Enchanted Algorithms: How Organizations Use Algorithms to Automate Decision-Making Routines

With technological advances that have contributed to the emergence of “Big Data,” organizations increasingly try to replace routines guided by human decision-makers with decision-making routines automated by complex algorithms that draw on sophisticated mathematical expertise. This trend creates challenges for organizations since the use of these algorithms requires organizations to appropriate this complex mathematical expertise while maintaining their ability to perform routines flexibly. I seek to provide insight into this phenomenon by examining the research question, how do organizations use algorithms to automate routines? To answer this question, I study how a law enforcement organization and an expert in game theory use an algorithm to automate a decision-making routine for scheduling and deploying security resources. I find that the law enforcement organization reconciles their grounded, embedded, and pragmatic approach to the routine with the algorithmic expert’s inherently abstract, disembedded, and mathematical approach to the routine. Although prior research highlights the role of algorithms in rationalizing decision-making processes, my findings suggest that algorithms compel organizational experts to inscribe their dynamic, multi-faceted domain expertise into an “enchanted” algorithm. By offering an empirically grounded analysis of how organizations use an algorithm to automate a routine, I contribute to strategy research on the emergence of routines and capabilities, as well as research on organizational decision-making more broadly.