Walmart and several other major retailers are investing in artificial intelligence (AI) transformation. This includes bots and machine learning that create reports for sales teams, help customers find and purchase products for special occasions, and help store employees resolve issues with business processes.
Walmart CEO John Furner recently said in a social media post that the retail giant is building the future of retail with the help of AI.
“But we know we don’t have all the answers,” Farner said. “Part of having a strong culture of listening and learning is knowing that the best way to stay on the cutting edge is to be close to the people who are at the cutting edge of technology.”
He said a recent meeting with Jesse Cheung, co-founder and CEO of San Francisco-based Decagon, convinced him how Walmart is approaching AI.
• Start with the problem, not the technology, and solve real frictions to make life easier.
• Turning information into action because the real value of AI comes from applying insights in ways that are relevant and useful to customers.
• Expectations are instantaneous, so retailers need to act faster. Customers want choice and want to get the right product quickly and seamlessly.
Donna Morris, Walmart’s chief human resources officer, recently announced plans to train its 2.1 million employees in some level of AI skills. Morris said Walmart has been experimenting with generative AI starting in 2022. She said the company uses up to four different agent platforms combined with custom builds and external large-scale language models. The company has more than 35,000 employees working in the technology industry, and AI is impacting every part of the organization, she said.
Brands that follow
Eric Howerton, co-founder of AdFury, said many companies that sell products to large retailers are not very adept at AI.
“Retailers are releasing very specialized AI tools to help customers find and purchase products,” Howerton said at a recent webinar event hosted by Vendormint. “I feel like retailers themselves are definitely leading the way, but I feel like brands haven’t really caught up to speed.”
Cheryl Yarbrough, vice president of partnerships at New Nexus Group, said during the webinar that AI is one of the biggest challenges for supplier brands.
“AI really demands real-time,” she said. “It’s no longer a differentiator. Retailers are using AI to uncover signals in real time, such as changes in speed, decreases in fill rate, price discrepancies, and changes in demand. That information is now right in front of the buyer’s eyes.”
“I don’t think brands are changing fast enough. Most brands are still finding that information weeks later. AI has significantly shortened that feedback loop from weeks to hours. This is both an opportunity and a warning for brands.”
Yarbrough said supplier brands are becoming more connected to AI technology and will continue to be visible and relevant. He said brands that can’t keep up with this situation will lose priority before the phones are manufactured. He said supply chain disruptions are becoming the norm and AI tools will be needed to better respond and adapt.
“Walmart has spoken publicly about how they are using AI to reroute logistics and disrupt ports in real time,” she said. “The brands that will benefit are the ones whose operational context is clear and accurate, and who can take advantage of those signals and act on them, rather than scrambling to update a spreadsheet while the system is essentially rerouting it.”
The future of agents
The Massachusetts Institute of Technology defines agent AI as an autonomous system that uses machine learning models to pursue complex goals with limited human supervision. This moves AI from being reactive and reactive to proactive and independent action. For example, agent AI bots can perform tasks such as booking travel or making retail purchases for consumers.
Rob Gandes, chief technology officer at Vendormint, said agent AI won’t work in the same way as the summarized, superficial data that humans need to make decisions.
“We need more depth,” Gandes said. “We need more detailed information. We need raw data to make more informed decisions. It improves over time and is predictive in nature. Agenttic AI can improve processes rather than simply mirroring them as they were done in traditional systems.”
He said businesses will struggle because their data isn’t ready, it’s not sanitized and it’s not clean. He said agent AI cannot be applied to legacy systems because business processes and workflows would need to be rewritten to support AI. He said suppliers that cannot implement agent-based AI systems risk being left behind.
Howerton said consumers are finding products through conversational commerce powered by AI shopping agents.
“Conversational commerce extracts more truth from discoveries and asks deeper questions,” he said. “We don’t get direct links like we used to. We don’t have advertising right now. Shoppers are finding this a very useful tool, and I think it’s building trust right away. Personalization will follow.”
Howerton said AI-powered conversational commerce ignores specific brands and ads, finds products with high ratings and criteria attributes based on research and recommendations, and returns the best solution based on the consumer’s needs. He said brands that want to appear in conversational commerce need detailed content, lots of images, and detailed specifications.
Yarbrough said the conversations happening now at the supplier and retailer table are being driven by the demands and outcomes of AI.
“This is a big change, and AI makes it very difficult to hide that distinction,” she says. “In other words, we are moving from a world where brands choose where they appear to a world where AI decides who to recommend based on relevance, credibility, and performance history.”
Yarbrough said retail is still a relationship business, but brands need to recognize that retail media networks are allocating visibility based on performance signals, not just advertising.
Gandes said many consumers will at some point fully delegate their purchasing decisions to AI agents.
“At that point, all marketing as we know it changes because you’re no longer trying to influence consumer decisions,” he said. “You’re now trying to influence the algorithm’s decisions.”
Editor’s note: Supply side section Talk Business & Politics focuses on the companies, organizations, issues, and individuals involved in providing products and services to retailers. The supply side is managed by Talk Business & Politics.
