Is Generative AI Really Going to Change the Retail Experience as Some Experts Claim?

Make no mistake: AI has been embedded in retail operations for a while now helping with everything from demand planning to product recommendations. But until recently, AI’s role in retail hasn’t been the talk of the town. Now that generative AI has captured consumers’ attention, everyone wants to know if this particular flavor of AI will change the world – or at least the shopping experience. 

Well, if consumers have it their way, store associates will be better equipped to help them in the coming months. Three-quarters of consumers who participated in Zebra’s 16th Annual Global Shopper Study said they have a better shopping experience when associates are using the latest technology to help them. To be fair, associates agree. Nearly eight in 10 store associates currently feel like shoppers know more about what’s going on in their stores than they do!

Perhaps that’s why most hope retailers will invest in AI. Seventy-five percent of associates believe they can provide a better customer experience if they were able to use generative AI, or “gen AI.” Slightly more – 78% – feel the same way about any AI and machine learning technology tools.

Will everyone’s wish come true?

I think so. 

Stuart Hubbard, the Senior Director of Artificial Intelligence and Advanced Development here at Zebra, agrees. He recently gave me the scoop on some very cool projects the Zebra Chief Technology Office (CTO) team is working on to improve the shopper experience, increase associates’ customer readiness, and help retail operations become more manageable and profitable. And his timing couldn’t have been more perfect, as three days later, I received the 16th Annual Global Shopper Study readout. 

To my surprise, everything the team is working on is exactly what shoppers, associates and even retail decision-makers are saying they want and need right now. This includes smart carts for shoppers, AI-based task guidance for associates, and a more detailed record of the customer buying journey to improve retailers’ inventory-related intelligence. (Wouldn’t it be great if every retailer knew – with complete confidence – what is in or out of stock, what customers do and don’t want, and more?!)

That’s why I sat down with Stuart again post-study readout to talk about how exactly gen AI is going to be put to work in these cases where shoppers and associates say greater technology use is needed. I captured the 20-minute conversation and want to share it with you, as I learned quite a lot from Stuart: 

I also participated in a roundtable discussion a few months back with three great thought leaders in the retail and consumer packaged goods (CPG) sectors:

• Avanade’s Rasmus Hyltegard, Global Data and AI Lead for Retail and Consumer Goods

• Microsoft’s Global Retail Strategy Lead, Michele Fisher

• Wayfair’s John Wall, Head of Studio, Creative Operations Management and Video Production 

We talked about what gen AI may be able to do for you and what it probably can’t (because there’s another, better-suited, type of AI), and noted where gen AI is more hype than help, at least for now. You can watch the full roundtable discussion here, but there’s one thing I want to call out because it’s important we all take this to heart:

Generative AI is not going to become the only AI that we turn to help solve every problem, make every retail experience possible, and ensure every aspect of CPG production and order fulfillment works to plan.

It is certainly worth asking whether or not gen AI is capable of supporting certain decisions, experiences or ambitions in retail or CPG environments. However, there are many different types of AI that you will ultimately need to bring in to play a part in the modern store, factory or warehouse. This includes the types of AI that made it possible for companies like Bimbo to keep your favorite baked goods on shelves during the pandemic and other challenging periods or for Walgreens to better anticipate and respond to customers’ needs.

So, when you start asking around about how others are using AI – or about how you could be using AI – to help optimize inventory, engage associates and elevate the customer experience, make sure you understand what type of AI is being used or recommended. More importantly, understand why that AI model is being used or recommended and then ask if there are other options.  

Global industries, and even technologists, are still very much in an exploratory stage with gen AI, and like any new technology, the “potential” of a new technology can overshadow its practicality. The incorrect assumption that all AIs are created equal can also lead to disappointment when the capabilities of different AI models aren’t properly distinguished and leveraged. All shoes serve the same purpose: to protect your feet. However, you wouldn’t wear dress shoes to hike up a mountain. (I know that well.) So, be inquisitive and intentional. 

I believe gen AI will live up to the current hype if (and only if) you and I work together with other AI and retail/CPG experts, including shoppers and associates, to understand which “shoe” fits best. Once we know what they want to do, then it will be easier for us to make the right changes and take the right actions. For example, let’s say they want to be able to walk into a store, ask an associate where their favorite brand of crackers are located, actually find the product on the shelf where the associate said it was, then be able to check out/pay without having to go through a lane. We can then say, “You need to leverage gen AI for XYZ and adaptive AI for XYZ alongside these types of vision, mobile computing and payment technologies.” 

These different types of AI can work together seamlessly to deliver the information you and your associates need to deliver the shopping experiences consumers want. For example, the vision AI can detect when something goes out of stock on the shelf but is available in the backroom. Then a voice assistant (which uses gen AI) can create a task and tell a specific associate about it. The associate can then grab the item from the back and get it out on the floor so that you don’t miss a sale. Or if the item is completely out of stock, then these signals can go direct to the ordering system to prompt action. 

Either way, vision AI system signals could be fed automatically into the inventory/ordering/planning system so you always have an accurate view of what’s in stock or not, when things are moving, etc. This data, along with other data the AI can pull in (i.e., is there a football game on?), helps planners and managers act on more accurate forecasting predictions based on time of day and local events.    

Plus, now that we have demonstrated gen AI can work on Zebra devices powered by Qualcomm, we can start to look at how else gen AI could be used to guide both associates and shoppers as they pick items, check prices, and more. You can read more about how the on-device gen AI works here. You may also want to watch this short video in which Zebra CTO Tom Bianculli also talks about how on-device gen AI can help you reduce costs, increase privacy and provide AI assistance in environments where internet is limited or not available: