It is a common and frequently implicit assumption in the literature on knowledge transfer and organizational learning that imitating practices from high-performing firms has a positive impact on the imitating firm. Although a large body of research has identified obstacles to successful imitation, not much is known about what breadth of imitation is most effective. In this paper, we use a simulation model to explore how context and firm similarity, interdependence among practices, context and firm similarity, and time horizon interact in nontrivial ways to determine the payoffs that arise from different breadths of imitation. The results of the model allow us to qualify and refine predictions of the extant literature on imitation. In particular, the results shed light on the conditions under which increases in imitation breadth, and hence investments that facilitate the faithful copying of more practices, are valuable. In addition, the results of the model highlight that imitation can serve two different functions-mimicking high performers, and generating search by dislodging a firm from its current set of practices-each requiring different organizational routines for its successful implementation. How much to copy? The contingent value of imitation capabilities Abstract: It is a common (and frequently implicit) assumption in the literature on knowledge exchange and organizational learning that imitating practices from high-performing firms has a positive impact on the imitating firm. While a large body of research has pointed out obstacles to successful imitation, very little is known about what degree of imitation is most effective. In this paper, we use a simulation model to explore the role that interdependence among practices, firm-similarity, and time horizon play in influencing the value of different degrees of imitation, and show how they interact in non-trivial ways.For instance, we find that in the presence of interactions, the most effective imitation strategy between similar firms with long time horizons is the worst strategy for short time horizons. One implication of our results is that even if a firm has the capability to copy all practices from a high-performing firm, this will only occasionally be the most appropriate imitation strategy. We also show that imitation can serve two different functions-mimicking high performers and dislodging a firm from its current set of practiceseach one requiring very different organizational routines for its successful implementation. Lastly, we use the model to shed light on three previous disputes in the literature: the controversy between slow and fast learning, whether imitation is effective only at the start of operations or on a continuing basis, and whether firm similarity increases or decreases learning opportunities.

This paper develops a parsimonious process-level theory that connects organizational structure to exploration and exploitation. Toward this end, it develops a mathematical model of organizational decision making that combines an information processing approach in the spirit of Sah and Stiglitz [Sah RK, Stiglitz JE (1986) The architecture of economic systems: Hierarchies and polyarchies. Amer. Econom. Rev. 76(4):716–727] with elements from signal detection theory. The model is first used to explore a “design space” of organizations and identify trade-offs and dominance relationships among alternative organization designs. The paper then studies open questions in the organization design literature, such as the extent to which exploration and exploitation can be produced by one organization and what is the effect of organization size on exploration. More broadly, this research speaks to calls for the introduction of more process-level explanations in the organizations literature. The paper concludes with testable hypotheses and managerially relevant insights.

Sah and Stiglitz (1986).