Towards Routines Mining – Designing and Implementing the Argos Miner, a Design Science Artifact for Studying Routine Dynamics with Process Mining

Abstract

Digital artifacts increasingly support actors in carrying out organizational routines. These artifacts leave digital trace data, that is, time-stamped data about what actions actors performed. While extant research on routines largely builds on qualitative methods, the increasing ubiquitousness and prevalence of trace data enable novel methodological opportunities. However, several challenges currently hinder the adoption of trace data in empirical research on routines in general and their dynamics in particular. Promising approaches such as process mining are neither designed for nor sensitive to the concept of routines. In this paper, we follow a design science research approach to develop the first iteration of an artifact, which we coin Argos Miner. This artifact is based on process mining algorithms and overcomes challenges inherent in adopting process mining in routine dynamics research. It enables scholars to capture reality in flight by analyzing routine dynamics using a computational, mixed-methods approach.