Between control and consent: insights from dark stores into middle managers’ roles and organizational dynamics in the age of algorithmic management

Last decade’s technological advancements have introduced a major transformation in the world of work: the inception of algorithmic management. While workplace monitoring and assessment through performance metrics are not new, algorithmic management takes a giant leap forward by leveraging the computational and deep learning abilities of algorithms to assemble and process vast amounts of data, enabling automated or semi-automated decision-making. In light of this, some scholars have explicitly or implicitly predicted a future where the role of human managers, particularly low- and middle-level managers, will become insignificant compared to that of algorithms.

However, against this potentially overly deterministic prediction, other studies have argued that the deployment of algorithmic management tools does not, and will not in the foreseeable future, eliminate the need for human managers. Indeed, middle managers remain a salient and everyday presence in the labor process of organizations despite the deployment of algorithmic tools. This is because, despite the potential of artificial intelligence (AI) and computational algorithms to improve firms’ efficiency, some critical managerial traits cannot be automated.

This article aims to understand how algorithmic tools reshape the role of middle managers and how this, in turn, transforms the dynamics of managerial control in workplaces. Drawing on 51 semi-structured interviews with managers and shop-floor workers from last-mile grocery delivery warehouses – a sector we consider both revelatory and critical for our purposes – our findings show that, notwithstanding the automation of several coordination-related tasks, middle and low management retains a crucial role in organizations. This role has two facets that appear contradictory but are, indeed, complementary.