As artificial intelligence starts to replace human resource, Drapers examines how retailers are using the technology to leverage business opportunities
Success in fashion retail was once predicated on having an instinct for the next big thing. But in an omnichannel, social media-saturated consumer environment, human powers of divination alone – even those informed by trend forecasters – are not enough.
The best fashion businesses are increasingly combining human with artificial intelligence (AI) to improve customer service, identify consumer trends, offer greater personalisation and to ensure their supply chain is as responsive as it can be.
The system recognises human speech and routes calls to relevant departments, and will be used in all 640 M&S UK stores by the end of this month, as well as its 13 UK call centres. It has the capacity to handle more than 1 million customer calls a month. M&S said 100 switchboard staff will be reassigned to roles in store.
Although M&S is using AI to create efficiencies and redeploying skilled staff elsewhere, a report from the Oxford Martin School at the University of Oxford last year forecast 80% of retail jobs are under threat from robotics and AI.
“We’re going to see millions of lower-level frontline data-heavy jobs move over to AI,” says Doug Stephens, retail industry futurist and founder of consultancy Retail Prophet.
However, many view AI as more of a saviour than a threat to human employees, as the technology becomes a key resource. AI can help fashion retailers predict and react more quickly to trends. The technology has the ability to crunch vast swathes of data, from catwalk shows and social media channels such as Instagram, to colour and style trends, enabling designers to create collections faster and merchandisers to plan more efficiently.
Within a few hours, AI can do what it takes fashion analysts a week to do
Danny Bagge, digital front office lead at IBM UK and Ireland
IBM worked with Tommy Hilfiger earlier this year to demonstrate this technology in a project called Reimagine Retail, which used AI to identify upcoming trends to enhance the design process.
“In fast fashion, merchandising and design teams’ time is buried in that boring [trend analysis] research, which is non-value-adding because every fast fashion retailer is doing the same research,” says Danny Bagge, digital front office lead at IBM UK and Ireland. “Within a few hours, AI can do what it takes fashion analysts a week to do.”
Joel Edmondson, C&H business development at M&S, sees AI as a way to tackle issues with productivity: “There is such potential in AI in making every human more productive, not putting them out of work.” Edmondson heads M&S’s AI-powered personal shopper service Try Tuesday, where shoppers share their preferences online and a team of stylists pick outfits for them.
“Every [Try Tuesday] outfit is built by a real stylist, and we use AI to make sure that outfit goes to the right people and is edited in the right way,” he says. “We use AI to let the stylist focus on the most creative parts of the job.”
Personalising the shopping experience relies on AI to process vast amounts of customer data. In this way, retailers can try to recreate a long-lost intimacy with their customers, redolent of a time when shopkeepers knew their customers by name.
“We can use AI to reclaim a lot of intelligence we had about our businesses when they were small – when they were the local merchants that understood their customers, trends and each of their markets,” says Stephens.
Technology, which some have blamed for contributing to the death of the high street, could also, paradoxically, help breathe new life into stores by offering customers a tailored experience – for example, providing sales staff with access to customer profiles and vast amounts of personal information on the individual customer and offering features such as smart mirrors in changing rooms.
We can use AI to reclaim a lot of intelligence we had about our businesses when they were small
Doug Stephens, retail industry futurist and founder of consultancy Retail Prophet
Brands from H&M to Burberry use chatbots to answer customer questions or offer advice. But AI is helping personalise interactions further with tools such as visual search, where shoppers can take a photo of an item of clothing they like in the real world and be directed to similar items on a brand’s site.
Asos, which recently announced it is boosting investment in AI, offers visual search, product recommendation and a fit analytics tool, where customers can virtually compare the fit of items they want to purchase with that of clothes they already own, with the aim of reducing returns.
Ben Chamberlain, lead data scientist at Asos, sees AI as integral to the brand’s operations: “Eventually, the more difficult question to answer will be ‘where isn’t AI used at Asos?’,” he says. “We see a future where AI is the key operating system for all of our digital products and, beyond that, making important contributions to how we buy, design, store and transport our products.”
Retailers are also using the technology behind the scenes of their operations to make stock more targeted and reduce markdowns and wastage. H&M, which has previously stocked its store portfolio with similar products, has started using algorithms to help decide which stock should be carried at which store with the aim of improving poor sales performance.
AI’s potential to help sell more items at full price will significantly boost profits for retailers, IBM’s Danny Bagge says: “Fashion retailers are saying to us, ‘what do I design for my Chinese market versus my London market?’, and going further, ‘what do I design for my London store versus my Birmingham store?’ – and that’s when you start adding money to the bottom line [with AI],” he says.
When you’ve got AI suggesting what customers want and re-ordering, then you get less waste and fewer returns
Cate Trotter, head of trends at retail consultancy Insider Trends
Conversely, if a product is flying off the shelves, AI can optimise the price and prevent the store running out of stock.
Cate Trotter, head of trends at retail consultancy Insider Trends, says: “You can predict the demand on a product will go up, which is not just better for sales, it’s better for long-term customer loyalty.
“When you’ve got AI suggesting what customers want and re-ordering, then you get less waste and fewer returns.”
But for all the buzz around AI and machine learning, the tech is still in the early stages and has limitations.
Mark Leach, managing director of the agency User Conversion and former head of ecommerce at Missguided, notes that many in the industry wrongly assume that AI can solve every issue: “There’s a misinterpretation of what it’s capable of. Some of the use cases are still a bit more about showing off the technology. Where it has the ability to make an impact is in data and analytics.”
AI is going to help brands feel that they can take informed risks, and that’s important
Doug Stephens, IBM
Ultimately, to succeed in fashion retail, the fundamentals have not changed, believes Stephens. Brands need to genuinely fascinate and be of value to consumers, he says. But AI’s ability to shift through data and help predict what is coming next could make impressing those consumers easier in future.
“Consumer expectations are that every time they come into your store, they’re going to see something new and exciting,” he says. “[AI] is going to help brands feel that they can take informed risks, and that’s important.”