Track: Graduate Student Paper Competition

Abstract

Problem definition: On-demand platforms operating on a gig-contractor model rely on self-scheduling gig workers to meet real-time demand, while maintaining desired levels of productivity and service quality. Such platforms usually allocate gig workers to tasks on a first-come-first-serve (FCFS) basis, which is often inefficient. This paper seeks to investigate whether allocation of workers to tasks based on prior experience and task characteristics could act as an alternative to the FCFS method. Since on-demand platforms frequently expand to new geographies where workers might not have prior experience, it is unlikely that allocating tasks to workers based on experience alone would significantly improve performance.We therefore investigate whether recent (same-day) experience can positively impact gig workers’ task performance, and thus be utilized as a ranking parameter in task allocations. Methodology/results: Utilizing data from an on-demand grocery platform, we first develop an econometric model to analyze gig workers’ task performance based on their same-day experience, while accounting for sample selection and endogeneity. Our findings reveal that as same-day experience increases, task productivity and service quality improves, even after controlling for worker and task characteristics. The positive effect is stronger for batched tasks. We further observe
that when task complexity is higher, although same-day experience improves productivity, service quality is negatively affected. A similar effect is observed when gig workers use their discretion to perform complex tasks. Managerial implications: Using prior- and same-day-experience, we rank available gig workers. We then develop two task allocation algorithms to allocate tasks based on workers’ rank, and task characteristics. Using FCFS allocations as a baseline, we calculate predicted improvements in productivity and service quality from the new allocations. Results from the task allocation algorithms demonstrate that allocating
higher ranked workers to complex tasks, or to new customers leads to an improvement in task performance and customer repurchase probability.