Frankfurt School of Finance and Management, Germany
Extensive research highlights the role of interdependencies, or couplings, among policy choices in shaping organizational outcomes. However, less is known about how managers understand and adapt to these couplings over time, and how such adaptation impacts organizational performance. To address the gap, we investigate how managers form and adapt mental models of coupled policies through feedback from implementation, and the conditions under which such adaptation improves long-term organizational performance. In line with Ashby’s law of requisite variety, we find that complex mental models excel in complex task environments, while simpler models are more effective in simpler environments. This is explained by the bias-variance tradeoff—overly simple mental models underfit, yielding poor predictions in known regions of the task environment, while overly complex models overfit, reducing generalization to unexplored regions. Adopting mental models whose complexity is aligned with the task environment helps organizations optimize performance by balancing the need for prediction accuracy with adaptability. We argue that such balanced mental models enhance managers’ ability to theorize effectively about value creation and discuss the implications of these findings for managing strategic search in complex task environments.
