Learning From the Past: Leveraging Defect Data

If test improvement activities are to be successful, we must convince management that our efforts are focused on areas with significant payback opportunities. Brian Robinson reports that in his organization a data-driven approach to improvement has led management, developers, and testers to adopt new approaches and strategies. They collect data from their existing defect tracking system, source code repository, and a document management system used in development. From this data, they analyze and classify defects that impacted schedule (late phase test failures) or cost (customer failures). Each defect type is then mapped to a test phase that is responsible for finding it. This mapping helps define a test strategy for each phase of testing. At the same time, areas for test improvement become obvious. Join Brian to discover how your organization can enjoy measurable reductions in defects found in late phases of testing and a significant reduction in defects that your customers find.

• Defect analysis and classification techniques that work
• Focused test improvement based on meaningful data
• Ways to reduce defects found late in testing