Driving Repeat Customers for Food Delivery Apps

CUHK research underlines the role of delivery performance in the success of an on-demand meal delivery service platform

By Nicole Chabot

The spread of COVID-19 has led to a surge in demand for home delivery services – including meal delivery – in many places around the world. However, it’s no secret that, as food delivery apps boom in our already convenience-obsessed world, their revenues remain cut-throat.

In the U.S., market leader DoorDash faces off against rivals the likes of Grubhub and Uber Eats. In China, the food delivery market is dominated by a heated rivalry between Meituan, backed by Tencent, and Alibaba-owned Ele.me. For many of these companies, losing a customer only takes as long as the download time of a rival app.

It is because of this that most food delivery apps remain loss-making. In the fourth quarter of 2019, Uber Eats lost US$461 million and is not expected to be profitable until 2024. Rival Grubhub lost US$33 million in the first quarter, while DoorDash itself lost US$450 million in 2019. In such a ruthless market, how can companies increase their revenues by fostering repeat business? This is what drove a group of researchers to examine how delivery performance affected the revenue growth of on-demand meal delivery platforms.

The research, entitled Faster Deliveries and Smarter Order Assignments for an On-Demand Meal Delivery Platform, utilizing a comprehensive dataset from an online meal delivery platform in Hangzhou, China, was conducted by Liu Ming, Assistant Professor at the School of Management and Economics at The Chinese University of Hong Kong, Shenzhen, in conjunction with Profs. Ying Rong and Huan Zheng at Shanghai Jiao Tong University; Prof. Christopher Tang at the University of California, Los Angeles; and Wenzheng Mao, a PhD candidate at the University of Hong Kong.

It found that food delivery app customers are likely to punish late deliveries more severely than they are to reward food that arrives early. It also found that driver experience and local knowledge played a crucial role in reducing delivery time.

Burgeoning Growth

According to a 2019 Statistica Report, the global online food delivery sector netted US$94.4 billion in revenues during the year, with China being the single biggest market, accounting for more than 42 percent. Meituan, which is publicly traded, has over 300 million registered members, 3.6 million registered restaurants and 2.7 million delivery drivers delivering 24 million meals across China daily. Ele.me boasts 260 million customers.

Prof. Ming attributes this growth to China’s great urban migration, which has seen the country’s massive rural population take to the cities for better job opportunities and living conditions. This came in tandem with the advent of GPS location tracking and mobile payment which has also spawned online platforms such as Uber and Lyft that bring together buyers and sellers in transactions using various internet-connected devices.

“Meal delivery companies need to pay particular attention to the fact that customers are more likely to punish late deliveries than they are to reward early deliveries.” – Prof. Liu Ming

The researchers began their study by looking at the impact of early or late meal delivery on a customer’s future order and that of drivers’ local area knowledge and experience on the earliness or lateness of a meal delivery.

Prof. Ming and his collaborators discovered that, while different meal delivery platforms used in-house algorithms to assign orders to drivers to minimise average delivery time, they did not yet take the driver’s experience and local area knowledge into consideration when assigning orders to drivers. This gave rise to a third research question: How should a platform assign customer orders to drivers in order to effectively increase a customer’s future orders?

The study found that, unsurprisingly, early deliveries increased future customer orders. Specifically, a 10-minute earlier delivery is associated with an increase of one order per month from each customer. Meanwhile, late deliveries strongly decreased future customer orders — specifically by 2.7 orders per month.

“We found an asymmetric effect whereby early delivery increased future orders a little bit, but late deliveries severely cut the likelihood of future orders,” he says.

The second finding was that a driver’s local area knowledge and delivery experience significantly affected a driver’s delivery time, reducing late deliveries significantly. A driver who has had 30 days of additional work experience reduced delivery time by 5.10 minutes per order, and a driver with local knowledge reduced delivery time by 3.33 minutes per order.

Using simulations, the researchers also found that meal delivery platforms that focus on maximising a customer’s future orders (by making sure deliveries arrived on time) resulted in 0.72 more orders per month, compared to a focus on minimising the delivery time alone. It also led to 1.11 more orders per month compared to a policy focused on minimising delivery distance.

Capturing Market Share

Prof. Ming notes that restaurants are increasingly partnering with food ordering and delivery platforms such as Grubhub, Deliveroo (UK) and Ele.me to capture office workers and consumers who prefer not to eat out or take out, with the platforms charging restaurants a commission fee based on the value of each order. Such partnerships have resulted in 9.3 percent annual growth in the online meal on demand sector.

However, he cautions that as more delivery platforms enter the market, they must compete on customer satisfaction concerning their meal delivery service to retain existing customers and acquire new ones — especially as the customer’s cost of switching platforms is low.

“To increase future customer orders, on-demand service platforms should address the issues arising from both the supply side, by improving drivers’ local area knowledge and delivery experience, as well as from the demand side,” says Prof. Ming.

“Meal delivery companies need to pay particular attention to the fact that customers are more likely to punish late deliveries than they are to reward early deliveries,” he says, adding that meal delivery platforms should therefore focus on striking a balance between early and on-time deliveries, instead of purely on the shortest delivery time.

He also warns against a pure focus on punishing drivers for late deliveries, a system that has been implemented by some apps such as Meituan and Ele.me in China. Prof. Ming notes that this must be done with caution, as drivers in China have demonstrated that they may prioritise speed over safety as a result, leading to traffic accidents.

He adds that, as a platform grows its business by hiring new drivers, some will be less experienced than others. An online delivery platform can offset this by leveraging technologies that can suggest the most efficient indoor route to deliver orders quickly, as well as encourage experience-sharing among drivers.