In recent years, the emergence of shared mobility services has created new opportunities for more flexible intermodal travel, but the behavioral barriers to their adoption remain poorly understood. Using stated-preference data collected in 2024 in Geneva, Switzerland, we estimate mode-pair-specific pure transfer penalties across public transport, personal bikes, and shared mobility services. A set of itinerary choice models are specified using an assisted algorithm and refined manually to ensure plausibility and consistency with behavioral theory. Results highlight significant variation in perceived transfer disutility, influenced by mode familiarity and user characteristics. These findings contribute to a deeper understanding of intermodal travel behaviors, enabling their integration into simulation models and informing the design of mobility hubs in practice.