Business Performance Analytics (BPA) entails the systematic use of data and analytical methods (mathematical, econometric, statistical) for performance measurement and management. Although potentially overcoming some traditional diagnostic issues related to Performance Management Systems (PMS), such as information overload, absence of cause-effect relationships, lack of a holistic view of the organization, research in the field is still in its infancy. A comprehensive model for operationalising analytics for diagnostic and interactive PMS is still lacking. Adopting an action research approach, this paper addresses this gap and develops a five-step framework applied to a company operating in the construction industry. The results show that in addition to encouraging dialogue, BPA can contribute to identifying critical performance variables, potential sources of risk and related interdependencies. A number of critical issues in implementing data-based approaches are also highlighted including data quality, organizational competences and cultural shifts.