# Test strategy template

*Ismail 3000 kit · version 1.0 · copy it and fill it in. Text in italics is guidance to replace.*

## 1. Context

| Item | To fill in |
| --- | --- |
| Product | *Product name and a one-sentence description* |
| Users | *Who uses it, how many, on which devices* |
| Team | *Number of developers, QA, PO; release cadence* |
| Stakes | *What a bug costs: money, data, reputation, compliance* |
| Document date and version | |

## 2. Quality goals

*Three to five measurable goals, aligned with the stakes.*

- No blocking bug on the critical journeys in production.
- A regression caught before release in *X*% of cases.
- Test feedback in under *X* minutes on every pull request.

## 3. Scope

| In scope | Out of scope |
| --- | --- |
| *Applications, modules, APIs, platforms tested* | *What won't be tested, and why* |

## 4. Risk analysis

*List the features, rate likelihood and impact from 1 to 3, then multiply. Risks scoring 6 and 9 come first.*

| Feature | Likelihood (1-3) | Impact (1-3) | Score | Priority |
| --- | --- | --- | --- | --- |
| *Payment* | | | | |
| *Sign-up / login* | | | | |
| | | | | |

## 5. Critical journeys

*The journeys that must never break. They are tested on every release and automated first.*

1. *Journey 1*
2. *Journey 2*
3. *Journey 3*

## 6. Test levels and types

| Level | Who | Tools | When |
| --- | --- | --- | --- |
| Unit tests | Developers | *Jest, JUnit...* | On every commit |
| API / integration tests | Developers, QA | *Postman, Karate, Playwright API...* | On every pull request |
| End-to-end tests | QA | *Playwright, Cypress, Maestro...* | On every pull request (critical journeys), nightly (full suite) |
| Exploratory testing | QA, PO | Time-boxed sessions | Every sprint |
| Non-functional | *Who* | *k6, Lighthouse, axe...* | *Frequency* |
| AI feature testing | QA | *Scenario set, LLM judge* | On every prompt or model change |

## 7. Automation

- **What we automate first**: the critical journeys and regressions we've already hit.
- **What we don't automate**: *very unstable screens, one-off tests...*
- **CI integration**: *CI tool, triggers, target run time*.
- **Stability rule**: a flaky test is fixed or removed from the blocking suite within *X* days.

## 8. Environments and test data

| Environment | Use | Data |
| --- | --- | --- |
| *Acceptance / staging* | *End-to-end and acceptance testing* | *Anonymised data sets, test accounts* |
| | | |

## 9. Entry and exit criteria

- **A feature enters testing when**: *acceptance criteria are written, the build is deployed, unit tests are green*.
- **A version goes to production when**: *zero open blocking bugs, critical journeys green, acceptance checklist completed*.

## 10. Defect management

- Tool: *Jira, Linear, GitHub Issues...*
- Ticket template: steps, actual result, expected result, environment, screenshot.
- Severities: **Blocker** (fixed before release), **Major** (fixed within the sprint), **Minor** (prioritised in the backlog).

## 11. Metrics tracked

| Metric | Target | Frequency |
| --- | --- | --- |
| Automated critical journey coverage | *X%* | Every sprint |
| Flaky test rate | *< X%* | Every sprint |
| Test suite run time in CI | *< X min* | Every sprint |
| Bugs found in production | *Trending down* | Every month |

## 12. Roles

| Role | Responsibilities |
| --- | --- |
| *QA* | *Strategy, automation, acceptance testing, metrics* |
| *Developers* | *Unit tests, fixes, test reviews* |
| *PO* | *Acceptance criteria, functional sign-off* |

---

*This template is free to use in your projects. Find more kits and articles on AI-augmented QA at ismail3000.*