Implementing a Data Infrastructure to Enable Business Analytics in the Public Sector: a case study

Abstract

This paper describes the initial stages of the process of implementing the data infrastructure required to develop analytics capabilities in a public sector organization. Helfo (the Norwegian Health Economics Administration) is responsible for making payments on from the National Insurance scheme to healthcare actors who submit reimbursement claims. An important task for Helfo is also to prevent and detect errors, and the organization is currently strengthening this capacity though employing data analytics and artificial intelligence. Implementing data analytics entails more than a “plug-and-play” process, and we analyze the initial stages of the implementation process as a sociotechnical change process. As a starting point we employ the CRISP-DM process model and enrich this with additional steps and tasks that was found to be central in the case. In particular, we describe in more detail the preparatory work relating to the technical setup and data infrastructures, and the implications for the information processing routines of the organization more broadly. The case study shows that also the early-phase improvements in data access and utilizing basic analytics capabilities yielded benefits directly, without employing advanced analytics and artificial intelligence. The rich description of the early stages of the implementation process can be valuable for other public sector organizations that seek to build data analytic capabilities.