Gig Worker Learning: An Empirical Study on On-Demand Delivery Platform

Abstract

The advent of on-demand delivery facilitated by gig platforms is ushering in a new era of business operations, workforce management, and customer experiences. Problem definition: This study, motivated by the dynamics of the emerging gig economy, aims to explore how prior delivery experience impacts gig worker productivity and service quality in this new business context. Methodology results: Utilizing data from a leading on-demand delivery platform in Asia, our empirical analysis with the Heckman selection model uncovers several key findings. First, we observe that experience enhances worker performance by increasing the earning rate and reducing the average delay time. However, the probability of order delays initially increases with experience, indicating a temporary drop in service quality. As experience grows, this probability declines, leading to an eventual improvement in service quality. To understand these findings, we show that experience improves order processing times, thereby increasing productivity. Moreover, when gig workers have less experience, they prioritize delivering individual (non-batched) orders to explore new areas and gain delivery knowledge. This exploratory behavior boosts productivity but can decrease service quality due to visits to unfamiliar locations. As gig workers accumulate more experience, they shift towards batching more orders and concentrating on familiar regions and stores. This transition, facilitated by accumulated experience, enhances service quality. Finally, our results show that concentrated experience in specific regions or stores can enhance delivery quality by reducing delays, though it may also lower productivity. Experience with batched orders tends to improve service quality, whereas working during peak hours boosts productivity but may reduce service quality. Managerial implications: This study explores the learning curve of gig workers in the emerging gig economy and provides insights into how platforms can support the sustained long-term development of gig workers as they navigate the evolving future of work.

Keywords: learning, gig workers, productivity, quality, empirical., experience, exploratory behavior