A bug you wrote is in production. What are your next steps?

What's really being asked

This isn't a test of your debugging skills. It's a test of your maturity as a senior engineer. The interviewer is evaluating your sense of ownership, your communication discipline under pressure, and your ability to navigate a team process for incident response. They want to see if you can balance the technical urgency of a fix with the business need for clear information and prioritization.

The full answer

A great answer follows a clear, logical sequence. First, acknowledge and own the issue without defensiveness. Second, immediately triage the bug: reproduce it, check logs and metrics to quantify the blast radius (e.g., 1% of users, only on mobile) and severity (e.g., cosmetic UI vs. data loss). Third, communicate your initial findings clearly and concisely in the designated team channel, tagging the Product Owner and team lead. Propose a severity level. Fourth, collaborate with the PO and team on the right course of action—a hotfix, a rollback, or a scheduled fix. Fifth, execute the plan, ensuring the fix is reviewed and verified in production. Finally, ensure the issue is discussed in a blameless retro to identify process or testing improvements.

The mistakes people make

The most common red flag is the 'Lone Hero' who says, "I'd immediately start coding a fix." This response completely bypasses triage, communication, and team collaboration, creating information blackouts. Another major red flag is defensiveness or blame, such as, "The spec was ambiguous," or "QA should have caught this." Also, watch out for process extremes: the 'Process Robot' who just creates a low-priority ticket and moves on, or the 'Panicker' who declares an all-hands emergency for a minor visual bug.

What usually comes next

Expect questions that test your prioritization skills, such as, "What if the Product Owner disagrees with your severity assessment?" or "How does your response change if it's a critical data corruption bug versus a minor UI glitch?" Another common follow-up is, "What specific steps would you take to prevent this entire class of bug from happening again?" which tests your commitment to long-term solutions over short-term fixes.

A concrete example

"If a bug in my code was causing a 5% checkout failure rate, my first step is to confirm the impact in our monitoring tools. I'd then post in our #prod-incidents channel: 'Investigating checkout failures, initial impact ~5% of users, P2 severity. Will update in 15 mins.' I'd tag the PO and my EM. With the PO, I'd confirm priority. Given the revenue impact, this is likely a hotfix. I'd aim to have a fix deployed and verified within 1-2 hours, keeping the channel updated at key milestones: fix identified, PR up, deploying, and verified."