All-for-one and one-for-all? A multi-level analysis of communication patterns and individual performance in geographically distributed software development

Abstract

It is well established that distributed software projects benefit from informal communication. However, it is less clear how patterns of informal communication impact the performance of the individual developers. In a study of communication networks in a large commercial software project, we found that individuals performed better when they were central within a team's communication network but their performance worsened if they were central within the communication for the whole project. On the other hand, individuals embedded in a dense communication cluster at the team and at the project level perform better than those who were not embedded. The effects for both network positions were maintained even after controlling for formal role, individual differences in communication, workload and other factors that drive communication. We discuss the implications of the results for intra- and inter-team communication and for the inclusion of network structure into the design of collaborative and awareness tools. © 2012 ACM.