Customer Satisfaction in City Delivery Systems and Its Implications for Delivery Efficiency and Environmental Impacts: A Machine Learning Analysis

The rapid growth of e-commerce has intensified last-mile delivery activities in urban areas, creating challenges for city logistics systems related to operational efficiency, congestion, and environmental impacts. In this context, understanding the factors that influence customer satisfaction with logistics operators is increasingly important, as mismatches between customer expectations and delivery service characteristics may lead to operational inefficiencies such as failed delivery attempts and repeated delivery rounds. This study proposes a machine learning framework for predicting customer satisfaction with postal and logistics operators in urban delivery systems using survey data on customer characteristics, preferences, and service perceptions. Several machine learning algorithms were developed and evaluated to identify the key determinants of customer satisfaction and assess their predictive performance. Beyond predictive accuracy, the study interprets customer satisfaction as an indicator of the alignment between customer expectations and delivery service configurations. Improved alignment may support service configurations that reduce delivery mismatches and repeated delivery attempts, which are recognized as a significant source of additional transport activity in urban freight systems. By identifying customer segments whose expectations are not adequately addressed by existing delivery services, the proposed framework can support more informed service design and operational decision-making. From a city logistics perspective, the potential reduction in failed deliveries and repeated delivery rounds may contribute to lower vehicle kilometers travelled, congestion, energy consumption, and emissions associated with urban freight transport, although these operational and environmental indicators were not directly measured in this study. The proposed approach therefore provides a data-driven decision-support tool that can help operators improve service quality and serve as a basis for future integration with operational and environmental indicators in sustainable last-mile delivery planning.