Disengagement through Algorithms: How Traditional Organizations Aim for Experts' Satisfaction

Abstract

This study examines the use of algorithmic tools in traditional organizational decision-making processes. Through forty semi-structured interviews with managers, engineers, and (expert) users across six European projects, we suggest that initiators deploy algorithms not to automate actions or replace users, but to disengage themselves from prescriptive decision-making. Consequently, the responsibility to choose, select, and decide falls upon the users; they become engaged. Therefore, algorithm evaluation is oriented towards utility, interpretability, and, more broadly, user satisfaction. Further research is encouraged to analyze the advent of a 'satisfaction regime', from platforms to traditional organizations.