App-based Mobility as a Service (MaaS) platforms combining public transport, car- and micro-mobility-shared services with dynamic carpooling are emerging as viable alternatives to solo car use for sub-urban contexts. Insights from real-life implementation are however still limited. Which practical challenges affect MaaS platforms leveraging carpooling? We tackle this question from the perspective of potential users of the Swiss-based MixMyRide platform, engaging them in co-design workshops. We find four elements of practical interest, resonating with limitations already identified for carpooling. First, carpooling increases the number of inter-changes potentially affected by delays. This requires real-time traffic information data, re-scheduling tools, and features for quick interaction between users. Second, as social control is low, features to create trust between strangers are needed, which calls for trade-offs between strict identity checks and quick registration. Third, carpooling pick-up/drop-off may endanger safety if bus stops are used. This requires in-advance agreements, negatively affecting the MaaS’ dynamism. Fourth, car-pooling offer is not granted. To accept possible discomfort, decreased flexibility, and effort to enter ride offers, drivers need incentives, such as sharing of travel expenses, virtual or tangible rewards by public institutions, and feedback on saved emissions.