Artificial intelligence has stirred up retail for years, but has become of particular interest since late 2022, upon the public-facing launch of OpenAI’s ChatGPT.
What started as much of the world’s foray into generative AI has become an obsession with personalization, back-end efficiencies and omnichannel strategy—all powered by technology. But where is retail’s attention headed next, and how can AI realize its full promise with the help of other, less flashy technologies?
Brendan Witcher, VP and principal analyst for Forrester, hosted a session with Trang To, VP of Omni for Tapestry, Ellen Svanström, chief digital information officer at H&M and Karen Etzkorn, EVP and chief information officer for Qurate Retail Group at the National Retail Federation (NRF) conference in New York City on Monday. The group discussed AI’s current capabilities and where complementary systems could be headed as retailers implement technology-based solutions.
Related Stories
Is generative AI being realized in business?
At last year’s NRF conference, discussions about retailers and brands’ lofty expectations for generative AI dominated the conversation; not all of the use cases leaders expected to implement came to fruition, but many retailers are beginning to see results from the systems they have put in place, both on the customer side and for internal use cases.
Like many other retailers, Tapestry has used generative AI to guide the customer experience on its brands’ sites, helping consumers find products that match their customer persona. But some of the most apparent displays of generative AI’s true impact comes because of its internal implementation. For instance, Tapestry has seen a 40 percent reduction in code and a 15 percent reduction of re-coding among engineers, To said.
And while technology certainly changes corporate employees’ jobs, Tapestry has put the spotlight on what generative AI has done for its in-store associates. As the company works to gather greater insights about how consumers want to shop its brands in physical locations, it has given employees a tool to share observations. Initially Tapestry launched the tool with Coach; it calls it Tell Rexy, a play on the brand’s T-rex mascot.
“That tool aggregates feedback from all of our associates and leverages AI to develop themes for action,” To said. “We’ve collected tens of thousands of pieces of feedback from our…associates.”
Tapestry has now launched a similar program for its Kate Spade brand.
Over at H&M, generative AI use cases are still cooking. While Svanström said the brand sees “enormous potential” and has begun to invest in the technology, it has yet to see fully baked business impact from the technology that saw so much hype last year.
Personalization’s future
Personalization seems to be a buzzword in the retail industry, and while many companies have started to deliver on the promise of tailoring the customer experience, other still have a ways to go.
In the conversation, To said she believes that retailers that already have a fairly robust personalization system in place may start to look to hyper personalization—that is to say, segmenting consumers into much smaller groups or communicating with them as single individuals.
But with that shift could come a problem: scaling.
“The biggest hurdle for us in personalization is the speed at which we can create creative content to address all individuals versus segments,” she explained.
But that issue could soon resolve itself with other AI applications; Tapestry has started piloting AI-generated imagery, which has the potential to disrupt the way it communicates with and markets to consumers at an individual level.
What type of technology can complement AI best?
One of the most intriguing promises of AI is that it can enable stronger outcomes for legacy technology, like radio frequency identification (RFID). Algorithms can use the data collected by RFID to offer analyses of what’s happening in a retailer’s store. The same is true of other low-cost, sensor-based technologies.
Svanström said H&M has already started to experiment with such technology, primarily to better understand how shoppers navigate the company’s stores and which pieces of its assortment are hot at any given moment.
“The enormous interest in AI and gen AI also amplifies other related technologies,” she said.
And computer vision, which is a type of AI that is typically enabled by run-of-the-mill sensors and cameras, also has the ability to revolutionize how retailers think about stores, To said. In implementing computer vision software behind existing hardware, retailers can unlock insights about customer demographics, associates’ interactions with consumers and more.
“That information has the power to transform how we think about store, design, merchandising, store associate training and even labor productivity,” To told the audience.
