Dynamic Patterns of Industry Convergence: Evidence from a Large Amount of Unstructured Data

Because of the accelerated life cycle in technology and correspondingly rapid technological saturation in markets, firms are not only accelerating the rate of technological innovation but also expanding the scope of their products or services by combining product or service features of other markets, which eventually leads to industry convergence. However, despite the significant impact of industry convergence on the economy, our understanding of the phenomenon is still limited because previous studies explored only a few cases and come largely from the technological perspective. Therefore, it is still questionable whether industry convergence is a general phenomenon that is prevalent across entire industries. In this paper, we analyze the phenomenon in entire U.S. industries, focusing on its trends and patterns. To do so, we conduct a co-occurrence-based analysis of text mining for a large volume of unstructured data – 2 million newspaper articles from 1989 to 2012 – and suggest using an industry convergence (IC) index based on normalized pointwise mutual information (PMI). We find that overall industry convergence is increasing over time. Moreover, the rate of the increase has been greater within industry than between industries at a given industry level. However, when we cluster the dynamic patterns of industry convergence among industry pairs, the patterns are mixed, and, while some industry groups are converging over time, others are stationary. These findings suggest that significant transformation is under way in the economy, but this phenomenon is not yet prevalent across entire industries. In addition, this study provides a method for anticipating the future direction of industry convergence.