The relationship, though, seems to have soured of late, as people slowly resume their eating-out habits.

One reason is an inherent flaw in the revenue-sharing contracts between the restaurants and delivery services, said Professor Robert Swinney of Duke University’s Fuqua School of Business.

Under the predominant revenue-sharing contract, the delivery service that receives an order generally keeps 15-30% of the revenue and returns the rest to the restaurant. This percentage cut not only makes a dent in already low restaurant margins, it also causes the delivery service to make decisions in a way that fails to account for the negative impact that delivery orders have on the dine-in experience, said Swinney.

“Delivery orders placed on the platform generate congestion in the kitchen,” Swinney said. “This makes service worse for dine-in customers, potentially leading to restaurants losing dine-in revenue.”

Swinney, an associate professor in Operations Management, and colleagues-- Pnina Feldman of Boston University and Andrew E. Frazelle (PhD, ‘18) of the University of Texas at Dallas--research coordination problems in supply chains, and they suspected similar coordination problems were happening here. “In a supply chain, where one firm makes a product and another firm sells that product, because certain costs are incurred only by one firm or the other, firms can make uncoordinated decisions, taking actions that may be in their own best interest, but that decrease the profit of the whole supply chain. Delivery platforms are a type of supply chain, with the restaurant making the food and the platform selling and delivering that food to customers,” Swinney said.

Under the current contract between apps and restaurants, the “externality” cost (the deterioration of service for dine-in customers generated by delivery orders) is felt by restaurants but doesn’t affect the food-delivery services at all. Because platforms don’t feel this cost, they set prices that are too low, generating too much congestion in the kitchen. This decreases the restaurant’s dine-in revenues, and as a result, the combined profit of the platform and the restaurant.

The simple fix, described in an online paper, involves the delivery service paying a fixed fee to restaurants for each delivery order, in addition to the revenue-sharing split.

This fixed fee, Swinney explained, “charges the negative externality back to the platform,” which will lead it “to set its prices appropriately.” The current revenue-sharing model does work when most customers are concentrated on either the delivery or the dine-in side, Swinney added. It worked well during COVID because almost all the customers were ordering for delivery. Similarly, it worked well years ago, when the delivery services first entered the market, because most customers were still dining in restaurants.

“The problem is when you have a closer mixture between the two,” Swinney said. That’s when the delivery and dine-in channels have a large impact on one another, and the need for coordination between the platform and the restaurant is greatest.

Food-delivery apps won’t go away, Swinney said, because by now people are used to them.

Swinney decided to study this issue after reading about restaurants complaining about the platforms’ high commissions.

“Our message with this paper is that the problem isn’t merely that platform commissions are too high,” Swinney said. “In fact, we show in our research that commission caps don’t fix the problem, because even with a cap, platforms still don’t account for the externality costs they generate for the restaurant, leading to miscoordination.”

“We hope to influence platforms to consider a different payment structure by showing that this new contract maximizes the combined profit of platforms and restaurants,” he said. “This maximizes the size of the pie. And when you maximize the size of the pie, you can adjust the terms of the contract to cut the pieces in different ways and make everybody happy.”