Optimizing the Configuration of Development Teams Using EVA

Abstract

Resource constrained software development projects are likely to exceed the planned duration as well as the originally planned budget. As variances arise, the revision of team-task-assignments and a resource reallocation might be advisable to avoid a project escalation. In doing so, the development management has to consider the impact on team expertise and possible productivity losses if team members are shifted to other work packages than originally planned. This paper proposes the use of a modified Earned Value Analysis based on new resource-based performance measures, which provides information on the performance effects of adjustment alternatives. Furthermore, it will be demonstrated how these figures can support the decision which human resources should be shifted if a delay exists in one work package that belongs to the critical chain. In this regard, the idea of the knapsack problem is transformed into the challenging task of shifting resources to find the work package that should release the resources to minimize an overall performance decrease. A case study clarifies the implementation.