New Study Finds Uber Eats Reduces Uber Rideshare Volumes

Zhou and her co-authors, Michigan Ross PhD graduates Hyuck David Chung and Christine Choi, analyzed data from New York City’s rideshare and food delivery markets in 2015 and 2016, with additional tests in 2019. They found that a 1% increase in local restaurants joining Uber Eats led to 2% fewer Uber trips and nearly 7% fewer Lyft trips.

The drop in rideshare volumes is largely due to drivers switching between two services to maximize their earnings, the researchers say. Uber’s expansion into the food delivery business allows drivers to use their idle time for food delivery.

When a significant number of drivers flock to the new business, the rider who requests the trip now has to wait for a longer time to be connected with the next available driver. When such experiences accumulate, the rider might decide to leave the platform, causing other riders to leave (due to direct network effects) and some drivers to abandon the platform (due to indirect network effects). As a result, the rideshare volumes dropped following the launch of Uber Eats.

“Given the network effects and the ‘winner-take-all’ nature of platform business, the loss of riders and drivers might be fatal to the platform,” Zhou said.

Lyft also suffered from Uber Eats’ impact. Before Uber launched the food delivery business, Lyft drivers interested in food delivery had to juggle multiple apps, such as Lyft and DoorDash, to avoid receiving requests from a platform while completing transactions on another.

Declining requests from a platform may lead to a lower performance rate and fewer orders assigned from the platform in the future. Oftentimes, drivers had to use multiple smartphones to manage scheduling burdens across different apps. The coordination costs of managing multiple platforms discouraged some rideshare drivers from working in the food delivery business.

With Uber’s diversification into the food delivery business, Uber drivers were batched with either rideshare trips or food delivery orders when they turned on the Uber app. Drivers could choose which business to work on by modifying the settings on the Uber app instead of forgoing unwanted orders.

This not only lowers the coordination costs across different Uber apps but also minimizes Uber drivers’ idle time, the researchers say. As a result, Lyft drivers may switch from working on Lyft and DoorDash to working on Uber and Uber Eats, leading to the drop of Lyft’s rideshare volume.

“Our research highlights the hidden cost of Uber’s diversification,” Zhou said.

Annually, the launch of Uber Eats reduced Uber’s potential trip volume in Manhattan by about 3.3 million trips. With an average fare of $15 per trip and Uber collecting 20%, this translates to a $10 million annual revenue loss in Manhattan. For Lyft, the impact was a 1.3 million trip reduction and a $4 million revenue loss.

To reduce cannibalization, platform businesses can increase compensation for existing services or encourage within-platform multihoming, the researchers say. For instance, Uber could increase rideshare fees for the drivers and lower prices for the riders.

Additionally, rewarding drivers who complete more rideshare trips or food delivery orders might attract new drivers or divert drivers from competitors like Lyft. Zhou also suggests lowering the coordination costs of operating across different platforms within the same platform firm to attract drivers from rival firms.

Originally published on Michigan News.