Estimating the Effort Overhead in Global Software Development
This work addresses cost estimation challenges for distributed software development teams, but it is incremental as it applies an existing method to a new context.
The paper tackled the problem of insufficient cost estimation models for global software development by developing a cost overhead model for a Spanish organization, identifying and quantifying specific cost drivers empirically.
Models for effort and cost estimation are important for distributed software development as well as for collocated software and system development. Standard cost models only insufficiently consider the characteristics of distributed development such as dissimilar abilities at the different sites or significant overhead due to remote collaboration. Therefore, explicit cost models for distributed development are needed. In this article, we present the initial development of a cost overhead model for a Spanish global software development organization. The model was developed using the CoBRA approach for cost estimation. As a result, cost drivers for the specific distributed development context were identified and their impact was quantified on an empirical basis. The article presents related work, an overview of the approach, and its application in the industrial context. Finally, we sketch the inclusion of the model in an approach for systematic task allocation and give an overview of future work.