Knowledge Transfer in Software Maintenance Outsourcing: The Key Roles of Software Knowledge and Guided Learning Tasks

Software maintenance eats up the lion’s share of corporate software expenses, and many organizations attempt to reduce these costs through outsourcing and offshoring. A key challenge in these initiatives is to transfer knowledge to the new service delivery unit (a vendor or a captive center). Even though knowledge transfer plays a key role across theoretical perspectives in sourcing research (such as transaction cost economics, knowledge-based perspectives, and social perspectives), we know surprisingly little about what knowledge is most critical and through what mechanisms this knowledge is transferred in software maintenance outsourcing and offshoring. Insights from a multiple-case study of five knowledge transfers at a Swiss bank suggest that the most critical knowledge is software knowledge and that software knowledge is transferred through guided learning tasks. Software knowledge (i.e., knowledge about the application software, including its structure, functionality and behavior) is most critical because it allows engineers to cope with the cognitive burden imposed by enormous amounts of code, data, and documents. While engineers in settings of low knowledge specificity may possess sufficient software knowledge from the beginning, engineers in settings of high knowledge specificity acquire this knowledge through a series of guided learning tasks, i.e., by working on real or realistic maintenance tasks while receiving direction and task-specific information from experts. Our study adds to the emerging literature on transitions and offers important implications for the discourses on transaction cost economics and on knowledge-based perspectives in sourcing research.