Describe a failure and what you learned

## What a strong answer contains Interviewers typically want to see: - **Ownership:** You acknowledge your part without over-deflecting. - **Judgment:** You can diagnose root cause (not just symptoms). - **Recovery:** You acted quickly to mitigate impact. - **Learning loop:** Concrete process changes, not vague “I’ll be careful.” - **Scope & impact clarity:** Who/what was affected, how much, and for how long. --- ## Recommended structure (STAR + learning) Use **STAR** and add an explicit **Learning/Prevention** section: 1. **S (Situation):** Brief context (team, project, constraints). 2. **T (Task):** Your responsibility and success criteria. 3. **A (Action):** What you did, including the mistake. 4. **R (Result):** Impact + mitigation outcome. 5. **Learning/Prevention:** What you changed (process, tooling, communication). --- ## Picking the right failure Choose a story that is: - **Real but not catastrophic** (avoid ethics violations, security negligence, massive data loss unless you handled it exceptionally well and it’s appropriate). - **Actionable** (has a clear fix and measurable improvement). - **You had agency** (you can show ownership and change). Good categories: - Underestimating scope → missed deadline → implemented estimation / milestones. - Miscommunication with stakeholders → wrong requirements → added requirement reviews. - Quality gap → bug escaped → added tests, monitoring, canary releases. --- ## Root-cause and prevention examples (make it concrete) Instead of “I learned to communicate,” say: - Added a **design review checklist** and required sign-off from X. - Introduced **unit/integration tests** for critical paths; set coverage targets for modules. - Added **monitoring + alerting** (e.g., error rate, latency SLOs). - Used **feature flags** and staged rollouts. - Started **weekly stakeholder sync** and wrote decision docs. Include a small metric if possible: - “Reduced production incidents by ~30% over the next quarter.” - “Cut rollback rate from 5% to 1%.” --- ## Common pitfalls - Blaming others or making yourself the hero while others failed. - Choosing a failure with no clear lesson (“it just happened”). - Too much storytelling, not enough remediation. - No measurable outcome or no evidence the change stuck.