Academic journal article Journal of Electronic Commerce Research

The Life Cycle of Open Source Software Development Communities

Academic journal article Journal of Electronic Commerce Research

The Life Cycle of Open Source Software Development Communities

Article excerpt


Drawing from the concept of entropy in open systems theory, this article contributes to organizational theory by illuminating organizational life cycle theory and exploring open source software development communities (OSSDC) with quantitative longitudinal data. In particular, this study uses functional data analysis to uncover the development patterns of open source software projects in terms of effectiveness and activity levels. Our findings show that the life cycles of OSSDC display an inverted-U shape in terms of effectiveness level and an inverted-S shape in terms of activity level. Although our results provide some evidence of distinct states, they do not imply that such states are predetermined or irreversible. On the contrary, these numerous states are viewed here as intrinsically dynamic. These findings not only give empirical support to the organizational life cycle metaphor in the context of OSSDC, but also aid practitioners and policy-makers in assessing online communities. Taking an open systems view of organizations, this study aids in reconciling some issues in life cycle theory, such as the irreversibility and pre-determinacy of life cycle models, and adds to a young but fast growing stream of literature on open source projects. Lastly, our findings remark the importance of fostering active communities for superior effectiveness and long-term survival of the community.

Keywords: Online communities; open source software development; life cycle.

1. Introduction

The notion of the organizational life cycle has been in the literature for more than forty years [Greiner 1972; Lippit & Schmidt 1967]. In spite of the long history, relatively few empirical studies have been conducted [Levie and Lichtenstein 2008]. Furthermore, those few studies have often brought in significant assumptions (e.g., pre-determined number of stages) and have been conducted in settings that might be perceived as less dynamic [D¡Aveni 1994] than those faced by most organizations today.

Research on the development of online communities also lacks empirical grounding [for a recent review, see Iriberri & Leroy 2009]. Some notable exceptions are: the study of the relationship between turnover and collaboration in Wikipedia [Ransbotham & Kane 2011]; the study of the relationship between critical mass and online survival that also remarks the relationship between activity levels and community survival [Raban et al. 2010]; and the changes in individual roles in Twitter tag-based communities [Sonnenbichler & Bazant 2012].

This research seeks to fill this gap, expanding our horizons in relation to life cycle and growth models and exploring the new field of online communities with quantitative data. Instead of assuming the existence of life cycle stages, this research aims to determine whether the theorized pattern of organizational development can be verified empirically.

As the Internet changes our culture [Rettie 2002], online communities are quickly becoming commonplace, and corporations are increasingly turning to online communities to create valuable information [Ransbotham & Kane 2011], enabling virtual self-managing teams and embracing openness as a new strategy to improve innovation and competitiveness [Chesbrough 2003]. In such scenario, it seems important to understand whether life cycle models are applicable to online communities or whether there are other new patterns that can be uncovered that might help us better understand how these organizations develop over time. This study attempts to do this through the application of functional data analysis to the study of open source software development communities (OSSDC), such as the ones that have created the Apache web server, the Linux and Android operating systems, and quite numerous other applications.

An important distinction of this study is the attention to activity, in addition to the classic view of effectiveness patterns over time. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed


An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.