Using tech to tackle retail’s returns problem

UK fashion retailers have long struggled with the cost of returns. But as online retail has risen sharply in recent years, it is a cost that many are now struggling to absorb.

A survey in February 2022 by fashion AI company Dressipi of its UK clients, which represent 15% of the online womenswear market, found that around a third (29%) of clothing and footwear purchased is returned.

A February 2022 survey by fashion AI company Dressipi of its UK clients, which represent 15% of the online womenswear market, found that returns rates rose 1 percentage point from August to October
2021 compared with 2019 levels. It estimated that represented 90 million more returns in 2021 than in 2019, and predicts that will reach 121 million in 2022, costing £1bn.

Practices such as “bracketing”, whereby shoppers buy multiple sizes online to try to get the best fitt and return all but one, have fuelled the increase in returns. Meanwhile, the cost of living crisis is forcing shoppers to tighten their belts: one footwear retailer tells Drapers anecdotally that his returns rates have risen since the beginning of September.

For some companies the solution lies in putting a stop to free returns. In May Zara announced it would begin charging £1.95 for online returns, although returns to stores remain free. Boohoo followed in July with a charge of £1.99 for returns.

However, this may a risky strategy: a survey of 7,873 European consumers in May 2021 by market researcher Nielsen found that more than two-thirds (67%) of consumers will abandon online shopping carts, across all sectors, because of returns policies. As a result, many brands and retailers are turning to tech to resolve the problem.

From digital avatars, to personalised sizing tools and algorithmic questionnaires, here are some of the latest technologies being rolled out by the industry now.

Next-generation virtual try-ons

Virtual fit in stores took on mainstream appeal during Covid lockdowns that closed changing rooms for months on end. Bodyscanning technology is used to create avatars of shoppers on to which digital garments can be overlaid.

H&M offered shoppers the chance to trial these 3D bodyscanners in two of its German stores, one in Berlin and one in Hamburg in October and November 2021 respectively. A survey by provider NeXR Technologies of the 1,025 shoppers who tried the virtual fitting room showed 60% found it useful to identify the right size. It says  plans for any future rollouts are “classified".

It is online, however, that these tools are really gaining traction through investment and acquisitions.

In August 2021, Gap acquired Drapr, an ecommerce start-up that enables customers to quickly create 3D avatars and virtually try on clothing. The brand owner has said it plans to implement the technology as part of its Bodequality initiative for US men's and women’s wear brand Old Navy. The brand has removed categories such as "plus size" and "one size fits all" online to create greater inclusivity.

In March, Walmart announced its acquisition of Israeli virtual clothing start-up Zeekit, which has worked with Asos, Adidas and Tommy Hilfiger. It is now integrated into Walmart’s “Choose My Model” tool: online shoppers can pick from 50 different models to “try on” clothes ahead of purchase, filtering by skin tone, height and body shape. Walmart plans to launch a further 70 models.

Social media companies have also spotted an opportunity. In March, Snapchat partnered with clothing brand and materials research company Pangaia to debut a virtual try-on AR campaign using the platform’s in-built tools. Three months later, Snapchat said two-thirds of shoppers were less likely to return products bought using the feature.

James Fox, associate director at Canvas8, an agency that advises fashion brands on consumer behaviour, says: “This highlights both the business and sustainability benefits of consumers getting comfortable with tech-led shopping.”

Making more of the metaverse

While virtual fitting rooms allow shoppers to simulate how clothes will look and fit using a 3D model of their approximate height and weight, avatars are the key to real accuracy, believes Emma Ridderstad, co-founder and CEO of Warpin, a Stockholm-based immersive tech start-up that works with brands such as H&M and Disney.

“By entering their vital statistics, customers can now create photorealistic avatars of themselves to try on clothes without even leaving their bedroom,” she says. “Seeing what an item of clothing will look like on your body, rather than that of a model, or to test it in various background scenarios, will undoubtedly result in fewer returned items.”

One example already in use is YooxMirror. Developed by Italian online clothing retailer Yoox Net-A-Porter Group, users dress digital avatars of themselves in virtual clothing. In October 2020 it launched an upgraded version of the tool that can be used to try on 50,000 digitalised products on the retailer’s site, and a “share” function that allows users to send their avatars to their social media channels.

The metaverse is revolutionising the way we shop.

Emma Ridderstad, co-founder and CEO of Warpin

The tool has helped drive a 70% reduction in returns since 2018, says Ridderstad, adding: “The metaverse is revolutionising the way we shop, and it can also revolutionise the impact of returns.”

In fact, the role of avatars in creating a more accurate fit, reducing returns and waste, is also being used at the design and manufacturing stage. In launching its latest product update in February, 3D technology provider Browzwear said its customers, which include Patagonia, Nike and Adidas, would now be able to arrange 2D pattern pieces directly on 3D avatars to easily make tweaks to shape and style.

Using AI to customise size and fit

Size and fit remain the biggest factors driving returns in UK fashion, and several brands and retailers are integrating AI-powered sizing tools, both to  create greater sizing accuracy during design and help customers select the right size at the point of purchase.

In May, online retailer The Very Group partnered with US provider True Fit. The platform, which already works with retailers such as Quiz, Kate Spade and Boden, leverages AIpowered machine learning to provide tailored size advice to shoppers. The technology asks consumers to upload the details of an item, from any brand, that they feel “fits them perfectly”. It then uses AI to recommend personalised advice for each item clicked on. To reduce size bracketing, an automated pop-up appears when consumers add multiple sizes to their basket, nudging them to use the size fit tool instead.

True Fit claims its tech reduces retailers’ average returns by 5%, and cuts size bracketing-related returns by 40%.

In March, AI firm 3D Look, which has worked with workwear brand Dickies and womenswear denim brand 1822 Denim, also rolled out the latest iteration of its mobile bodyscanning technology. Its "YourFit 2.0" software combines a photorealistic virtual try-on, instant feedback on what size would fit best, personalised product recommendations based on consumers’ unique body shape and measurements, and a social-sharing feature that encourages shoppers to share their try-on. The company says the technology reduces product returns by 6%.

“Six months into using YourFit, we’ve seen great results: our size-related return rate is at a record low and nearly half the market average,” says Leila Shams, CEO and founder of online US-based swimwear brand TA3 Swim. “We also see little to no bracketing.”

Asos, meanwhile, says it continues to expand on its own plug-ins that deliver more accurate sizing advice: “We’ve invested in AR [augmented reality] and AI technology, including a size recommendation tool that offers customers personalisation by recommending them a size based on what similar shoppers have bought and kept,” says a spokeswoman. “The technology allows customers to visualise the fit of
products and pick the right size for them.”

Crunching data to spot serial returns

Many brands and retailers are leveraging comprehensive data sets to deliver more accurate recommendations for customers, and thereby reduce returns.

For example, personal styling service Stich Fix, which launched in the UK in 2019, has developed its own algorithmic tools to tailor size and fit. It is now sharing the data gathered from its “Shop Your Look” function with the brands it stocks, too, supporting athleisure brand Beyond Yoga last year with the launch of its plus-size range.

A Stich Fix spokeswoman says: “Beyond the baseline questions of clothing sizes and measurements, we ask questions that allow us to more deeply understand clients’ preferences. How do you like your pants to fit? What challenges do you typically have with sleeve lengths? What areas of your body do you like to show off?”

This continues after purchase, when the brand requests online feedback on size, fit and preference from customers when they receive items. “These all help us continue to refine our algorithms and create spot-on recommendations in the future, to ensure the items we send are a perfect match.”

Equally many retailers are crunching their inventory data to understand better what is contributing to returns – and take action.

“With immediate insights, retailers can observe the SKUs [stock-keeping units] that are most returned, and decide to put the price up, increase shipping costs or decrease promotions on problem products instantly,” says Drew Smith, co-founder and CEO at retail tech company Upp Technologies, which works with fashion brands such as men’s formalwear retailer Charles Tyrwhitt.

“Machines can learn to identify bundles of SKUs that work well together or, conversely, are beyond the returns threshold and take action to either promote or de-prioritise them immediately, while reporting this back into the business.”

Using automation to create seamless returns

With some level of returns an inevitability of online retail, using AI and tech to automate returns processes – and thus reduce their impact on the bottom line – is another key tech innovation. When The Very Group opened its Skygate fulfilment centre in 2020, for example, creating a more efficient returns process was top of mind. The East Midlands state-of-the-art fulfilment site deploys an automated pocket sortation system. Once a returned item has been manually inspected, its RFID [radio-frequency identification] label can be scanned and dropped into a pocket, where it is automatically categorised and put in the right sequence for customer orders at packing stations.

One strategy retailers are engaging in more is proactive communications with consumers.

Marie Hamlin, retail director at Deloitte

The returns process now takes less than 24 hours, and most items are resold in 72, says Very's fulfilment director Michael Wheeler: “It has allowed us to refund our customers more quickly, also making suitable items available for re-sale faster than ever before – improving product availability.”

Marie Hamlin, retail director at Deloitte, observes fashion retailers seeking to simplify the returns process at the consumer end: “One strategy retailers are engaging in more is proactive communications with consumers specifically encouraging the use of more cost-e cient returns options, such as going in store or using locker drop-off points.”

In July 2021, for example, Marks & Spencer rolled out a new click-and-collect service that lets customers make returns, as well as pick up orders at 78 of its stores using self-service digital screens and drop-off bins. It also updated its mobile app to allow users to choose how they want to return purchases, using product barcodes to navigate the process. As of March this year, the 78 stores with automated returns drop-off available had collected about 70,000 returns.