How AI is disrupting every stage of the customer journey

The journey from innovative disruptor to outdated dinosaur is short and fast. Just ask Frederick Tudor. 

In the early 1800s, Frederick Tudor had the idea to ship and sell ice extracted from his family’s New England farm pond to the Caribbean by boat. “About half the ice melts on the way from New England to the Bahamas, but some of that ice did arrive in the Bahamas,” said Publicis Groupe Chief Commerce Strategy Officer Jason Goldberg during a keynote session at NRF 2026: Retail’s Big Show. 

“Frederick Tudor gave free ice to every bar in the Bahamas and said, ‘Serve a cold drink to your customers and I’ll bet you they come back for more, and you can buy that ice from the Tudor Ice Company.’” 

The Tudor Ice Company was a huge success, and by 1840, Tudor was worth millions. His idea spawned a whole industry of ice harvesters from New York to Norway. But what Tudor and his cohorts didn’t count on was future disruptors, including the invention of artificial ice made in factories, and, eventually, home ice makers in every refrigerator door. 

“That is sort of a cautionary tale for the incumbents in this room,” Goldberg said, “about what could happen because of digital commerce.” 

Today’s major retail disruptor is clearly artificial intelligence — specifically agentic commerce, which goes far beyond AI-created efficiencies. 

“I’m a lot more interested in what I think is a much bigger disruptive change, which are entirely new consumer behaviors because of AI,” Goldberg said. “I’m talking about people that discover products and buy them in new ways that they didn’t a year ago.” 

Just as shopper marketing and discovery in-store was replaced by social media marketing and viral videos, AI tools such as ChatGPT, Perplexity, Copilot and Gemini are uprooting social media, search engine ads and even retailers’ own websites. 

Agentic commerce allows consumers to tap into all the research of AI, along with tailored product recommendations, and the ability to “even buy products, and you never have to pick a retailer. You never have to go to Amazon. You never have to go to Walmart. So, the onsite agents may well lose some or all of the traffic to the offsite agents,” Goldberg said. 

In fact, with 52 million prompts a day, ChatGPT is already “the ninth largest ecommerce site in the world, without even having a checkout engine,” he said. 

Jason ‘Retailgeek’ Goldberg speaks at NRF 2026: Retail's Big Show.

“They’re getting more traffic than Apple. They’re getting more traffic than Shopify. They were a top 10 ecommerce site in the world without commerce, just based on 1.9% of their prompts being commerce-related.” 

To win with the “robots,” as Goldberg called them, “we have to understand what’s important to the robots, which is going to change every day. So, we need tools and we need business processes to understand what the robots care about.” 

This shift has given rise to a whole industry, including generative engine optimization or answer engine optimization. “Those are important, but what you actually need is something more strategic. You actually need agentic commerce optimization,” he said. 

“You’re not just optimizing for the robots. You have to optimize for your product catalog. You have to optimize for the business outcomes. You have to optimize for the customer expectations.” 

Retailers will also need a change strategy to help their organization and employees think differently about new innovations and technologies. 

“While the tech is important, the strategy is important,” Goldberg said. “We have to think about how we’re going to take all those human beings, all those coworkers in our organization, and get them comfortable with this disruption so that we don’t end up being the Frederick Tudor ice harvesters.”