An Empirical Approach for Optimizing Globally Distributed Software Product Engineering

A distributed software product engineering team has to deal with the additional issue of distribution apart from the usual expectations around cost, quality, time to market and innovation. Distribution of work brings in the issues of temporal, geographical and cultural separation. In spite of individually following the prescribed software engineering processes, often the distributed teams fail to act as a single product team. It is also a well accepted truth that misaligned distributed teams rarely produce satisfactory outcome. The short life cycle for Internet based products adds to these challenges. As a result, global software product engineering is still somewhat immature. In this paper we describe an empirical study to diagnose, detect and cure the health issues in a distributed software product engineering program. This consists of steps such as problem identification, research review, study design, validity consideration, data analysis and conclusion. The key hypothesis in this approach is the assumption that most distributed software product engineering teams in the same organization require alignment rather than bottom up retooling in the form of an elaborate software engineering initiative and this alignment can be achieved in a fast and effective manner by aligning the key interface leaders. By focusing first on causes of misalignment between different parts of distributed team, the root causes behind lack of satisfactory outcome can be quickly surfaced. This top down method is a departure from the usual bottom up approach of starting from process adherence. This approach allows the distributed teams to retain their cultural idiosyncrasies that are hard to remove. This paper shares the experience from such an initiative undertaken in an Internet based product company doing software product engineering across India and USA.