Software quality is no longer the QA team's exclusive responsibility. High-performing teams have adopted a new paradigm where everyone —developers, product managers, designers— is a guardian of quality. This fundamental shift transforms the QA role from "final inspector" to "strategic quality advisor".
The problem: QA as the final bottleneck
For decades, the industry has operated under a model where testing is a separate phase at the end of the development cycle. Developers "throw code over the wall" and QAs validate it before release.
This model creates three critical dysfunctions:
🚧 Slow deliveries
QA becomes the bottleneck. Finished code waits days or weeks for validation.
💸 Expensive bugs
Defects discovered in late testing cost 10-30x more than if they're caught in development.
😤 Reactive culture
Developers "ship and forget". Quality is "QA's problem", not the team's.
The shift: from "Quality Assurance" to "Quality Advocacy"
Modern teams adopt a radically different approach inspired by leading consultancies like Thoughtworks and the practices of high-performing tech companies:
The new paradigm
❌ Before: QA validates at the end
A separate QA team that inspects finished code before release
✅ Now: quality from the start
Quality Advisors embedded in teams, quality as a shared responsibility, continuous testing
The 5 pillars of Quality Advocacy
Drawing on frameworks from leading tech consultancies, we identify five fundamental practices that transform quality culture:
1. Empowerment through knowledge
Teams need to understand why quality practices matter, not just follow processes mechanically. The Quality Advisor acts as a continuous educator:
- Constant communication: Explaining testing approaches, tools and methods in standups, refinements and retros
- Mob testing sessions: Sessions where the whole team participates in designing test cases for a new feature
- Testing showcases: Demos of successful testing strategies to inspire adoption
- Living documentation: Test cases documented in Gherkin or BDD that serve as an executable specification
In practice: Instead of writing test cases in Jira that nobody reads, the Quality Advisor facilitates Example Mapping sessions where PMs, developers and QAs define scenarios together. The examples become automated tests that document the expected behavior.
2. Cross-functional collaboration
Quality Advisors work side by side with developers, PMs and stakeholders, not as a separate gate:
Collaborative practices
- ✓ Pair programming with developers to write tests
- ✓ Participating in story kickoffs before writing code
- ✓ PR reviews focused on testability
- ✓ Mob testing to explore new functionality
Benefits
- 🚀 Bugs detected in hours, not days
- 🧠 The whole team understands edge cases
- ⚡ More testable code from the design stage
- 💬 Continuous feedback, not just at the end
3. A metrics-driven approach
Quality Advisors track indicators that matter, aligned with the "4 key metrics" from DORA (DevOps Research and Assessment):
Key quality metrics
Stability
- • Change Failure Rate: % of deploys that cause incidents
- • MTTR (Mean Time To Restore): Time to recover from a failure
- • Production defects: Severity and root cause
- • Regression defects: Bugs in existing functionality
Speed
- • Deployment Frequency: Release frequency
- • Lead Time: Commit to production
- • Test execution time: Test suite duration
- • Automation coverage: % of automated tests
Coverage
- • Unit test coverage: % of code covered (target: >80%)
- • Integration test coverage: Critical E2E flows
- • Defect detection rate: % of bugs found per testing layer
- • API contract coverage: % of endpoints with contract tests
Culture
- • Test ownership: % of tests written by developers (not just QA)
- • Time to first test: Delay between code and test
- • PR test discussion: % of PRs with comments about tests
- • Quality retrospective items: Frequency of quality discussions
4. Build trust, not compliance
Teams adopt quality practices when they see the value, not because it's mandatory. The Quality Advisor demonstrates impact:
- Bug bash sessions: The whole company tests features before release, finding UX bugs that automated tests don't capture
- Iteration showcases: Demos showing how the testing strategy prevented incidents
- Retrospectives with data: "Thanks to contract tests, we avoided 3 breaking changes this sprint"
- Stakeholder walkthroughs: Involving internal clients in acceptance testing
5. Act as a beacon and alert system
The Quality Advisor continuously monitors the health of the system and acts as an early warning system:
Signs of quality degradation
- ⚠️ Flaky tests increasing: A sign of technical debt or poor test architecture
- ⚠️ Test suite getting slower: Indicates a need for parallelization or refactoring
- ⚠️ Coverage dropping sprint by sprint: Pressure for features without a balance with testing
- ⚠️ Rising number of hotfixes: Testing isn't catching regressions
- ⚠️ Developers avoiding writing tests: A tooling, knowledge or culture problem
But it also celebrates successes: recognizing when the team writes excellent tests, improves coverage, or prevents bugs proactively.
The modern testing pyramid
High-performing teams invest in different testing layers strategically:
🔺
Layer 1: Unit Tests (70%)
The foundation. Fast tests (<5ms), abundant, run on every commit. They cover business logic, edge cases, validations.
Tools: Jest, Vitest, pytest, JUnit | Run: Pre-commit hook + CI
🔶
Layer 2: Integration Tests (20%)
Tests that validate integration between components: APIs, databases, queues, external services (mocked).
Tools: Testcontainers, Supertest, Pact (contract testing) | Run: CI on every PR
🔸
Layer 3: E2E Tests (10%)
Tests that simulate a real user navigating the application. Only critical paths (login, checkout, signup). Slow, fragile, expensive to maintain.
Tools: Playwright, Cypress, Selenium | Run: Nightly or pre-release
Common anti-pattern: Teams that invert the pyramid (many E2E, few unit tests) suffer from slow suites, flaky tests, and low confidence. The right pyramid maximizes speed and reliability.
Success stories: Quality Advocacy in action
SaaS startup: from 0 to 85% coverage in 6 months
A B2B startup with 8 developers had no automated tests. They hired a Quality Advisor who implemented:
- Weekly testing dojos (1h pairing to write tests)
- A test coverage dashboard visible in Slack
- A PR template with a testing checklist
- Mutation testing with Stryker to validate test quality
Results: In 6 months they reached 85% unit test coverage, 60% integration coverage. They reduced production bugs by 70%. Deployment frequency went from weekly to daily.
Fintech scaleup: eliminating flaky tests
A fintech with 40 engineers had 250 E2E tests, of which 30-40 failed randomly (flaky). Every PR required 2-3 re-runs, slowing deliveries.
The QA team implemented:
- Automatic quarantine of flaky tests (they don't block the build)
- Refactoring 60% of E2E tests into integration tests (faster, more stable)
- Contract testing with Pact to validate APIs
- Visual regression testing with Percy for the UI
Results: The flaky rate dropped from 15% to <2%. Test suite execution time went from 45min to 12min. Confidence in tests increased dramatically.
Roadmap to transform into Quality Advocacy
Phase 1: Foundations (Month 1-2)
- ✓ Audit the current state: coverage, flaky tests, suite speed, production metrics
- ✓ Define a "definition of done" that includes testing (e.g. >80% coverage on new features)
- ✓ Set up metrics dashboards visible to the whole team
- ✓ Training session: testing pyramid, TDD basics, available tooling
Phase 2: Culture and practices (Month 3-4)
- ✓ Launch weekly testing dojos (mob programming of tests)
- ✓ Implement a PR checklist with testing criteria
- ✓ Introduce pair testing between QA and developers
- ✓ Monthly bug bash with the whole team (including PMs, designers)
Phase 3: Advanced automation (Month 5-6)
- ✓ Contract testing for microservices (Pact, Spring Cloud Contract)
- ✓ Visual regression testing (Percy, Chromatic)
- ✓ Performance testing in CI (Lighthouse CI, k6)
- ✓ Mutation testing to validate unit test quality
Phase 4: Continuous optimization (Ongoing)
- ✓ Refactoring slow E2E tests into integration tests
- ✓ Test suite parallelization (Cypress Parallelization, Jest sharding)
- ✓ Synthetic monitoring in production (Datadog Synthetics, Checkly)
- ✓ Quarterly retrospectives of the testing strategy
The role of the QA & Testing CoE
For organizations with multiple teams, a QA Center of Excellence scales best practices:
- Defines and maintains the standard testing framework (libraries, helpers, patterns)
- Creates and updates the testing training academy (from basic to advanced)
- Manages shared tooling (Sauce Labs, BrowserStack, Percy licenses)
- Organizes communities of practice (CoPs) where Quality Advisors share learnings
- Develops unified quality metrics at the company level
Conclusion: quality as a competitive advantage
Teams that adopt Quality Advocacy don't just reduce bugs. They build the confidence to move fast. With fast, reliable tests, developers can refactor without fear, experiment with new architectures, and deploy multiple times a day.
Quality stops being a trade-off against speed and becomes its enabler. As the Continuous Delivery mantra goes: "If it hurts, do it more often, and bring the pain forward."
In a market where the ability to iterate fast defines who wins, having a solid culture of shared quality isn't a nice-to-have. It's a critical competitive advantage.
Want to transform your testing culture without stopping deliveries?
At onext we've helped development teams adopt Quality Advocacy without sacrificing speed. Our on-demand QA & Testing CoE accompanies you from the initial audit to the implementation of a modern, sustainable testing strategy.
