A growing number of ridesharing apps is available to help users arrange real-time shared rides. Ridesharing service providers implement platforms that mediate between the requirements of drivers and riders to tackle the challenges of the chicken and egg problem: an intermediary should attract passengers by means of creating a large installed base of drivers who will be willing to register only if they expect that many passengers will adopt the service. Therefore, dynamic ridesharing services need to reach an initial critical mass of users in order to provide desirable level of matching between drivers and passengers. To help in this task, the objective of this work is to investigate some of the levers that could boost the diffusion of a new ridesharing service. This is done by the means of a System Dynamics (SD) model. The model integrates existing diffusion models retrieved from the literature with the main outcomes of a project to which the authors contributed. Three main results are found out of simulation runs and case-scenario analyses. First, the service is expected to work effectively in densely populated urban areas, due to higher contact rates between users that turn in increased matching opportunities. Second, 2 since dynamic ridesharing has yet to be established as a valid alternative for private urban mobility, a high level of service is required. The combined effect of lower population density with high level of desired matching by users is substantially detrimental for the adoption of the service. Third, should the service provider decide to focus on just one type of user, it would have a negative long-term effect on the matching, leading to a higher discard rate of the service.