Escaping real (non-benign) competency traps: linking the dynamics of organizational structure to the dynamics of search

Abstract

Focusing on the effect that organizational structure exerts on organizational search, we show under which conditions a change in structure increases performance even in stable environments. We model five different organizational structures (centralized, decentralized, ambidextrous, hybrid and team-based) and study with the help of an agent-based simulation whether transitions among these structures are beneficial. We find that sequences of structures can achieve higher performance than fixed structures. Alternative structures differ in their competency traps or sets of sticking points, that is the sets of points at which a search process in a given structure will terminate. As a result, a shift in structure may dislodge an organization from its current configuration of choices and provoke further search. Changes in organizational structure effectively differentiate between settings in which remaining at a competency trap is in fact a trap and settings in which remaining at a competency trap indicates competence. In particular, a shift in organizational structure differentially sorts among more or less favorable sticking points, as sticking points that are common to two structures tend to be higher-performing. Thus, behavior that remains inert when structures change will tend to be associated with particularly high-performing sticking points. Moreover, behavior should be dislodged from the prior sticking point, this provides a favorable starting point for subsequent search. Consequently, a shift in organizational structure need not be a response to new environmental contingency, but a mechanism to overcome the challenge of competency traps as well.