Abstract
The growth of e-commerce in Asia Pacific is outpacing the rest of the world, and last mile logistics companies which serve the region are exploring ways to improve their services as home deliveries gradually replace traditional retail shopping. In Europe, people are increasingly turning to the use of lockers and retail stores to receive their parcels, particularly when home deliveries are not feasible. We define these locations collectively as Delivery Points, which are crucial in facilitating the growing number of parcel deliveries, especially in urban environments. We study the case of Singapore, in which an incumbent dominates the postal services landscape but is facing competition from up and coming last-mile logistics companies that offer parcel deliveries for e-commerce retailers. Utilizing data provided by Ninja Van, a last-mile logistics start-up that is rolling out a network of Delivery Points to supplement its home delivery services, we observe and report several characteristics about urban e-commerce deliveries and the effectiveness of Delivery Points. About 75% of e-commerce parcels of the logistics company’s 16,500 deliveries over an observed 3-day period are small, meaning that they weigh less than 4kg and have dimensions that do not exceed 80cm in total. The failure rate for first attempted home deliveries is 7.6%, but this rate more than doubles for the second and third attempted deliveries due to challenges in rescheduling. Customers in Singapore were not receptive to using Delivery Points in the first 3 months they were introduced by the logistics provider in the study, although this situation is gradually improving as new Delivery Points are added and people become more aware of the service. As of January 2016, around 12% of customers agreed to have their parcels redirected to Delivery Points when they were asked. They also appeared to respond more positively when these requests were made on Tuesday, Friday, and Saturday. Yet, parcels sent to Delivery Points constituted a miniscule 5-6% of all deliveries in selected service regions. Hence, even though the delivery failure rate for Delivery Points is substantially lower than home deliveries, they do not currently have a meaningful impact on the overall failure rate for deliveries made. Despite these low figures, Delivery Points are effective in reducing the cost of e-commerce deliveries as multiple packages can be sent to a single location. This is especially important during peak seasons, when diverting a small percentage of parcels to Delivery Points could help to maintain high service standards. In terms of reducing the vehicle kilometres travelled for delivery drivers, Delivery Points also appear to be proportionally more effective the greater the number of parcels scheduled for a driver’s daily route. We find that Delivery Points with the highest volume are those within shopping malls that are connected to the MRT network. For deployment planning purposes, the effective range of Delivery Points in Singapore is a radius of 1.6km, as 90% of all parcels sent to Delivery Points have associated home delivery addresses within that distance. Finally, we suggest directions for future research to supplement findings in this thesis. These include developing a better understanding of the dimensions of small e-commerce parcels to build clusters of larger letterboxes, comparing different solutions for Delivery Points, and understanding the system-wide impact of diverting more parcels to Delivery Points.