Predicting Relevant KPIs for Roles in Organizations

Abstract

Business Intelligence (BI) and Business Process Monitoring (BPM) have become staples of modern management. BPM and BI keep track of business processes and capture them in Key Performance Indicators (KPIs). KPIs are visualizations of metrics that measure the status of business processes. By consulting KPIs, the status of a business process can be easily assessed and acted upon. Organizations provide KPIs to their employees to improve their decisions relating to the captured business processes. Creating KPIs is a time and cost-intensive process. To reduce the effort required for this process, some research has been done to automatically determine relevant KPIs. These research efforts have yielded approaches that either require manual labor or focus on organizations instead of end-users. However, predictions that fit organizations might not fit end-users. To provide KPIs relevant to end-users, their roles in their organizations should be taken into account when predicting relevant KPIs. By taking the characteristics of roles into account when predicting relevant KPIs, predictions can be specified for information needs specific to a role. This specification can increase the relevance of KPIs to end-users, improving their decision making. There is currently no approach available for automatically predicting relevant KPIs for roles in organizations. This study shows that the relevance of KPIs can be predicted based on role characteristics. The relevance of a KPI to a role is shown to be determined by the role's responsibilities and its place in the organizational structure. The approach presented in this study was applied in a real-life setting and evaluated with experts. The results show that role characteristics can be used to predict relevant KPIs for roles in organizations. This research presents an approach that is able to specify the prediction of relevant KPIs from organizations to roles while decreasing manual effort required.