7 / 8 · Book 22 · From Ad Hoc Testing to Strategic Quality Engineering · drill: interview Q&A← prev⊞ allnext →Get the book →
1.13Question 2
Prompt: How do you convince a team that 400 well-targeted tests are better than 2,000 untargeted ones? What a strong answer should cover:
- The concept of risk-based testing and effort proportionality
- Concrete metrics: defect escape rate, cost per bug found, time-to-feedback
- Framing in terms leadership cares about (revenue risk, release velocity) Example answer:
- I would pull the defect escape data and show which areas produce the most production bugs. Then I would map the current 2,000 tests to those areas. Typically, you find that the riskiest 20% of features receive the same coverage as the least risky 20%. I would propose an experiment: for two sprints, reallocate effort toward the high-risk areas and measure whether escaped defects decrease. Data speaks louder than opinions.