Defect removal efficiency (DRE)
The percentage of total defects found before release — the inverse of defect escape rate and a measure of testing thoroughness.
// Formula
// About this metric
Defect removal efficiency (DRE) measures what fraction of all defects in a product were caught before reaching production. The formula is: defects found pre-release divided by total defects (pre-release plus post-release), expressed as a percentage.
DRE is mathematically the inverse of defect escape rate. If your escape rate is 8%, your DRE is 92%. They carry the same information, so you should track one or the other — tracking both creates an illusion of more data than you actually have.
Industry benchmarks from SWEBOK and Capers Jones surveys suggest 85% DRE as a typical target for commercial software, with world-class performance above 95%. Safety-critical systems in aerospace, medical devices, and automotive domains are expected to operate at or above 99%.
The metric gained widespread adoption from Barry Boehm's cost-of-quality research showing that defect removal efficiency was one of the strongest predictors of total cost to deliver. High-DRE teams spend more on pre-release testing but far less on production support, yielding lower total cost.
// Calculator
🧮 Calculator
// Benchmark
Source: SWEBOK + Capers Jones industry surveys
DRE is the inverse of defect escape rate. Measure one or the other — they're the same signal.
// When to use this metric
Use DRE when you want to measure how much of your quality effort is being spent in pre-release (cheap) versus post-release (expensive) phases. It is a useful input for decisions about test investment: if DRE is 70%, you are finding 30% of defects in production, and the question becomes whether more pre-release testing investment would pay off.
Choose either DRE or defect escape rate as your primary metric — don't track both. DRE is more intuitive when framing a positive narrative ("we catch X% before release"); escape rate is more intuitive when framing a risk narrative ("X% escapes"). The underlying data is identical.
// Common pitfall
DRE can look excellent because of incomplete production defect tracking, not because testing is thorough. Before celebrating a DRE above 95%, verify that production defects are being consistently logged — check whether your support ticket volume and incident count are consistent with a near-perfect pre-release catch rate.