Strategic categories and competition: significant clustering for strategic groups

Abstract

Purpose Strategic groups research has been hampered by the poor alignment between theory and methods. This has been due in large part to the lack of significance tests for cluster analysis. Now that significance tests are available, the theoretical and methodological implications are discussed. The paper aims to discuss these issues. Design/methodology/approach The theory behind strategic groups is reframed to capitalize on the available significance tests. Subsequently, the significance tests are also modified to fit the proposed theory. Due to this integrative approach, this is both a theoretical and a methodological paper. Findings In lieu of significance tests, finding differences in performance emerged as the litmus test for the existence of discrete strategic groups. The concept of strategic groups gradually evolved to fit this requirement. Now that significance tests are available, these legacy effects of the structure-performance link can be removed. This reveals that three conflicting concepts have been sharing the label of strategic groups: strategic categories, interdependent strategic groups and strategic performance groups. The theory also reveals that the significance tests developed in ecological research need modifications for use in strategic groups research. Research limitations/implications A theory is proposed for interdependent strategic groups and a significance test of external isolation is proposed as part of this integrative solution. Originality/value This integrative solution appears to resolve the historical mismatch between theory and methods that has plagued this field since its inception. This creates a variety of intriguing areas for future research.