Senior retailers and technology providers from around the world converged on New York City for Retail’s Big Show, organised by the National Retail Federation (NRF), where they gained insights into the latest solutions affecting the sector.
This year’s event attracted around 40,000 people from 105 countries representing 6,200 brands who heard about a variety of themes. But the overwhelming focus was on artificial intelligence (AI) as there has been a growing recognition by the retail industry that the technology is going to change many parts of their businesses.
The tone was set by Doug Herrington, CEO of Worldwide stores at Amazon: “AI is becoming transformative. We’ve not had a technology revolution like this since the internet. Mobile was large and social media was also large, but they were nothing like this. It’s requiring eye-opening investment, but the breadth of gains makes it worthwhile.”
Almost every booth in the EXPO halls and all the presentations given over the event’s three days involved AI in some form or another, such is its incredibly broad reach. Very much making its presence felt this year was AI chip supplier and platform provider Nvidia.
Representing the business, Azita Martin, vice-president and general manager of retail and CPG at Nvidia, recommended retailers definitely need to start on their AI journeys if they are not doing so already. Just like with e-commerce, she suggests the leaders in the field will gain advantage.
“AI is real, so get started. It needs executive sponsorship – they need to believe in it. Retailers need to pick specific challenges and assign teams to it and determine the metrics to measure success,” she says.
Among those on the journey already are US-based DIY chain Lowe’s, which has created digital twins of its 1,700 stores. “They can be updated with inventory data to optimise layouts and the way customers shop the stores. They can move things around on the digital twin, just like a video game, without physically having to do it,” says Martin.
Seemantini Godbole, chief digital and information officer at Lowe’s, says: “Our space planning leaders can mark SKUs on the twin so they can move out the low-selling items and replace them with higher sellers and visualise what that would look like. Hundreds of such experiments can be undertaken by our designers.”
Computer vision
The company has also deployed computer vision in its stores, which can help in a myriad of ways when aligned with AI, such as reducing shrinkage by identifying issues in real-time and turning off the point of sale, while also assigning an associate to the problem.
There is also a growing use of AI in the supply chain, according to Martin, who says combining digital twins with computer vision can optimise the throughput in distribution centres. “Optimising layouts is very disruptive, but with AI twin, you can create a warehouse and simulate layouts and see how they behave.”
These solutions are relying on what Martin describes as “physics AI” that is particularly suited to distribution centres as it involves understanding the weight and depth of packages.
Retailers continue to be excited by the opportunity of AI empowering their employees. At Lowe’s, when the computer vision in its stores is aligned with AI, it can allocate staff in store in real time to customers who need assistance.
“The computer vision spots them and then alerts the employee on their [personal] Zebra device,” says Godbole. The device also incorporates a generative AI (GenAI) chatbot that can be used to answer customers’ questions. “We’re giving superpowers to associates. Everyone becomes an expert.”
Computer vision will also be playing a part in Tapestry’s plans. Trang To, vice-president of omni-channel at Tapestry, says: “When it’s adopted, it will have the ability to inform us of who is in the store and where are they shopping. It will help with store design, training and labour scheduling. It will be enormous.”
Similar AI-powered initiatives are taking place at Walmart, according to Greg Cathey, senior vice-president of transportation and innovation at the firm, who says all employees have a phone and access to the company’s app. “Now that GenAI is involved, we’ve democratised data. How do you take all the knowledge in the company and give it to the associates? This is people-led technology power,” he says.
There are myriad use-cases. For instance, by taking in various data points, the technology now enables managers to plan an overnight shift in under five minutes, compared with an hour before.
“We spend a lot of time listening to the operational team. As we build a use-case, what are the KPIs for success? We want to know the ROI, but we start with the associates and NPS scores. [And ask whether] we are reducing friction and pain-points,” says Cathey.
Ellen Svanstrom, chief digital information officer at H&M, says: “It’s all about connecting the digital assets with the physical. RFID technology could be connected to an application layer powered by AI that can then empower colleagues. It will be key linking these things. It is low-cost tech, but it’s a game-changer for us.”
She says H&M invested early in AI and learnt a lot, but realises it needs to up the ante: “We’re now in the second wave, and the pace of change means we need to be more collaborative. We’re investing in our data foundation and platform with our partners.”
These partners can include the likes of data specialists including Coveo and Stibo Systems that help retailers deal with the increasing complexity around the collection and handling of vast amounts of data that is being pulled in from in-house and third-party sources to power the hungry AI engines.
Gustavo Amorim, CMO of Stibo Systems, says: “We enable retailers to gain trust in the data and we support them to ensure the data is clean and the governance is in place. It sounds simple but it is not…especially when the data comes from all sorts of sources.”
Online fashion rental business Stitch Fix works with Stibo Systems for the management of its product information. When combining this with AI, it helps its stylists deliver a more effective service to Stitch Fix’s customers.
Tony Bacos, chief product and technology officer at Stitch Fix, says AI is used to collect the information on customers and trends, as well as to prepare the notes that are given to customers to describe why each item in their “fix” has been selected for them. “We also continue to use AI to drive our assortment to ensure we have the right mix of new and favourite items,” says Bacos.
Volume handling and RFID
The capabilities of AI has created an appetite for personalisation, but the flip side of this is the increased volumes of creative materials that that is driving. Needless to say, AI can sort this issue.
David Hartman, vice-president of creative at Walmart, is working with Adobe to better handle the volumes of advertising and marketing materials that it is creating at its new studio in Los Angeles. “The tracking and management tools has helped us handle the volumes.”
Marta Frattini, director of retail industry strategy and marketplace at Adobe, says: “With the promise of personalisation, there will be five-times the demand for content. It’s mind blowing. GenAI can address this through speed that has never been available before.”
DigitalWave plays in this space, enabling retailers’ creative teams to take a simple 2D photo of a product and provide basic context of the model to be featured. The solution then provides hundreds of alternative images. The retailer can then select the one they wish to use in their campaigns.
Although AI dominated the agenda at NRF 2025, the event was not exclusively focused on this all-pervading technology. At the low-tech end of the market is RFID, but this does not mean it is less exciting, according to Dean Frew, president of RFID solutions at SML, who says only 15% of retailers currently use RFID technology, but he predicts this will change rapidly, forecasting that within five years only 15% will not be using it.
This will be driven by the fact RFID usage has moved beyond simply providing stock visibility in an organisation to also helping retailers with a variety of other factors including handling returns, powering click and collect, and reducing shrinkage. It is now about boosting the customer experience rather than being simply a back-end solution.
Boosting the in-store customer experience was evident in various solutions being showcased in the NRF Innovators Zone, which has grown in size at the event since its introduction. Among them is Mira, which has created the first unsupervised computer vision platform that is powering a self-checkout solution. It removes the need for barcode scanning, and its compact design enables it to be retro-fitted to existing till structures.
Despite these exceptions to the general rule at NRF 2025, there is no doubt that AI is affecting all parts of the retail landscape. We are likely to see it be increasingly utilised by operators of all descriptions, from the largest to the smallest, as each recognises the upsides it can deliver to their organisations.