There is no universal percentage
Testing is sufficient when the available evidence provides enough confidence for the decision being made, relative to the risk of failure.
Do not decide based on test-count percentage alone.
Evaluate five factors
- Business criticality
- Change impact
- Critical flow coverage
- Open defects
- Residual uncertainty
- Integration risk
- Historical fragility
- Environment limitations
Use risk-based completion
| Question | Why it matters |
|---|---|
| Did all P1 flows pass? | Critical business risk may be protected. |
| What changed? | Changed/shared components need deeper regression. |
| What remains untested? | Unvalidated scope defines residual uncertainty. |
| What defects remain? | Known risk may outweigh execution percentage. |
AskAQA recommendation
Recommendation
Define “enough” as sufficient risk coverage for the release decision. Prioritize critical and impacted areas, document untested scope, and explicitly communicate residual risk.
What to document
Record critical coverage, incomplete areas, open defects, environment constraints, known limitations, and the rationale for the release recommendation.