Defect Analysis & DRE: Measure Software Quality by Phase
What is Defect Removal Efficiency (DRE)?
Defect removal efficiency measures how well a phase or the whole process removes defects: DRE = defects removed in the phase divided by (defects removed in the phase plus defects leaked to later phases), times 100%. For example, if a requirements review finds 10 defects and 2 requirement defects are later found downstream, DRE = 10 / (10 + 2) = 83.3%. Higher DRE means the phase is doing a better job of catching its own defects.
When to Use It
Use DRE and phase-based defect analysis in software quality programs to see where defects enter and where they are caught, across requirements, design, coding, unit test, integration test, system test, acceptance and post-release. It is also useful for building a defect baseline, comparing projects, and justifying added reviews or testing at weak upstream phases. A low upstream DRE signals that reviews and early testing should be strengthened.
How to Use It (Step by Step)
Enter the defects found in each phase and, optionally, defect type or module details. The tool computes per-phase and cumulative DRE, defect density (defects per KLOC or function point) and the phase distribution, then draws a Pareto chart of defects by type or module and a phase trend chart. Focus prevention investment on the few defect types or modules that account for about 80% of defects, and update the Pareto analysis each iteration to track improvement.
Key Formulas / Example
DRE = defects removed in phase divided by (defects removed in phase plus defects leaked downstream), times 100%; defect density = defects divided by size in KLOC or function points. For the whole process, overall DRE = defects found before release divided by (defects before release plus defects found after release), times 100%. Mature organizations typically hold overall DRE in the 85%-95% range or higher.