The shift from traditional waterfall development to Agile has necessitated a transformation in the quality assurance (QA) function. In waterfall models, QA would be involved only after the product was fully built, to test and report defects before release.
However, Agile teams focus on iterative development, where small releases build upon each other in rapid cycles. This calls for QA to collaborate closely with development teams in each sprint.
So how can QA managers drive the quality culture in Agile teams? What does their role entail in this new paradigm? This comprehensive guide covers all aspects.
Building a Cohesive QA Team
The first responsibility is to put together a skilled QA team capable of keeping pace with Agile sprints. Based on the type of product and technologies used, find testers with expertise in those domains. For example, testing a machine learning application requires knowledge of AI/ML concepts.
The right mix of manual, automation, performance, security and other specialties is key too. "As the saying goes – you cannot inspect quality into a product, it has to be built right from the start," says Lisa Crispin, pioneer of Agile testing practices.
Defining Quality Standards and Metrics
The old testing metrics of defect counts or test case pass percentages don‘t work well. Your definition of "done" for a user story must include objective acceptance criteria that confirm its completion as per specifications.
Set quantifiable quality gates aligned to business goals – peak concurrent users supported, minimum acceptable page load time, max failed transactions percentage and so on. These provide visibility into progress.
Enabling Collaboration Between Teams
Break down silos and foster a shared vision of success across cross-functional feature teams. As Trish Khoo, Director of Engineering in AWS, puts it, "Quality is everyone’s job: developers, testers, ops. Everyone has quality responsibility."
Encourage testers to team up with developers to guide testing practices. Charlotte Hanna, QA Lead at Shopify, notes that developers who write unit tests tend to build more testable systems.
Promote pairing testers with product managers too for writing acceptance tests, and with customers to validate requirements.
Advocating Test Automation
With iterative changes, regression risks multiply. Relying solely on repetitive manual testing severely compromises test coverage and cycle times.
As QA manager, sell the value of test automation to the development teams and higher management. Automated scripts amplify testing velocity, freeing up cycles for exploratory testing – where a tester‘s intuition and skill unearth hard-to-catch defects.
Open source tools like Selenium and experiment-friendly frameworks like Cypress have made test automation more accessible. Evaluate options to determine what works best.
Choosing the Right Testing Tools
Standardize tools for requirements traceability, test management, defect logging and test automation. With distributed teams, choose cloud-based tools providing real-time visibility and centralized control.
Carefully evaluate free trials before committing to paid tool sets. Assemble an open source toolkit where possible. For test environments, leverage cloud device labs rather than expensive device labs.
Integrate tools with popular productivity platforms like Jira, Trello, Slack etc. Reduce context switching and log ins for teams.
Promoting Shift Left Testing
In Agile teams, quality cannot be an after-thought or checkpoint at the end. Advocate starting testing from the very first sprint, parallelly with development.
Shift left testing focuses on evaluating code as it is written, preventing defects from going downstream. Unit, integration, UI and API testing act as safety nets, reducing escapes into production.
Over 75% of respondents in Capgemini’s World Quality Report state this has improved time-to-market and customer satisfaction.
Creating a Quality Culture
As QA manager, lead from the front in establishing a culture of quality. Evaluate team behaviors regarding target quality, rework, collaboration and process discipline.
Design incentive programs for meeting quality goals. Highlight positive deviance. Conduct peer reviews of approach and published best practices.
Uncover unhealthy patterns like cutting corners to meet deadlines, inadequate unit testing, tangled code and technical debt. Build quality engineering practices addressing these.
Using Exploratory Testing to Uncover the Unknowns
While automation testing brings speed, unscripted exploratory testing unlocks creativity in discovering hard to anticipate side-effects.
Put together a tiger team of seasoned QA experts to conduct intense exploratory testing sessions before major releases. Empower them to improvise tests, question assumptions and probe edge scenarios.
Techniques like boundary value analysis, state transition testing and data combination testing help explore uncharted territory. The goal is to discover the unknowns before customers do.
Reporting Quality KPIs to Stakeholders
Standards mean little unless backed by evidence of meeting them sprint after sprint. Monitoring and publishing quality KPIs builds confidence in releases.
Share test coverage relative to risk hotspots, defect removal efficiency, escaping defects and other base measures. Provide drill-downs to dev teams on areas needing attention.
Brief management and customers on quality health to instill trust that team velocity isn’t compromising product stability or user experience. Certify confidence in production-readiness.
Continuously Improving Processes
Retrospectives offer opportunity for testers and developers to review accomplishments, impediments faced and areas of improvement.
Draw up plans to plug gaps revealed through end-to-end traceability – from requirement to design to test case. Analyze types of defects injected and escaped into next stages.
As Agile teams mature their technical and soft skills, reassess workflows. Realign them to optimize for productivity, quality and flow of value to customers.
