Retail stores of the future will be powered by generative AI, but truly frictionless shopping experiences envisioned by tech vendors may be a long way off.
Retailers may be bedazzled by generative AI, but for many companies it will take much data wrangling and investment to realize bottom-line results from the technology.
That was the general sentiment from tech buyers, vendors and other attendees at the National Retail Federation's NRF '25: Retail's Big Show held in New York Jan. 12-14.
There's much potential for generative AI to reduce labor and deliver more real-time insights up and down the supply chain and into stores, said John Furner, Walmart U.S. president and CEO, and Azita Martin, Nvidia vice president and general manager for retail and consumer packaged goods in a keynote sit-down conversation.
But the futuristic scenarios presented by more than 125 vendors with AI software and hardware on the NRF exhibit floor -- such as sensors that detect spills in a warehouse aisle and generative AI digital twins that swoop in, redirect forklifts around it and dispatch people to clean it up -- require data, people and processes to set up and execute.
Take, for example, personalization, which has been a difficult puzzle for retailers to solve and one that generative AI tech can potentially solve.
"We all know people in our lives, we know our customers individually," Furner said. "But in many cases, [retailers] are trying to understand, recognize and learn about millions of people at the same time -- and the data is always changing."
E-commerce personalization occurs with web search, Martin said. Generative AI understands searcher intent and sends that data downstream to enhance digital commerce on mobile apps and websites.
"Search has become AI-enabled, and it uses multiple generative AI models," Martin said. "It uses catalog enrichment, reasoning and embedded text, and all of that enables your e-commerce site, your mobile apps, to really deliver a much more personalized experience."
At NRF '25: Retail's Big Show, John Furner, Walmart U.S. president and CEO, discusses the potential for generative AI to aid retailers with Azita Martin, Nvidia vice president and general manager for retail and consumer packaged goods.
Some retailers stuck with legacy tech
While Nvidia's keynote showed what can be done with its fastest chips on the bleeding edge of tech, tech buyers on the NRF show floor were still trying to figure out tactical applications; things like how to use predictive and generative AI for analytics and to simply show store associates what is and isn't available in stock.
Market leaders -- especially companies that launched digitally -- such as Amazon, Walmart and eBay are ready for advanced generative AI tools, said Oscar Mullin, vice president of technology at Mercado Libre, an online retailer with 100 million users in South and Central America. His company has embraced generative AI for chatbots to interact one-to-one with shoppers but also beyond that to create text, image and video content.
Retailers with large online footprints have evolved their own technology quite a lot, Mullin added. Most of them either started in the cloud or have migrated their operations there. Most have used AI in the form of machine learning for more than 10 years to enable product recommendations and forecasting.
"Those that are leading are more than ready to build on top of this new technology," Mullin said. "But I also think these new technologies are going to enable many other new competitors to join the game, particularly in the digital space. It's easier than ever now to create a digital product, and it's easier than ever to handle customer support and things like that because you don't need a huge workforce and you don't need complex workflows to do that."
SharkNinja, the popular maker of vacuum cleaners, kitchen appliances and other household devices, is about to rip-and-replace various legacy applications in favor of an integrated Salesforce stack. The company wants to take advantage of generative AI tools embedded in Agentforce to connect with customers on a deeper level.
For me, [the benefit of generative AI is] the ability to take a lot of things and be able to move the data … to just process so much and to articulate it in a way that we can do something with it.
Velia CarboniCIO, SharkNinja
While SharkNinja does have a considerable direct-to-consumer operation through infomercials, it also sells through retailers such as Target. Driving more customers to its website, automating some aspects of customer service and more personalized marketing are the initial goals of its generative AI adoption, said Velia Carboni, SharkNinja CIO.
"For me, [the benefit of generative AI is] the ability to take a lot of things and be able to move the data … to just process so much and to articulate it in a way that we can do something with it versus having to throw a lot of bodies at Excel spreadsheets," Carboni said.
Some retailers are starting with GenAI with e-commerce sites, remixing content with basic ChatGPT or putting data into Perplexity and polling it for ideas, said Dominik Angerer, CEO of content management vendor StoryBlok.
"You can ask your own data questions," Angerer said, "One thing AI is actually really good at, the way we have it right now, is to analyze data and give you feedback.
"It's really important to still check if that feedback is valid," he added, "because there are a lot of hallucinations still going on."
Cashless store runs on GenAI
The most advanced application of generative AI in retail -- the latest iteration of the autonomous store -- was also on display at NRF.
One of the exhibitors, SenseiTech, is popular in Europe and claims to have the largest autonomous store in the world -- a 1,200-square-foot supermarket in Lisbon. No cashiers, cash or scanning is needed. The tricky part? It requires many surveillance cameras in the ceilings, tracking each individual shopper through the store, and taking a payment when they step into a small space before a checkout kiosk. Then, because of EU privacy rules, the store must forget customers unless they opt in to sharing it.
SenseiTech illustrates the limits of collecting consumer data for personalization -- and by extension, the potential limits of generative AI to execute it -- as more privacy rules come online in the United States, too. Such regulations, said SenseiTech founder and CEO Vasco Portugal, make loyalty programs that are much more valuable to retailers.
Nevertheless, he is bullish on GenAI, for not just retail, but also all verticals.
"I don't think there's industry that won't leverage AI," Portugal said. "It's this general-purpose technology, like the internet, so at some point it will be part of [everything]. Companies need to embace it."
Don Fluckinger is a senior news writer for Informa TechTarget. He covers customer experience, digital experience management and end-user computing. Got a tip? Email him.
