Uncovering AI’s role in retail’s next frontier

Professor Saravanan Kesavan worked in artificial intelligence (AI) before most people knew its name.

Over 20 years ago, he was a consultant implementing supply chain software for businesses. The first time he implemented planning software at a large semiconductor firm, four days went by before it yielded a solution to the problem he was addressing.

“It wasn’t called AI at all,” says Kesavan, professor of operations and Sarah Graham Kenan Scholar at UNC Kenan-Flagler Business School. “We used to call it decision-support tools.”

Soon, the solution took eight hours. Today, what once took eight hours can now be done in a few seconds.

Now algorithms are extremely sophisticated. Microprocessors are designed specifically to speed up machine learning algorithms. Big data just keeps getting bigger, and AI is intertwined in the fundamental fabric of not just Amazon and Google but big-box department stores and in an increasing number of organizations in different industries.

But while the number of businesses using AI technology has dramatically increased, one thing remains the same as it was at the beginning of Kesavan’s career.

While companies understand the importance of AI in optimizing performance and they’re investing heavily in the technology, many managers still struggle with how best to use data analytics and business solutions that come from machine learning.

An important question remains: How should companies actually work with AI?

“I’ve worked with large companies that continue to grapple with how to use AI effectively to make operational decisions,” says Kesavan. “They’re trying to figure out how to go about implementing it, how to train a workforce in AI and how to hire the right people to help with AI. The biggest challenge, though, might be the lack of awareness on how to restructure the processes to capitalize on the power of AI.”

With his latest initiative, Kesavan hopes to give companies answers.

Kesavan has launched the Rethinc. AI Management Lab at UNC Kenan-Flagler, housed within Rethinc. Labs at the Frank Hawkins Kenan Institute of Private Enterprise, thanks to a three-year, $150,000 gift from founding partner Advance Auto Parts.

The lab’s focus is research at the cross-section of AI and operations management, studying the most effective ways for a wide range of industries to implement AI more thoughtfully. It will explore AI’s role in such fundamental issues as sales forecasting, merchandise planning and employee scheduling.

If graduates enter the healthcare field, there will be an AI system that’s making decisions on doctor scheduling and patient services. If they enter retail, they will encounter AI systems that look at how much inventory to order and suggest pricing.

Companies rely on the next generation of business students to help them understand the full potential of AI and find new ways to implement it successfully.

“The lab is going to help us conduct cutting-edge work on problems many companies share,” says Kesavan. “AI is absolutely critical, but managers have been naturally skeptical because they tend to make decisions on intuition.

“For them, it has long been about the art, rather than the science, of decision making. AI shifts it to the far end of the science of decision-making.”

Retail therapy

The AI Management Lab brings Kesavan’s career full-circle. He has had a passion for retail from the time he grew up in India where his family ran a successful bookstore chain that was profiled in The New York Times. That shaped Kesavan’s interest in nearly every aspect of business operations.

“I had experience in just about every part of the store. I would act as cashier, load products from distributors and handled customer service by the time I was in middle school,” he says. “My favorite part of running the store was talking to customers, understanding their needs and serving them. I got a lot of satisfaction from knowing that I was able to fulfill the needs of our customers.”

Kesavan earned bachelor’s and master’s degrees in engineering from the Indian Institute of Technology, Madras, and the University of Massachusetts at Amherst, respectively, before receiving a doctorate in technology and operations management from Harvard Business School.

He then launched his career as a consultant with i2 Technologies, helping big-name business clients such as the Texas Instruments, UPS and Nike solve supply chain management issues.

In 2007, he joined the faculty of UNC Kenan-Flagler, which was high on his list of places he’d like to work after earning his PhD. The research environment at UNC made the move a no-brainer.

Since then he has was named assistant dean of the Full-Time MBA Program received numerous honors for his retail operations and digital operations courses for MBA students, including the MBA Teaching All Star award 12 times and a Weatherspoon Award for Teaching Excellence in the Full-Time MBA Program in 2018.

He also has been a relentless researcher, studying labor productivity, worker well-being and ways big data can improve forecasting. His work has had immediate implications for retailers.

A study he coauthored  on how responsible scheduling practices can improve retail store performance at the Gap was cited as part of a 2022 bill introduced in the U.S. Congress. A study of global supply chain issues used 28 years of sales data from 183 publicly traded U.S. retailers to show how companies can better deal with shifts in demand.

Too many customers on a busy day in a self-service store can actually hurt sales, as Kesavan learned in a study with Vinayak Deshpande, his operations colleague, and then-student Hyon Seok Lee (PhD ’17).

Kesavan even asked his daughters to conduct retail research of their own. When they spent a year studying in India, he asked them to pay attention to cultural nuances related to retail customer behavior. His older daughter observed customers standing almost on top of each other when in line at a store to prevent other shoppers from cutting the line.

Much of Kesavan’s research ends up in his coursework. What he and his colleagues find through the new AI lab will, too.

“Our students will play an important role in shaping how AI is going to be adopted in organizations around the world,” says Kesavan. “They will be the ones leading the path of AI forward.”