Translocality in Global Software Development: the Dark Side of Global Agile

Abstract

What happens when agile methods are introduced in global outsourcing set-ups? Agile methods are designed to empower IT developers in decision-making through self-managing collocated teams. We studied how agile methods were introduced into global outsourcing from the Indian IT vendor’s perspective. We explored how agile processes in global outsourcing impacts work conditions of the Indian IT developers, and were surprised to find that agile methodologies, even after 3 years of implementation, created a stressful and inflexible work environment negatively impacting their personal lives. Many of the negative aspects of work, which agile methodologies were developed to reduce, were evident in the global agile outsourcing set-up. We propose translocality to repudiate the dichotomy of global/local reminding us that methodologies and technologies must be understood as immediately localized and situated. Translocality helps us to explain why we cannot determine the impact of global agile as a methodology independent of how it unfolds at particular sites. Instead, we must attend to concrete practices of use when evaluating the impact of new methods.

Acknowledgments

The argument in this article has benefited from many of the discussions, we have had with researchers concerning the difficulties talking about issues of politics in global software outsourcing. We want to thank the Indian IT service provider, which made this study possible, by inviting us to interview their employees over a period between December 2011 and February 2014. We appreciate all the time people have provided us to discuss these issues at events such as the ISTC all-hands event in 2014 and the Dagstuhl seminar Rethinking Software Productivity in 2017. We would like to acknowledge the great feedback and discussions we have had with several scholars about the issues discussed in this article including James Herbsleb, Susanne Bødker, Gloria Mark, Morten Kyng, Lone Koefoed Hansen, Jack O’Neil, Pamela Hinds, William Scherlis, Christian Kästner, Stina Matthiasen, Irina Shklovski, Paul Dourish, Mel Gregg, Silvia Lindtner, Katie Pine, Marisa Cohn, and Manju Ahuja. Finally, we would like to thank the critical and engaging feedback from the Associate Editor and anonymous reviewers, which helped us developing our argument.

Funding

The empirical study was conducted within the project “Next Generation Technologies and Processes for Global Software Development,” financially supported by the National Council for Strategic Research, Ministry of Science, Innovation and Higher Education in Denmark (Grant no. #10-092313).