Structure and evolution of communication networks in organizations

Abstract

We study three types of communication data--emails, phone calls, and meetings-in a modern IT firm. Using network descriptive statistics, we show how communication networks in an organization differ from random networks and other social networks. We also compare and contrast the three types of communication networks. Using Quadratic Assignment Procedure (QAP), Multiple Regression Quadratic Assignment Procedure (MRQAP) and Exponential Random Graph Models (ERGM), we identify significant factors affecting the size and shape of communication networks. These parameters include organizational structure, homophily, job position, and physical proximity. We record the evolution of the networks and discuss how the factors affecting initial network growth differ from the steady state of the network.Erik Brynjolfsson George and Sandi Schussel Professor of Management Director, Center for Digital Business, MIT Sloan School of Management