Understanding Coordination in Distributed Agile Software Development

Abstract

Coordination in modern Software Development involves high levels of collaboration between multiple individuals having different capabilities, skills, and knowledge to achieve common goals. Coordination in this work environment requires harmonisation between tasks and deliverables, knowledge sharing among team members, and a shared understanding of how to work together. As a result, effective coordination in co-located software development projects is associated with several contemporary issues and challenges. These issues are even more complicated in geographically distributed contexts because several coordination mechanisms supporting co-located projects are either absent or inadequate. In addition to the challenges associated with spatial, temporal, and configurational differences, projects in distributed settings face increased coordination complexity. Having witnessed the success of agile methods in small co-located projects, organisations are striving to blend agile methods and distributed development to reap both benefits. Therefore, this thesis explores how coordination is managed in Distributed Agile Software Development projects.

Evidence was drawn from a multi-case study approach to develop a comprehensive coordination model with four key concepts (i.e. dependency type, dependency risk, coordination strategy, and coordination effectiveness), along with their constituting components and associations. The model explains that there are knowledge, activity, and resource type dependencies in distributed agile software development. These dependency types are influenced by the dependency relationship and situational factors, and may involve vulnerabilities that could lead to dependency risks. A well-formulated coordination strategy contributes to eliminating dependency risks and, thus coordination effectiveness.

It contributes to the body of knowledge in several ways: first, it presents an empirically validated coordination model in distributed agile software development. The model includes three essential concepts of coordination: a taxonomy of dependency, coordination strategy, and coordination effectiveness. Second, it introduces the dependency risk concepts associated with the key dependencies that influence coordination effectiveness. This relationship also guides the formation of the coordination strategy in achieving coordination effectiveness in distributed agile software development. Finally, it prescribes a repertoire of mechanisms for the practitioners to improve coordination performance by alleviating dependency risks and coordination challenges.

The framework of analysis developed in this research can be adapted and applied in future coordination research in different contexts. The study’s findings also have important implications for practitioners. It suggests strategies for coordination effectiveness by considering the influence of the dependency antecedents and associated risks. The coordination strategy concept can be used to select agile practices that ensure effective coordination in distributed contexts.