Data-driven performance management of business units using process mining and DEA: case study of an Iranian chain store

Purpose - The underlying purpose of this paper is to propose a comprehensive framework evaluating the performance of business units of an organization with a process perspective, identifying the most influential performance indicators, enabling managers to make more informed decisions based on data recording every day in their operational information systems. Design/methodology/approach - For proposing the conceptual framework of performance evaluation a synchronized analysis of selected process' data, obtained from an integrated information system of an Iranian chain store, was performed. Findings - The superiority of the proposed framework results is demonstrated in comparison to applying the process mining solely; principal component analysis was identified as an efficient link between process mining and data envelopment analysis. Also, based on the final data analytics, the units' throughput times and the variety of brands and suppliers had the most impact on their performances. Research limitations/implications - The data of abundant business units and performance indicators, which would have allowed adding data prediction and other data analytics techniques for more insight, was not able to be accessed. Practical implications - Organizations' managers can use the framework to evaluate their business units' current status and then prioritize their resources based on the most influential performance indicators for overall improvement. Originality/value - The study contributes to the research on performance management and process mining by presenting a comprehensive framework with two levels of data analytics. It stresses discovering what is happening in business units, and how to prioritize their improvement opportunities learning the significant correlations between performance indicators and units' performance.