AI suffused the entire Shoptalk Spring event last week in Las Vegas, from an exhibit floor that was flooded with AI-powered tech, solutions and services to the scores of sessions exploring every nook and cranny of artificial intelligence in retail.
Across all the information (and opinions) expressed at the show, two broad themes stand out:
- Agentic commerce is already in operation and rapidly changing how retailers and brands go to market; and
- The overall pace of change and growth of AI are much faster than they’ve been with previous technology “earthquakes,” such as the introduction of mobile commerce and the development of social networks as shopping venues.
Here are some additional “hot takes” on the state of AI today from a few of the many experts and retail decision-makers at the event:
Agentic AI Could be a $3 Trillion Opportunity
“In AI, focusing on the customer is the most important thing. Don’t start with a tech-first mindset; it’s about being where customers are shopping. It’s a $3 trillion agentic commerce opportunity across the globe. [The next step is that] every brand needs to understand the value that brand stands for — like lower prices at Walmart or improved assortments at Target. It’s critical that your brand’s agent has a voice that aligns with yours. And finally, it’s about data — product data, metadata — because agents don’t browse like humans; they match, so a strong product and data foundation is essential.”
— Kapil Dabi, America’s Market Lead, Global Industries and Solutions, Retail and Consumer, Google
Customer Intent is the New Search and Discovery Battleground
Agentic AI’s more conversational searches can provide valuable clues to customer intent as well as what a potential purchase would be used for.
“With AI, searches are getting longer — 2X to 3X longer than you would typically see in [traditional] search. [With AI prompts people can] speak like a human, it’s much more conversational. You might say ‘I’m going to Atlanta and want to get a cashmere sweater that will be appropriate for spring there.’ Now, the advertiser can understand intent in a wholly new way, and perhaps match ads not just for sweaters but also for jeans or a purse. AI Mode is an immersive search experience, and [shoppers are] saying so much more with these intent signals — not just what they want but also why they want it, what this moment in their life means.”
— Courtney Rose, VP, Google Ads
“A customer may say ‘I need to shop for camping shoes in this price range,’ but if you don’t have products that fit into those parameters, it won’t show up in an [agentic] search — it will choke the recommendation. You need rich contextualization of the SKU, rich storytelling, Reddit reviews, social media scans — intent is key in discoverability.”
— Neha Kovach, Global Head of CRM, CX and Loyalty, David Yurman
Ongoing Task: Attuning Agents to Customers’ Personal Preferences
“An AI agent should have a history of your past purchases, to help insert contextualization. For example, if I shop all things gold in jewelry and I’m ‘self-gifting,’ the agent shouldn’t recommend silver jewelry. Or if I’m already wearing a brand, whether it’s Balenciaga or Levi’s, the agent might say ‘Maybe these two products go together,’ because it knows your trend and purchase history. Brands will need to contextualize all this within the product, for example by writing richer stories on your website. The question is, how do you keep making the agents smarter? Also, if your AI search is something like ‘Show me potential gifts for my 16-year-old daughter’s graduation,’ that specific experience needs to be carried forward to your site — the Product Landing Pages, the Product Detail Pages, the stories you tell — the future is that agent handoff.”
— Neha Kovach, Global Head of CRM, CX and Loyalty, David Yurman
[In terms of privacy concerns], “if a personalized experience was worth giving [the brand or agent] a bit more information, it’s worth it [to the shopper]. Today’s teenagers expect every digital interaction they’ve had will be remembered the next time they’re on.”
— Kelly Moran, Managing Director, Deloitte Digital
Retailers Must Determine which CX Experiences Still Require the Human Touch
With many customer exchanges, interacting with a human at the other end of a phone line or text chat actually doesn’t provide optimal customer service.
“If I ordered something online and it was supposed to arrive by 5 p.m. and it isn’t here yet, I don’t need a ‘human touch,’ just an answer: will it arrive? Having to wait on hold even 30 seconds to get that answer is disrespectful for any business. Retailers need to determine where they want that high-touch human experience, but for all the others — warranty claims, scheduling deliveries — an AI agent would be better, faster and cheaper and would likely drive up your NPS score, leaving you more to spend on those differentiated high-touch experiences. When clients want their problems solved instantly, that’s the opportunity with agents. It doesn’t remove the human touch; it amplifies the human experience.”
— Bret Taylor, Co-founder and CEO, Sierra, Chairman of the board, OpenAI
AI’s Behind-the-Scenes Impact Continues to Grow
“We’re seeing a lot of AI impact in things that benefit the consumer — even where they don’t know it’s happening. One example is order intelligence; at The Home Depot, with the products we’re delivering, it’s by no means all parcel delivery — we could be delivering lumber or building supplies via a flatbed truck, for example. We can access route intelligence that can see that, of the 20 deliveries that have taken place to this area, the road is too narrow for a flatbed truck, [so we would need to find an alternate route and adjust the delivery time.] We’re using AI to look for anomalies and take action — but all the customer knows is that they got the delivery in time. AI is helping us solve problems and see patterns we haven’t been able to see in the past.”
— Angie Brown, EVP and CIO, The Home Depot
Bonus: Historical Perspective and Words of Comfort
There are often high levels of uncertainty when new technologies come on the scene. Shopify’s Dustin Holmstrom noted that, for example, mobile commerce didn’t really take off until a few years after the introduction of the iPhone in 2008.
“The first thing we did [on our phones] was play solitaire; it took years for consumers to trust that form factor. I believe AI will hockey stick faster, but building customer service and trust is how it will happen.”
— Dustin Holmstrom, Field CTO, Shopify
“None of us are behind on AI. Even if you think you’re ahead, it’s going to change every day.”
— Andrew Laudato, COO, The Vitamin Shoppe
