Describe a mistake that caused a production incident

Why interviewers ask this

This is more technical and process-focused than the general "time you failed" question — interviewers want to see incident-response instinct: did you communicate clearly under pressure, and did the fix address the root cause or just the symptom?

What a strong answer covers

• What broke, in specific technical terms, and the actual user impact (how many users, how long, what they experienced)
• Your real-time actions during the incident — communication, mitigation, escalation — not just the eventual fix
• The root-cause analysis: was it a code bug, a process gap, a monitoring blind spot, or some combination?
• The concrete follow-up: a new alert, a new test, a new checklist item — something that would have caught this earlier next time

Common pitfalls

• No mention of user impact or how it was communicated to stakeholders, which can suggest poor incident hygiene
• A fix described only at the symptom level ("I reverted the change") without a root-cause explanation