This study explores user-centered personality design requirements for service robots, moving beyond traditional approaches based on human personality psychology. Through an experiential futures methodology, we investigated what personalities three types of service robots—medical, transportation, and hospitality—should have from users' perspectives. Twelve participants, divided into three groups, created future scenarios and identified personality traits that robots should pursue or avoid. Our analysis revealed both context-specific and universal requirements. Context-specific patterns emerged: "competent empathy" for medical robots, balancing professional competence with emotional care; "reliable adaptability" for transportation robots, combining rule adherence with situational flexibility; and "calibrated friendliness" for hospitality robots, merging warmth with professional efficiency. Four universal principles were identified across all robot types: trustworthy attentiveness, moderated expression, contextual responsiveness, and competent reliability. Notably, participants emphasized avoiding certain traits, such as being insensitive or exaggerating, as strongly as they pursued positive ones. These findings suggest that robot personality design requires a bidirectional approach, considering both desirable and undesirable traits, and adapting to specific service contexts. This study contributes a new user-centered framework for robot personality design that balances functional competence with appropriate social interaction, extending beyond mere application of human personality models to robots.