Self-service business intelligence systems in production domains: Introducing data democratization and mass customizable mashups

This paper explores the observation that making business intelligence available through self-service business intelligence (SSBI) systems has only to a lesser degree been successful in organizations. In this paper, we specifically explore SSBI in the production domain, where Internet-of-Things and production data are core data sources in production companies. This paper takes a design science approach, where a structured literature review is used to establish the current research frontier. From our analysis of the problem situation, and close empirical collaboration with a company experiencing problems with implementing SSBI systems, we propose and design two digital artifacts. The two artifacts are partly evaluated based on empirical data from the case company. In this paper, we find that data should not only be made available in SSBI systems, but must also be presented in accordance with domain knowledge of the users. This finding implies that the SSBI literature should to a larger degree draw upon insights from relevant IS literature on contextualized data. Practitioners should design IS systems with contextualized data at their core, as the digital twin discussed in this paper. The second main finding of the paper is that analytics in SSBI systems should be prescriptive analytics, which directly supports the users’ decision-making. A novel contribution of this paper is integrating the concept of prescriptive analytics SSBI systems, and in this paper we discuss and partly validate how integrating prescriptive analytics and SSBI systems can improve decision-making in a production context.