Evaluating consumer shopping, delivery demands, and last-mile preferences: An integrated MDCEV-HCM approach

Abstract

The effective implementation of innovative last-mile delivery approaches depends on understanding two key elements: (i) consumer demand (who places an online order, where, and how) and (ii) consumers’ delivery needs and preferences. This study, first, proposes a disaggregated online demand modeling framework utilizing Multiple Discrete-Continuous Extreme Values (MDCEV) to estimate consumer shopping behavior and households’ home delivery and pick-up demands across different commodity types. Second, a Hybrid Choice Model (HCM) is introduced to assess the competitiveness of three innovative last-mile delivery modes, (i) Autonomous Delivery Robots, (ii) Crowdsourced Delivery, and (iii)Automated Parcel Lockers, considering consumer attitudes toward these technologies. Subsequently, we conduct elasticity analysis for cost and commodity type, revealing consumers’ Willingness-to-Pay for various last-mile delivery methods. The proposed framework is applied to a dataset acquired through an online survey distributed among residents of the State of Tennessee, USA. Analyzing results show that consumers can be categorized into five latent segments according to their shopping preferences: traditional shoppers, benefit seekers, e-shopping enthusiasts, omnichannel consumers, and Indifferent customers. Results indicate that businesses should focus on delivery time for e-shopping enthusiasts and omnichannel consumers, while accessibility to APLs may encourage traditional shoppers and benefits seekers to transition to online shopping. Also, latent variable analysis shows that while perceived risk hinders adoption, perceived benefits and ease of use drive acceptance. The findings of this study highlight the importance of a tailored approach to adopting innovative delivery solutions, ensuring a balance between cost, accessibility, and consumer priorities to meet evolving demands.