Assessing inter- and intrapersonal variability in carsharing usage

Carsharing is rapidly gaining popularity worldwide as it provides fleets of shared vehicles that reduce car ownership and promote the use of transit and active transportation modes. Carsharing operators collect vast amounts of data that offer valuable insights into usage patterns. However, most existing studies rely on static analyses that capture system usage at a single point in time. They primarily focus on interpersonal variability—differences between users—while intrapersonal variability, i.e., changes within individual users over time, remains largely unexplored. Furthermore, few studies assess the stability of carsharing usage behaviors over extended periods. Most existing research focuses on early-stage systems, which attract early adopters in dense, homogeneous areas. As carsharing systems expand and mature, their user base becomes increasingly heterogeneous, giving rise to more diverse and evolving behavioral patterns that require further investigation. This study addresses these gaps by developing a comprehensive methodology to examine both interpersonal and intrapersonal variability in carsharing usage, and to assess the temporal stability of these behaviors. Using three years of reservation data from Communauto, Montreal's largest carsharing provider, we implement a four-step approach: data preprocessing, creation of a user-month typology, creation of user-year profiles, and analysis of behavioral stability. This framework identifies eight distinct behavior-based user groups across monthly periods and evaluates how individuals shift between them over time. The proposed indicators and analytical approach provide actionable insights for operators to better understand behavioral heterogeneity, anticipate evolving mobility needs, and improve service planning and management.