How artificial intelligence will change retail

Artificial intelligence has the power to change retail — intelligently. And for those who might be considering the worst-case scenario, the reality is going to be far different.

“People have all these preconceived notions of what AI is like, and it’s going to be nothing like what people think. Artificial intelligence programs are assisting and making people’s lives better,” says Ananda “Andy” Chakravarty, vice president of research at IDC Retail Insights.

It’s undeniable that AI is already reshaping retail. Levi Strauss & Co. recently announced it would be working with customized AI-generated models. American Eagle said it’s using AI for inventory tracking, and Puma is using AI forpersonalized customer styling.

Chakravarty admits that nobody can predict what artificial intelligence might be like in, say, 20 years. He can’t guarantee AI won’t sometimes overtake humans. But for now, “it’s not ready for independent usage yet. We don’t see very many AI technologies that don’t have humans involved. You have to have that oversight and ‘human in the loop,’ if you will.”

In any case, here’s where AI seems most likely to help revolutionize the retail industry.

Targeting customers

Levi’s plans on utilizing AI to help with its models in an effort to offer more diverse body types. When it first announced the plans, the company received some blowback from critics and quickly put up a post, clarifying that “we are not scaling back our plans for live photo shoots, the use of live models, or our commitment to working with diverse models.”

Levi Strauss & Co. told NRF that, later in the year, the company will experiment with a “small, controlled” project on using AI-generated models “with the objective of learning more and determining whether we can enhance the overall consumer experience.”

Lev's store front in NYC.

Currently, Levi’s has just one model for each product on or on its app. “It is physically impossible to shoot every single product at every single size at the SKU level,” the company says, “so this technology can potentially assist us by creating supplemental images and unlocking a future where we can offer a more personal, relevant and potentially more engaging shopping experience.”

The AI technology could help greatly with the bottom line, if customers feel more confident about what they’re buying.

It could also help the company if customers decide not to buy a pair of jeans online.

“It is especially hard to replicate the size and fit confidence a consumer can get from trying on a garment in a fitting room in a digital environment,” the company says. “Online apparel returns by some estimates exceed 25%, with size and fit being the number one reason for returns.”

More efficient business operations

Deborah von Kutzleben is the chief marketing officer at Tropical Smoothie Cafe, which currently has 1,235 locations in 44 states. From a scale of 1 to 10, with 10 being really excited about AI, Von Kutzleben says executives at Tropical Smoothie Café are “a solid 8.”

She adds, “Our team knows there is still a lot to learn about the capabilities — and concerns — of AI, and we are listening closely to conversations within the fast-casual industry and beyond to understand its full power.”

Tropical Smoothie store front.

Still, von Kutzleben says the executives can imagine a day not too far out in the future in which the restaurant uses AI software to collect broad data that could help run day-to-day operations.

“Key insights such as predicting order traffic, staffing, assisting with more accurate product ordering and menu pricing suggestions are all possible when AI works in tandem with an in-store POS system,” she says.

Von Kutzleben also says Tropical Smoothie Cafe is watching to see how AI impacts direct customer communication across the fast-casual and quick-serve restaurant space.

“With the capability to respond to online customer reviews or solve customer concerns via text or chat platform, AI could help with customer retention and reputation management,” she says.

Optimized pricing, forecasting and marketing

“Retailers today, 60% to 70% of them don’t know where all the things are located in their supply chain,” Chakravarty says. “They don’t have the ability to track things from one store to another. There are still a ton of huge, immense problems that still exist in even the most advanced corporations.”

He envisions a retailer with 100,000 SKUs, and an AI program keeping track of all of that, changing prices whenever it makes sense and ordering inventory based on its demand forecasting.

Explore practical applications of AI in retail from industry leaders at NRF Nexus, July 10-12, 2023.

“This is where you’re going to really see AI having value,” he says.

Chakravarty says certain elements of dynamic pricing can be enabled today, but it’s generally a niche market.

“There are execution limitations in the store,” he says. “Most retailers don’t have electronic labeling, so printing and tagging items will be the challenge every time a price change is made.” If AI is doing that for the retailer, dynamic pricing really starts to live up to its billing.

Additionally, “Marketing is another part where AI could really transform retail,” Chakravarty says. “Imagine a computer system that can pull the right copy assets and images and create its own email campaign that it can send to millions of customers based on the customer’s own engagement, behavior and purchases for a true one-to-one relationship with the retailer.”

What needs to happen first

Simply put, time needs to pass. Companies are in a wait-and-see mode, at least judging from von Kutzleben’s perspective.

“With how rapidly AI technology is changing and how little we know about how these integrations will affect businesses over time, we are waiting for the right opportunity to integrate AI that will benefit both our franchisees and our guests,” von Kutzleben says.

“Both the limitations and potential applications of AI can create hesitancy. Until those are addressed, we will continue to explore the technology but may wait to implement it. However, we do think Tropical Smoothie Cafe will be a fast follower in terms of implementation.”

There are practical reasons to wait, von Kutzleben says. AI technology can be expensive to implement, and right now the ROI is unproven.

“AI is constantly evolving so we need to be measured in how we move to make the best choices for our brand and our franchisees,” she says.

While AI might transform the world of retail and restaurants, the metamorphosis is not going to happen too fast.

For instance, the marketing campaign that Chakravarty envisioned is certainly plausible but not very practical right now.

“All of that requires a lot of infrastructure and the right tools to pull off something like that,” he says. “It’s not routine or simple. It may be a few years from now, but it’s coming.”

Christian Beckner, vice president of retail technology and cybersecurity at NRF says being ready for AI and developing best practices for managing it is why NRF recently started its AI Working Group.

“We’ve had discussions about AI for years in groups, but this is the first group dedicated to AI issues,” Beckner says.

He says over 70 NRF member companies were at the first meeting, many sending senior executives, and that there has been an unusual amount of interest in artificial intelligence compared with other times issue-oriented groups have been started.

Beckner says the AI Working Group, and NRF in general, will be working on discussing best practices with AI and advising legislators on policies. He does point out that other sectors, such as the medical field or law enforcement, have far more knotty problems to figure out.

AI advising how to care for a patient is far different than AI suggesting how to fight fraud or strengthen inventory decisions. Still, artificial intelligence will undoubtedly change retail. The question is how.

“We want to be a voice in the AI debate, to both information policymakers on where guardrails should be and to help ensure AI doesn’t get used in ways that aren’t appropriate,” Beckner says.