Generative Artificial Intelligence (GenAI) is increasingly integrated into business environments with the expectation of enhancing productivity across various organizational levels. However, while technical implementation is prioritized, critical aspects such as employee-perceived quality remain underexplored. Existing usability specifications, primarily developed for deterministic systems, fail to capture GenAI's adaptive and probabilistic nature, necessitating a redefinition of user-perceived quality to establish valid design requirements and measurement instruments. This study aims to conceptualize a user-centered quality framework for GenAI chatbots in a business context. Employing a mixed-method approach, the research includes 12 semi-structured interviews and a survey of 512 respondents, followed by an exploratory factor analysis to identify key quality dimensions. The findings reveal five principal dimensions of user-perceived quality: (1) Productivity of Responses, (2) Trustworthiness, (3) Contextual Fit, (4) Domain Suitability, and (5) Operational Usability. These dimensions provide a structured perspective on how employees evaluate GenAI chatbots, addressing critical gaps in existing usability models. The study contributes to the Human-Computer Interaction (HCI) field by extending traditional usability and user experience (UX) frameworks to accommodate the dynamic nature of GenAI. The identified quality dimensions offer organizations a structured approach to assessing and refining GenAI solutions based on user feedback. Future research will focus on further empirical validation through a large-scale study, investigating the correlation between quality dimensions, user satisfaction, and long-term GenAI adoption in business settings. By establishing a robust evaluation framework, this research supports the development of GenAI chatbots that enhance both employee satisfaction and organizational performance.