Understanding bundle choices and subscription patterns of mobility as a service (MaaS) trial participants: evidence from an Australian university community

Understanding choice preferences and patterns of subscriptions to bundles is a critical prerequisite for the successful introduction and widespread adoption of MaaS. Existing studies on choice preferences for MaaS bundles are limited to a single-choice scenario through stated preference surveys, or to limited choices in small-scale trials with few participants. Unlike behaviours observed in single-choice scenarios, analysing individuals’ bundle choices and budget allocations across subscriptions, based on their subscription sequences, offers a more objective way to understand their MaaS subscription patterns. To contribute new evidence, this study utilises data from a large-scale MaaS trial conducted within a university community in South East Queensland (SEQ), Australia, and examines participants’ bundle subscriptions during semester 1 of 2023. The descriptive analysis reveals that many participants repeatedly purchase the same bundle or select multiple bundle types throughout the semester. Further, there appear to be some factors that affect not only the bundle choices but also the semester-long commitment of students and staff to the MaaS program. To explore this further, our study models participants’ bundle choices and the within-semester subscription patterns using multiple discrete–continuous extreme value (MDCEV) models. The model results highlight the significant impacts of the participants’ gender, user type, age, ability to ride an e-scooter, and motivations for joining the MaaS trial do affect bundle choices and subscription patterns over the semester. Moreover, land use variables corresponding to residential bikeway density, population density, land use mix, percentage of land area dedicated to transport, also seem to affect a participant’s preferences for bundles and subscription patterns. The developed models also account for the heterogeneity in participants’ budget allocations into bundle subscriptions across participant profiles. The findings are insightful in understanding users’ subscription patterns for specific MaaS bundles, thus guiding operators and policymakers to enhance their bundle offerings in real-world MaaS deployment and operation.