The lean distributed scrum (LDScrum) model: investigating the relationship between lean and agile software development in a distributed development context

Agile software development (ASD) has emerged as a practice-led initiative which offers great promise in improving software productivity. However some confusion exists as to its relationship with Lean Software Development (LSD). Some treat LSD as more or less synonymous with ASD whereas others view LSD as a different concept. The definition and positioning of LSD relative to ASD is important as it gets to the heart of software development as craft versus science debate. The purpose of this thesis is to present a deeper understanding on the relationship between these two paradigms. It is proposed that LSD is more management philosophy than method. In order to facilitate this exploration, a set of core LSD values is induced from literature and proposed to ‘define’ LSD, much as the agile manifesto values unified and defined socalled ‘lightweight methods’ for ASD. In order to narrow the focus of this investigation, the broad management philosophy of LSD is explored in the context of a very specific instance of ASD: the application of the most dominant ASD method (Scrum) in the management of projects that use globally distributed teams (GSD). The main deliverables from this interpretive study are a research framework and a thick description of empirical findings in this regard. The Scrum method is described as comprising of a set of core Scrum practices. The research framework leverages method rationale analysis theory in order to propose LSD values that may underpin goals pursued by the application of those Scrum practices within globally distributed teams. It is recommended that this research framework be applied to further investigations into the influence of LSD on other Scrum manifestations as well as other ASD methods. An interpretive case study approach was applied to examine different Scrum teams and projects. Although all LSD values were found to be present in the case, certain values were more prevalent than others. ‘Effective collaboration’ (reduction of distance) and ‘Person-focus’ (individual and team empowerment) were the most prevalent LSD values observed. ‘Continuous improvement’ was also very evident. One project within the case showed very little awareness of the core LSD values of ‘Flow of value’ and ‘Waste reduction’. Across the entire case, the value of ‘Data-driven decisions’ was not seen to have a strong influence on developer intentions. This lack of statistical rigour may reflect the craft versus science controversy in relation to software development.