A two-stage EBM-based approach to evaluate operational performance of unattended convenience store

Purpose: Due to recent technological advances, the retail industry has changed significantly. This paper examines a novel unmanned retail mode-unattended convenience store to identify the possible operational problems and develop appropriate managerial recommendations. Design/methodology/approach: A data-driven two-stage epsilon-based measure (EBM) data envelopment analysis (DEA) method was developed to evaluate operational performance data from 33 unattended convenience stores and assess the impacts on efficiency of the internal factors, and a Tobit regression analysis was employed to examine the external environment. Findings: It was found that the overall economic performances were relatively low and fluctuated significantly; however, the social performances were slightly higher. The out-of-stock rate was found to have a negative impact on efficiency, and regional characteristics were found to have significant effects on performance. Practical implications: This study sought to identify current operational problems with unattended convenience stores to provide managerial insights. The cross-sectional assessment suggested that to achieve better performance, particular attention needed to be paid to store locations and surrounding store environments. Originality/value: First, this paper establishes a novel theoretical framework to evaluate the economic and social operational performances at unattended convenience stores. Second, it contributes to research on unattended convenience stores and the unmanned retail industry and offers significant guidance on detecting operational deficiencies and improving future performances.