ABSTRACT Significant performance differences in the process of knowledge creation may exist between organisations with exploratory and exploitative tendencies due to differences in their inner network structures. This study focuses on the subgroups widespread in intra-organisational networks and their structural characteristics. We construct an agent-based model that describes the two different organisational learning models that exist in organisations – exploration-oriented and exploitation-oriented. We find that higher network density in each subgroup helps increase the speed of knowledge creation but depletes the idiosyncratic sets of beliefs and leads to premature convergence. Only when the network scale is large enough or the organisation tends to focus more on exploration can a sparse network play a positive role in knowledge creation. Moderate levels of cross-group links prevent exploitation-oriented organisations from entering the stagnation stage of knowledge growth and increase the knowledge creation performance of exploration-oriented organisations. Knowledge creation performance in exploitation-oriented organisations depends more on boundary hubs than on internal hubs. The presence of twofold hubs can improve knowledge creation performance without significantly reducing the dissimilarity of collective beliefs.
Read full abstract