The hospitality industry, struggling with significant staff shortages, has increasingly turned to service robots as a solution. However, the prevalent service robot’s design with anthropomorphic appearance is considered inharmonious with the fine-dining restaurant ambiance and may harm the guests’ perception of the service. An alternative approach is exemplified by Rober, which adopts a design resembling a traditional cart. The non-anthropomorphic design offers flexibility, economic efficiency, and enhanced acceptability in hospitality settings. However, it also raises challenges in expressing intentions that are typically conveyed through human non-verbal cues. Consequently, the movement quality of service robots becomes a critical area of design to facilitate nuanced human-robot interaction (HRI) in hotel restaurant contexts. The research focused on two main questions: how to design robot movement to facilitate essential interaction and collaboration qualities during dining experiences, and how to craft these movements using a dramaturgic performative approach. The project employed methodologies like speculative enactment and Extended Reality (XR) experiments to explore and evaluate robot movements. These methods allowed for creative ideation and assessments of the robot’s movements in simulated dining scenarios. The project’s findings revealed that specific robot movements, including refined presence, prompted actions, and engaging addresses, significantly enhance the experience of guests, staff, and managers of the hotel restaurant. The robot’s role was envisioned as an ‘Ensemblist,’ a term encapsulating its function as an integral yet unobtrusive participant in the fine dining scene. This role demands the robot be ‘response-able,’ adapting to the fine dining rhythm. Furthermore, the project’s performative approach illuminated methods to design the robot’s movement as expressively meaningful and contextually appropriate. Methodological reflections revealed the effectiveness of speculative enactment and XR experiments in capturing the complexities of human-robot interactions, though suggesting future improvements in prototype fidelity, participant diversity, and advanced data treatment. This project contributes to the field of HRI in hospitality, bridging theoretical concepts with practical applications. It lays the groundwork for future research in service robot design, emphasizing the need for nuanced interaction designs that resonate with human users in the hospitality sector.