they can succeed.
How fit is your competitive advantage?
Fit and fit technology are customer experience tools– A lot of brands believe that implementing more lenient return policies can somehow impact the quantity of returns. In our view this is similar to putting a Band-Aid® on a gash… it simply doesn’t treat the real issue of customer expectation.
According to an article on online apparel returns myths:
• Most returns are made by one-time buyers.
• Good returns policies do not affect sales.
• Most shoppers don't think about returns before buying.
• Most people are not concerned with free return shipping.
• Bad returns policies don't affect sales, and a returns policy won't impact future sales.
By the time a customer has had to return an item, you have lost them for future opportunities. Customer expectations must be met and returns avoided. This can be done through building confidence with consumers, whether in-store or online, and helping them understand what to expect in regard to fit.
Fit and fit technology are loss management tools– Implementing fit technology helps to increase consumer confidence in products. $62.4 billion worth of apparel and footwear is returned every year due to incorrect fit. That works out to about 57% of footwear and 64% of apparel purchases, according to a recent Footwear Newsstudy. The same study found that if fit were not a concern, 51% of respondents would purchase footwear more often, both online and in-store, while 58% would purchase clothing more frequently.
Excellent communication around fit is important because it helps build confidence with the shopper – increasing sales and generating fewer returns. Implementing fit technology tools that create directive shopping experiences and manage expectations can help to reduce the amount of unsold inventory.
Fit can help reduce fashion’s carbon footprint– A recent op-ed piece published in The Business of Fashion revealed that dead inventory (unsold clothing) costs the US retail industry $50 billion a year. Although brands may be able to absorb some of these costs through write-offs on the balance sheet, the environment (through landfills, toxin pollution, etc.) cannot.
Newsweek published an article stating that Americans alone produced 15.1 million tons of textile waste in 2013 and around 85% of that ended-up in landfill, according to the Environmental Protection Agency.
Fit technology allows brands to create better-fitting clothing for shoppers, and helps to match them to their products – so clothing is not created unnecessarily, quickly ending up in landfill. Although changing the shopping habits of consumers is a difficult task, brands have to take more responsibility for their impact on the environment. Implementing fit technology can help to fix fashion’s misaligned supply and demand issues.
Fit is inclusive: more people shopping equals more money– In a survey conducted by Fung Global Research, some 72% of respondents did not believe that fashion designers create their designs with the average American woman in mind.
Approximately 78% of people would be willing to spend more money on clothing if more designers offered plus-size options. Some 68% are interested in participating in fashion trends, but 67% feel that there are not as many fashionable clothing options available in their size as they would like. This isn't just a plus-size issue.
According to a Business Insiderreporton petite people, over 70 million US women fall into the ‘special’ size category, and 50% of the population is under 5' 4" tall, but brands' size offerings do not reflect this. In addition to these categories, there are also tall women, big and tall men, petite men, and people with physical handicaps that are also opportunities for brands to target.
Fit Tool Desired output
When thinking about creating a fit tool, firstly, it’s necessary to think about what is needed from that tool. For instance, whilst producing a custom-made dress, a pattern with all the correct measurements will be required from the outset.
However, in this piece we’re not talking about bespoke garments, but clothes that are already manufactured, and are available in a finite number of sizes.
When shopping in a store for clothes, most consumers who are not sure what size to pick opt to try them on – and when a size doesn’t fit correctly they may examine different sizes until either finding a good fit, or deciding that none of the available options are suitable. It’s this process that we want to duplicate in a fit tool – essentially the algorithm “tries on” every available size on a body, selects the best size for that body, or concludes that none of the sizes are any good.
So really what is being asked from a tool is “best size, if any”.
Ideal garment measurements
In future, if manufacturing processes change so that fit plays a greater part, we may want the tool to provide us with a list of “ideal” measurements for a garment. This could, for example, be used as input into some sort of electronic manufacturing system that makes every garment to order.
But perhaps this is jumping ahead.
In order to get the best results from a tool it needs consumer information to work with. To return to the analogy of trying clothes on in a shop, there are two things involved: a body and a garment. A tool needs information about both.
Clearly, a fit tool needs the body in question’s measurements, and the most obvious way of obtaining them would simply be to measure with a tape, the way a tailor would. This is actually the best way to get accurate metrics, if it were a professional who was undertaking the measuring. However, for a customer at home, it’s not a great system. Firstly, the subject needs to possess a tape measure, and secondly, would be willing to stop in the middle of shopping in order to take measurements.
These issues are problematical in themselves, but worse, the majority of people don’t know the correct measuring method, so will ultimately supply inaccurate metrics. And if the data is inaccurate, there’s no way the tool can give a good result.
AI method - "pertinent questions"
An easier and more reliable way to get the information needed is to ask the customer some pertinent questions – age, weight, height – simple information that people already know about themselves. Once it has this information, a tool can use a neural net, armed with a great deal of knowledge that has been previously collected, to deduce that user’s measurements surprisingly well. This method is usually significantly more accurate than asking consumers to measure themselves.
The information that is required about a garment is a little more complicated. It’s not enough to simply know the physical dimensions (although these are necessary), since other considerations, such as how closely the garment is meant to fit at certain points, and how stretchy the material is, must be taken into account.
The easiest place to get this information is from the manufacturer. All the details about the apparel’s dimensions, the fabric’s ‘handle’, and the design’s ‘preferred fit’, are known to them, because this information is needed for the manufacture of the garment. However, sometimes the retailer doesn’t have a direct relationship with the manufacturer and won’t have access to that information.
Without these details, it’s necessary to use one of a number of methods. The most accurate would be for a garment technologist to acquire the apparel in each size and undertake accurate measurements, using their expertise, along with product photography to judge the preferred fit. However, with a large number of products, this approach becomes prohibitively expensive. Other available methods include generic size charts, information from similar garments, and artificial-intelligence inference from product descriptions and photography.
Manufacturers who want to make sure that an accurate fit could be calculated for all their products would be best advised to make all the measurements and design information easily and freely accessible.
If this became an industry norm, customers would find obtaining a good fit much easier, and the level of expectation and competition would ultimately cause manufacturers to raise their game with regards to fit.
To make sure a tool is reliable, developers need to check that the results make sense. There are certain ways to do this.
One simple method is for a specialist to test tools by entering lots of different measurements and then see if the recommended size “looks right”. Of course, this method can be subjective and inaccurate as, for example, it depends on the manufacturer’s idea of “size 10” broadly agreeing with the technologist’s.
More accurate testing can be done, albeit more expensively, by buying garments in the recommended sizes for many people of different shapes and sizes, and judging the fit when trying them on. Information from this process can then be fed back into the tool to improve its accuracy.
In Conclusion....Mark Charlton
The diversity that exists across the human race meshes with the complexity of each fashion brand's design aims, layered to the multiplicity of fabric properties and fit preferences, both of designers and consumers. These issues create a mind-bogglingly intricate problem of achieving the perfect fit.
But this is only part of the challenge: for example, optimal fit can also differ across POM’s (Points of Measurement). An instance of this would be where stretch jeans would require greater elasticity in some areas than in others, so that there is flexibility on the hips, but a snug fit on the waist: a combination of variable body shape, but also of preference.
No individual company, however great their resources, can solve the fit question in isolation: one brand can hope (at best) to supply a solution for their own apparel-which only represents a fraction of their consumer’s overall fit needs.
We need the vision to collaborate with fit solutions across the entire fashion industry, whilst still competing in this space. A necessary step towards this is to understand that we must solve the issue of apparel fit by rising above simply thinking of it as returns problem. It is far more important than that.