Niklaus Leemann, Dominik K. Kanbach

• Purpose This paper aims to categorize and organize dynamic capabilities that have been inductively identified in empirical research into a comprehensive taxonomy. Thus, it addresses calls in the literature for a better understanding of dynamic capabilities and integration of scattered empirical findings into theory. Design/methodology/approach A systematic literature review approach was adopted, with a total of 34 articles published between August 2007 and April 2020, from which 240 idiosyncratic dynamic capabilities were identified. The taxonomy was constructed using the Gioia-method. Findings The main finding is a three-level taxonomy of dynamic capabilities (DC). Level DC-1 is based on the existing triad of sensing, seizing and transforming. Level DC-2 is newly introduced to the literature by this study, consisting of 19 dynamic sub-capabilities that categorize and organize all 240 idiosyncratic dynamic capabilities in the sample (level DC-3). The taxonomy supports the existing claim that dynamic capabilities are common in key features and idiosyncratic in details. Moreover, theoretical connections to business model innovation and ambidexterity are indicated. Practical implications This study integrates scattered empirical findings of specific dynamic capabilities and translates them to a practitioner audience. The taxonomy allows the strategic manager to understand what they specifically are and, thus, assess the dynamic capability endowment of the firm which allows deploying, developing and fostering them. Originality/value The taxonomy provides a comprehensive and tangible picture of what dynamic capabilities look like in practice. It improves existing knowledge and understanding by bridging the rigor-relevance gap between rather rigorous conceptual literature and rather relevant empirical research as it integrates them. As such, it can serve as a “map” of dynamic capabilities for scholars and practitioners.