The Next Retail Revolution?

What if grocery stores set prices the way airlines and hotels do?

Based on the research of Ioannis Stamatopoulis

Anyone who has shopped for an airplane seat or a hotel room has experienced dynamic pricing. Prices fluctuate often, even daily, depending on availability and how soon the shopper needs to book. Consumer demand drives these fluctuations.

Supermarkets, by comparison, are behind the times. They take months to change prices storewide because of printing new labels and putting them on shelves. But new technologies could bring dynamic pricing to grocery aisles, says Ioannis “Yannis” Stamatopoulos, an associate professor of information, risk, and operations management at McCombs since 2016. In several recent studies, he finds the ripple effects of two technologies — electronic shelf labels and expanded bar codes — can benefit not only stores, but also consumers and suppliers.

Stamatopoulos’ field is operations management: the study of maximizing efficiency in business operations. It’s an interest that began years ago when he took part in math contests. “I found that I enjoyed solving practical problems, as opposed to talking about abstract objects,” he says.

He’s particularly interested in obstacles to efficiency, which he calls operational frictions. His research has examined ways to lower these obstacles in cleaning services, road salt procurement, and electric scooters.

Stamatopoulos, who has a doctorate in operations management and an M.A. in economics from Northwestern University, and a B.S. in mathematics from the University of Athens, describes ways to reduce frictions in setting supermarket prices. Modest operational changes could lead to multiple benefits, he says: cutting food waste, stabilizing supply chains, and discounting some prices for consumers. This interview has been edited for length and clarity.

How did you become interested in grocery store prices as an operational problem?

I was taking a macroeconomics class, and we talked about why prices don’t move more often. I thought to myself: “This is an operations question, because if there are frictions, they must lie in the micro processes of pricing itself. This micro problem has macro implications.”

How did you go about finding a solution?

I connected with a chain in the U.K. that installed electronic shelf labels. They’re small screens that display a price and can also display a discount. You can have one price for fresh produce and a different price for produce nearing expiration. If you want to change your prices, you can execute the process from a tablet.

This chain tried it in a few stores first to see whether it was worth it. I offered to do a statistical analysis, comparing the stores that received electronic shelf labels with the stores that did not.

What were the results from introducing digital labels?

The chain went from changing prices every month and a half to every three weeks. And by the way, this came with increases in revenue. But I was disappointed that they didn’t go all the way to changing prices daily. I asked them why this didn’t fully resolve the issue. One person told me, “The №1 reason we change our prices is to respond to inventories. If we have two months’ worth of something in inventory, we put it on a discount.”

What light bulb went off when you heard that?

Inventory records are garbage. There’s research demonstrating that they are wrong 65% of the time about what’s really on the shelves. So, when I heard him speak, it was a light bulb moment. Bad information about inventories is an information friction. Better information might speed up price changes from every month to every day.

But I needed some empirical proof of concept. In collaboration with two other professors, we found a chain that adopted two changes at the same time. They adopted electronic labels, and they adopted expanded bar codes to improve their inventory records.

Can you explain expanded bar codes and how they help with changing prices?

A typical UPC bar code only captures the kind of product. It tells you, “This is a pack of Tyson chicken thighs.” An expanded bar code captures what day it was packed on, at what facility. You have an ID for every package, as a unique item that you can follow through the chain.

With expanded bar codes, they had very detailed and much more accurate inventory records, including expiration dates. They were able to do dynamic pricing. If there was more of a product about to expire, they gave it a deeper discount. Or if they just got a new batch, and they had lots on the shelf, they might lower the price. They could make daily changes.

So, what impacts did dynamic pricing have on those stores?

They used to waste 20% to 30% of products. They went down to wasting almost nothing, and it was all through the smart discounts. Because they were able to move more quantity, their revenues and profits went up.

How did shoppers react to these fluctuating prices?

They did surveys, and the pricing seemed fair to their consumers. If a product was about to expire, they thought it was fair to lower the price. If the consumer wanted a fresher product, they thought it was OK to pay a little premium.

You found that suppliers as well as stores could benefit from dynamic pricing. How would that work?

With a colleague, I got data from a large supermarket chain in China and their distribution centers. I could observe all the product flows and prices. We put that empirical data into a model to simulate what would happen if there was no cost to changing prices.

We found that with dynamic pricing, when stores waste less, they can order more often. The amounts are more consistent, rather than seesawing up and down. It counteracts the bullwhip effect, that volatility increases as you go upstream in the supply chain. Instead, the whole supply chain is more efficient. Suppliers can plan better. They’ll know they won’t need an extra truck and that their labor level is going to suffice.

If these technologies have such big payoffs, why aren’t most U.S. grocers adopting them?

Their biggest problem is the high fixed costs of adoption. A lot of innovation in dynamic pricing is coming from Europe, because their labor costs are higher, and they can be harsher about food waste. But it’s going to come to the States. It’s just a question of when and how. If other competitors start adopting these technologies, you don’t want to be late to this party. I hope that my research will show them real data about the benefits.

What can be done to speed up adoption?

If we really want to combat food waste, it might make sense for policymakers to provide incentives to supermarkets to adopt some of these technologies. If they get help doing this, such as tax credits, they’re going to order more and waste less. This will benefit all of us.

Story by Steve Brooks