Market-driven drivers: Dynamic payout ratio means more money, less wait

The rise of on-demand services is great for getting what we want when we want it — but it’s also made us impatient.

And who can blame us? With everything else in the world moving so quickly, why shouldn’t we get groceries delivered to our doorstep as soon as possible or find a last-minute babysitter for the kids for an impromptu night out?

As technology has gotten better, wait-time expectations have gotten shorter. Now, wait time has become a critical factor in determining which service a customer will ultimately use. This especially applies to ride-sharing services such as Uber and Lyft.

“It’s easy to look at the app and see the expected wait time. If it’s going to take a long time for an Uber to arrive, people will either look at another service like Lyft or a traditional taxi service if it means faster service,” says Jiaru Bai, an assistant professor of supply chain management in Binghamton University’s School of Management.

Because ride-sharing drivers can choose when they want to work, and because the demand for rides isn’t consistent throughout the day, there is no consistent supply and demand. If not managed well, this can lead to pesky and inconsistent wait times, which can lead to fewer people using the service, which can lead to fewer drivers opting to work for the service, which can lead to less profit for the company running the service.

So how do you keep all parties happy? Is there an optimal way to keep the service providers, the customers and the company all satisfied?

Bai says yes, there is.

With a mountain of real data to work with from Didi, the largest on-demand ride-hailing service platform in China, Bai and her fellow researchers analyzed rides that took place in both peak and nonpeak times.

Bai found that the optimal solution boils down to allowing the percentage of a transaction the provider gets to be flexibly determined by the market characteristics of a given location. Basically, when demand is high, providers should get paid a higher percentage, and when demand is low, providers should get paid a lower percentage.

“Having a dynamic payout ratio almost always performs better than a fixed payout ratio, according to our model and data analysis, and it leads to benefits for all involved,” Bai says.

Adjusting the payout ratio led to fewer drivers waiting around for riders in nonpeak times (as the incentive to work during those hours is lower) and more drivers on the road to deal with higher demand during peak hours (knowing that the payout will be significantly higher). This led to consistently lower wait times for impatient riders and more satisfaction and usage of the service. Overall, this meant more profit and a better reputation for the company.

Bai says that in addition to utilizing a dynamic payout ratio, platforms should charge more in peak hours, something Uber already does with surge pricing during the busiest times.

“Usually, customers get this perception that they are being taken advantage of when the price surges. But if they notice much-lower wait times because there are more drivers available as a result, the rider will see the benefit in the increased pricing and not feel ripped off,” Bai says.

And for those customers who still don’t see the benefit in paying more during peak hours, Bai says they will be “priced out” and find another way to get to where they want to go, ultimately easing the wait time for everybody else trying to use the service.

Despite the overall benefits, Bai says implementing a dynamic payout ratio is easier said than done.

“When other companies are competing for the same pool of providers, it makes sense to guarantee them as fixed an income as possible. Saying your providers will get 80 percent of every transaction is easier than saying your providers will get anywhere between 10 percent and 90 percent of a transaction depending on the demand,” Bai says.

“Providers will ultimately go with the company they have more confidence in, which is very practical.”

Bai says some companies are currently experimenting with offering different types of incentives to providers working during peak times, which she sees as a step toward eventually deploying a dynamic payout ratio.

“As competition starts to go away in some areas, I think more platforms will be in a position to consider dynamic payout ratios,” Bai says. “It’s difficult to deploy, but eventually can lead to benefits for all.”

Bai’s research on this topic has been accepted into Manufacturing & Service Operations Management and was awarded first place in the Production and Operations Management Society (POMS) 2017 College of Supply Chain Management Student Paper Competition.