From left: Meera Bhatia, COO of Fabletics speaks with Christopher Thomas-Moore, chief digital officer at Domino’s Pizza LLC; Brian Seewald, executive vice president, digital, with Victoria’s Secret & Co. and Ananda “Andy” Chakravarty, vice president, research, IDC Retail Insights, IDC at NRF 2024.
Generative AI in theory is one thing. The reality can still be something different altogether — even in its ability to reshape customer relationships.
At NRF Nexus, Ananda “Andy” Chakravarty, vice president, research, IDC Retail Insights, IDC, sat down with a trio of leaders facing such challenges on a daily basis. The panel included Meera Bhatia, COO of Fabletics; Christopher Thomas-Moore, chief digital officer at Domino’s Pizza LLC; and Brian Seewald, executive vice president, digital, with Victoria’s Secret & Co.
Chakravarty kicked off by taking the audience back to GenAI’s roots and its exceptionally rapid growth. In November 2022, more than 1 million people began using generative AI within days of the launch of ChatGPT. In January 2023, more than 100 million people had used it.
Take look at the highlights from NRF Nexus at the Terranea Resort in California.
It has been the subject of hype from the start — and Chakravarty said he remembers writing reports on artificial intelligence in ninth grade. “Some of the things I was writing in those days are actually coming true,” he said.
As for the rest? It’s still “early stages,” according to the panel, and this is a time for experimentation, learning and finding value.
“I absolutely am bullish on it being the future of how we can improve operationally, and also better delight customers in the end,” Bhatia said. Thomas-Moore spoke of the “promise of a better future” with generative AI in terms of personalizing experiences and being relevant to what’s being provided to customers. “But it’s an evolution,” he said.
The panelists pondered a number of topics: how masked or forward the use of generative AI is (or should be) to consumers; the introduction of GenAI tools for operational efficiencies; legal concerns raised; privacy issues; current use cases under experimentation; what might be possible as technology improves; and what can still be done with traditional data.
For the operations aspect in particular, Thomas-Moore said, there’s a lot of potential — specifically in “how you pull that cognitive load out of the store, out of our team members, so that they can focus on the things that really require their attention, versus the things that now use their attention because we don’t have the technology for it.”
Seewald, meanwhile, talked about Victoria’s Secret’s “very scrappy” data science team, and how they were incorporating ChatGPT to provide direct answers to simple questions on store operations. It was something they were just trying out, he said, “but it seems super cool and useful to me.”
Victoria’s Secret is also in beta on a home try-on program. Consumers can get a curated box every month or every quarter to help build out a wardrobe. Here, GenAI is useful in incorporating feedback to improve curation and increase keep rates. In essence, it can help figure out what should go into the box. “It is going to help us make that program better,” Seewald said.
The group also offered a bit of advice and caution. In terms of advice, Thomas-Moore recommended starting with learning and testing. “It’s going to be different for every company,” he said. It might even be different for every end user, and he doesn’t know how the desired end result will be. “But I will say that the full spectrum of AI is going to be leveraged in different ways to bring about a better experience.”
"The full spectrum of AI is going to be leveraged in different ways to bring about a better experience."
As for that caution, the panelists noted how leveraging AI can open up a variety of considerations. There must, for example, be the right kind of teams in place to understand the technology and to manage and drive innovation in the generative AI space.
Seewald told a story of using ChatGPT to write a job description. “What I learned was that there’s an inherent bias sometimes in the answers that it gives,” he said. “I was like, well, that’s OK, because I then edited it and put a Victoria’s Secret spin on it. But if somebody didn’t know that, and they hadn’t been properly trained, what do you do with that?”
Bhatia also advised not putting company secrets into ChatGPT, as well as having a security team “review anything before you put it out there.”
In all, the panelists said, it’s important to consider operational alignment, to pay attention to change management and to get that testing underway — even as the technology continues to mature.
Related content
