Skip to main content

Test Strategy

Test Strategy

Project: {{PROJECT_NAME}} Version: {{VERSION}} Date: {{DATE}} Author: {{AUTHOR}} Status: Draft | In Review | Approved Reviewers: {{REVIEWERS}}

Document History

Version Date Author Changes
0.1 {{DATE}} {{AUTHOR}} Initial draft

1. Testing Philosophy & Principles

Core Principles:

  1. Tests are first-class code — reviewed, maintained, refactored like production code
  2. Test the behavior, not the implementation — tests enable safe refactoring
  3. Fast feedback — unit tests run in seconds; blocking tests run in < 5 minutes
  4. Tests document intent — a failing test explains what broke and why it matters
  5. Shift left — find bugs as early as possible, before they reach users

Testing philosophy: {{PHILOSOPHY}}

2. Test Pyramid

pyramid
    title Test Distribution
    accTitle: Test Distribution by Layer
    "Unit Tests (70%)" : 70
    "Integration Tests (20%)" : 20
    "E2E Tests (10%)" : 10

graph TB
    subgraph E2E["E2E Tests — 10%"]
        E2E_DESC["Critical user journeys<br/>Browser / API end-to-end<br/>Slow, expensive, high-confidence"]
    end
    subgraph INT["Integration Tests — 20%"]
        INT_DESC["Service-to-service<br/>DB + cache + external APIs<br/>Medium speed, high-value"]
    end
    subgraph UNIT["Unit Tests — 70%"]
        UNIT_DESC["Functions, classes, modules<br/>Pure business logic<br/>Fast, cheap, developer-owned"]
    end

2.1 Unit Tests

Attribute Value
Scope Individual functions, classes, pure business logic
External dependencies Mocked (no real DB, network, or filesystem calls)
Framework {{UNIT_FRAMEWORK}}
Coverage target ≥ {{UNIT_COVERAGE}}% lines, ≥ {{BRANCH_COVERAGE}}% branches
Execution time Full suite < {{UNIT_TIME}} minutes
Runs on Every commit, pre-commit hook
Written by Developer who writes the feature

What to unit test:

  • Business logic and algorithms
  • Edge cases and error conditions
  • Data transformations and validations
  • Utility functions

What NOT to unit test:

  • Third-party library internals
  • Simple property getters/setters with no logic
  • Framework boilerplate

2.2 Integration Tests

Attribute Value
Scope Service interactions: DB, cache, message queues, internal APIs
External dependencies Real (test containers or shared test environment)
Framework {{INT_FRAMEWORK}}
Coverage target All service boundaries tested; ≥ {{INT_COVERAGE}}% of integration paths
Execution time Full suite < {{INT_TIME}} minutes
Runs on Every PR, blocking merge
Written by Developer who writes the integration

What to integration test:

  • Database queries and ORM mappings
  • API endpoint request/response contracts
  • Message queue publish/consume
  • Cache read/write behavior
  • Authentication middleware

2.3 E2E Tests

Attribute Value
Scope Critical user journeys through the real deployed application
External dependencies Real (staging environment)
Framework {{E2E_FRAMEWORK}}
Coverage target Top {{E2E_JOURNEY_COUNT}} critical user journeys
Execution time Full suite < {{E2E_TIME}} minutes
Runs on Post-staging deploy, pre-production gate
Written by QA + Developer collaboration

Critical journeys to E2E test:

  1. {{JOURNEY_1}}
  2. {{JOURNEY_2}}
  3. {{JOURNEY_3}}

3. Testing Tools

Type Tool Version Purpose Config File
Unit testing {{UNIT_TOOL}} {{VERSION}} Unit tests {{CONFIG_PATH}}
Test runner {{RUNNER}} {{VERSION}} Parallel execution {{CONFIG_PATH}}
Mocking {{MOCK_TOOL}} {{VERSION}} Mock external deps Built-in
Integration testing {{INT_TOOL}} {{VERSION}} Integration tests {{CONFIG_PATH}}
Test containers {{CONTAINER_TOOL}} {{VERSION}} Real DB/Redis in tests {{CONFIG_PATH}}
E2E testing {{E2E_TOOL}} {{VERSION}} Browser E2E {{CONFIG_PATH}}
Coverage {{COV_TOOL}} {{VERSION}} Coverage reports {{CONFIG_PATH}}
Performance testing {{PERF_TOOL}} {{VERSION}} Load/stress tests {{CONFIG_PATH}}
Visual regression {{VIS_TOOL}} {{VERSION}} UI screenshot diff {{CONFIG_PATH}}

4. Test Environments & Data Management

Test Environments

Test Type Environment Database External Services
Unit Local (no env) None / in-memory Mocked
Integration Local + Docker TestContainers Stubbed
E2E Staging Dedicated test DB Sandbox/test mode
Performance Staging (production-sized) Production-scale data Sandbox

Test Data Management

Approach Used For Tool Notes
Fixtures / factories Unit + integration tests {{FACTORY_TOOL}} Reset per test
Database seeding E2E tests scripts/seed.sh Reset per test run
Anonymized production data Performance tests scripts/anonymize.sh Weekly refresh
Shared test accounts E2E (cross-service) Test account vault Never modified by tests

Data cleanup policy: All test data cleaned up after test run (via teardown hooks or transaction rollback)

5. Test Automation Strategy

Test Type Automation Trigger Tooling
Unit tests 100% automated Pre-commit + CI {{UNIT_TOOL}}
Integration tests 100% automated CI on PR {{INT_TOOL}}
E2E (critical paths) 100% automated Post-staging deploy {{E2E_TOOL}}
E2E (edge cases) Partially automated Weekly scheduled run {{E2E_TOOL}}
Performance tests Automated baseline Weekly + pre-release {{PERF_TOOL}}
Security (SAST/SCA) 100% automated Every PR {{SAST_TOOL}}
Visual regression Automated On UI changes {{VIS_TOOL}}
Accessibility Automated + manual On UI changes {{A11Y_TOOL}}
Manual exploratory Manual Sprint end / pre-release —

6. Manual Testing Scope

Manual testing is required for:

  • New feature exploratory testing (first sprint of a feature)
  • Usability and UX review
  • Accessibility testing (beyond automated checks)
  • Complex multi-step business scenarios not yet automated
  • Third-party integrations with limited test environments
  • Devices/browsers outside automated matrix

Manual testing is NOT required for:

  • Regression of previously tested features (automate these)
  • CRUD operations on standard forms
  • Unit-level logic covered by automated tests

7. Code Coverage Requirements

Layer Lines Branches Functions Notes
Business logic ≥ {{BIZ_COV}}% ≥ {{BIZ_BRANCH}}% ≥ {{BIZ_FN}}% Strictly enforced
API handlers ≥ {{API_COV}}% ≥ {{API_BRANCH}}% ≥ {{API_FN}}%
Utilities ≥ {{UTIL_COV}}% ≥ {{UTIL_BRANCH}}% ≥ {{UTIL_FN}}%
UI components ≥ {{UI_COV}}% — ≥ {{UI_FN}}% Render + interaction
Overall minimum ≥ {{TOTAL_COV}}% — — CI gate

Coverage enforcement: CI pipeline fails if coverage drops below minimum Coverage report: Published to {{COV_REPORT_URL}} on every PR

8. Quality Gates

PR Merge Gate (must pass before merge)

  • All unit tests pass
  • All integration tests pass
  • Coverage ≥ minimum thresholds
  • No HIGH/CRITICAL security findings (SAST/SCA)
  • No secrets detected
  • Linting passes
  • Type checking passes (if typed language)

Staging Deploy Gate (must pass before staging deploy)

  • All PR gates passed
  • Build artifact created and signed

Production Deploy Gate (must pass before production deploy)

  • All E2E tests pass on staging
  • Performance baseline not degraded > {{PERF_REGRESSION}}%
  • Manual QA sign-off (if sprint includes new features)
  • Manual approval in CI pipeline

9. Responsibility Matrix

Test Type Writes Tests Reviews Tests Maintains Tests Signs Off
Unit tests Developer PR reviewer Developer Tech lead
Integration tests Developer QA engineer Developer Tech lead
E2E tests QA + Developer Tech lead QA engineer QA lead
Performance tests DevOps + Developer Tech lead DevOps Eng manager
Security tests DevOps (automated) Security DevOps Security lead

10. Test Reporting & Metrics

Metric Target Reporting
Test pass rate ≥ 99% (unit), ≥ 95% (E2E) CI dashboard
Flaky test rate < {{FLAKY_RATE}}% Weekly report
Test execution time < {{TOTAL_TEST_TIME}} min (full suite) CI dashboard
Coverage trend Stable or improving PR comments
Tests added per sprint ≥ {{TESTS_PER_SPRINT}} Sprint metrics

11. Continuous Testing in CI/CD

See CI/CD Pipeline for full pipeline details.

Stage Tests Run Blocking
Pre-commit (local) Unit tests, linting Yes (can be bypassed with --no-verify with justification)
PR open / update Unit + integration + SAST + SCA Yes — blocks merge
Staging deploy E2E, visual regression Yes — blocks production
Production deploy Smoke tests, monitoring check Yes — auto-rollback on failure
Scheduled (nightly) Full E2E suite, performance baseline No — alerts only

Approval

Role Name Date Signature
Author
Reviewer
Approver
Back to top