Empirical analysis of a distributed software development project

Abstract

In spite of the abundant research that promotes different methods for software development, and the current method war amidst agile and disciplined methods, little research is done to actually figure out whether real projects, carried out in industrial environments, benefit more from either approach. This paper analyses a real project team's opinions and feelings about project management techniques, software development methods and cultural difference in a multi-site project where traveling and communication are made difficult by restrictions and low-quality infrastructure. The different sites also worked in different time zones and with different working week patterns. The project team members almost unanimously indicated that the presence of a local team leader with authority and flexibility to cover a role that is not exactly as assigned in the beginning, is key factor for the success of this sort of projects. While there was no consensus on whether the project was agile or disciplined, evidence seems to hint towards a more disciplined approach, probably as a compensation for the higher degree of uncertainty that derives from the distributed setup. While the findings of the case study cannot be extended to other organizations without caution, we do infer a number of conclusions on cultural differences, project management tools and techniques. © 2008 Springer-Verlag Berlin Heidelberg.