Strategy Experiments in Nonexperimental Settings: Challenges of Theory, Inference, and Persuasion in Business Strategy

Abstract

We introduce two dimensions that help situate the approach to strategy as an exercise of theory building and experimentation. First, we consider whether a given experiment is repeatable or nonrepeatable: the extent to which the very undertaking of a given experiment will change how the world—consumers, competitors, ecosystem partners—will react to future experiments, even if identical to the original. Second, we consider whether a given experiment can be undertaken by a single actor or requires the joint effort and buy-in of other actors in the internal or external ecosystem of the firm, and, in turn, the extent to which persuasion becomes a critical role of theory. We show how these two properties—nonrepeatability and joint action—impact the nature of inference from strategic actions and the ability to benefit from learning. Further, these properties suggest a more deliberative calculus regarding the cost of experimentation and highlight the role of theory as a basis for both prediction and persuasion in the service of choosing how to initiate and learn from strategic acts in nonexperimental settings.