Persistent high growth firms have been suggested to be essential for advancing economies and resolving societal challenges. Yet only a small minority of high growth firms persist over time and scholars have debated about whether persistent high growth is just a random process or may emerge from deliberate and skillful management. Our study uses a configurational perspective and combines advanced machine learning with high-end capability computing to explore how growth, restructuring and market actions interact to predict persistent high growth. Our findings suggest that persistent high growth can be predicted very accurately, and as such, our study implies that persistent high growth is not a random walk, but rather emerges from a concerted and interrelated configuration of managerial actions. With that, our study applies algorithmic inductive theorizing that provides important implications for understanding persistent high growth, the role of specific and effective configurations of managerial actions, as well as shows the potential machine learning approaches may provide management scholars to research forward and to predict complex outcomes in general.
