Project Size and Software Maintenance Productivity: Empirical Evidence on Economies of Scale in Software Maintenance

Although appropriate sizing of software projects is a concern in software development, Information Systems managers generally do not consider maintenance project size as a potentid influence on software maintenance productivity. This view ignores potential efficiency gains which may arise from proactive renovation strategies such as batching similar maintenance requests into larger projects. In this study, we explore the relationship between project size and productivity for software rmintenance projects at a major national mass merchandising retailer. Using a non-parametric methodology called Da& Envelopment Analysis (DEA) for estimating the functional relationship between maintenance inputs and outputs, we determine the most productive scale size for a set of maintenance projects at this organization. In addition, we also employ DEA-based heuristics to test for the existence of returns to scale for the projects. Our results indicate the presence of significant scale economies in these software maintenance projects. The most productive scale size is larger than 90% of the projects included in our sample. These results imply that there may be potential to increase productivity in software maintenance at this organization by grouping smaller modification projects into larger planned releases.