Understanding Defect Leakage
How we calculate. Leakage % = (post-release defects ÷ total defects) × 100 — keep the same defect definition in both counts. The form uses the same arithmetic as the worked examples on this page. See our methodology and accuracy policy.
Real-world scenario: A typical Defect Leakage case uses defects found after release 8 and total defects found (all phases) 80. Enter the same figures below to reproduce the worked path.
What is Defect Leakage?
Defect leakage is the share of all discovered defects that slipped past earlier QA phases. Teams use it to judge whether unit, integration, and system testing are catching issues before release.
- Target: many teams aim for leakage under 5–10%, depending on product risk
- Not the same as density: leakage is a phase-escape rate; density is defects per size unit
- Count consistently: include only confirmed defects, not duplicate tickets
The Formula
Worked Example
Common Use Cases
- Release retrospectives: compare leakage sprint to sprint
- Test process audits: high leakage points to weak earlier phases
- Vendor / outsourced QA: track escapes into your production environment
Pro Tips
- Split by severity: critical escapes matter more than cosmetic ones
- Exclude rejected tickets: invalid defects inflate totals
- Pair with DRE: leakage and DRE tell complementary stories
Limitations: Defect Leakage results are estimates for learning and QA planning—not contractual SLAs. Align definitions with your org’s quality glossary before executive reporting.
FAQ
How is defect leakage different from escaped defect rate?
They are closely related. This calculator uses post-release defects ÷ total defects. Some orgs define escaped defects only as customer-found issues—use the Escaped Defect Rate page if that is your definition.
Should blocked or deferred defects count?
Usually no for leakage math. Count confirmed defects that were valid in the product; document your org’s rule so trends stay comparable.
Authoritative References
For software quality metrics definitions, consult established engineering sources:
- ISTQB — software testing glossary and practices
- ISO/IEC 25010 — systems and software quality models
- IEEE — software engineering standards