In Agile, the sprint is sacred — typically 2 weeks, always time-boxed. Fitting all the necessary testing types into that window requires deliberate prioritisation and smart automation.
Testing Types Within a Sprint
1. Smoke Testing (Day 1-2)
Run smoke tests on every new build to verify the build is stable enough for deeper testing. Keep smoke suites small (15-30 minutes) and fully automated.
2. Functional / Story Testing (Throughout)
Test new user stories as soon as developers move them to "Ready for QA". Shift-left by reviewing acceptance criteria during sprint planning.
3. Integration Testing (Mid-sprint)
Verify that new features integrate correctly with existing modules and external services. API contract testing (Pact, REST Assured) fits well here.
4. Regression Testing (Day 8-10)
Run automated regression suites to ensure new changes haven't broken existing functionality. A well-maintained regression suite should run overnight in CI.
5. System Testing (Pre-release)
End-to-end flows across the entire application. These are the most expensive tests — run the critical paths manually, automate the stable ones.
6. User Acceptance Testing (Sprint Review)
Involve the Product Owner and stakeholders in validating stories against acceptance criteria during the sprint review or a dedicated UAT session.
Key Principles
- Automate relentlessly — anything that can be automated should be automated
- Shift left — involve QA in story refinement, not just execution
- Risk-based prioritisation — test the highest-risk areas first
- Definition of Done — include testing completion as a DoD criterion