Prompt and scope
Tell me about a time you disagreed with a team or leader’s decision but fully executed it after the decision was made. Explain your evidence, how you influenced the discussion, the final decision, your commitments, and the result.
This tests the boundary between principled disagreement and team alignment after a decision. Amazon describes the behavior as respectfully challenging a decision and then committing fully; MIT’s STAR guidance asks candidates to make their own actions and measurable results explicit. The story should contain a real disagreement, not “I agreed all along” presented as conflict.
What the interviewer is testing
First, did you use facts and customer impact rather than rank, emotion, or preference? Second, can you release the need to win once a decision is made and state your execution responsibility? Third, when the result is poor, do you still review and own improvements?
Questions to clarify before answering
- What level was the disagreement? Goal, approach, risk, schedule, or resource allocation?
- Who had final decision rights? You influenced the discussion but did not overrule the owner.
- What evidence did you have? Be specific about data, customer feedback, experiments, incidents, and constraints.
- What did you own after the decision? Describe proactive work, not merely “following orders.”
- Was there a safety, compliance, or ethics red line? If so, explain escalation rather than blind execution.
A 30-second answer framework
“Our team planned to launch a redesign in six weeks. Based on three user studies and historical error data, I worried that a critical flow could see a 15% failure increase. I put the risk, alternatives, and validation cost into a one-page decision memo, raised the objection in review, and answered follow-ups. The owner chose the date but accepted a small-traffic monitor and rollback threshold. I committed to instrumentation, rollout, and release communication; the failure rate rose only 3%, and we used that evidence for the next fix. In the review I acknowledged that my forecast was too pessimistic while keeping the early-validation practice.”
Step-by-step deep dive
Step 1: Choose a real story with decision tension
Choose a clear decision, verifiable disagreement, and result you could influence. Avoid a story where you only complained without evidence or where the final decision simply followed your recommendation; neither shows commitment after disagreement.
Step 2: Describe the objection with facts
Connect your view to data, customer impact, risk, or reversibility. For example, historical logs showed an 8% failure rate and the new design could merge two critical steps. Be specific without exposing confidential information or turning colleagues into opponents.
Step 3: Explain how you influenced the discussion
State what you prepared, which roles you invited, what questions you asked, and how you separated facts from assumptions. Influence is not making people surrender; it makes options, costs, and unknowns visible.
Step 4: Acknowledge final decision rights
Explain how the owner weighed time, benefit, risk, and resources, then state the decision plainly. Do not say “they ignored me” or imply a private parallel plan. If there was a safety or compliance red line, explain escalation to the authorized body.
Step 5: Turn commitment into action
After the decision, own risk-related work: add monitoring, set rollout gates, update docs, notify support, or rehearse rollback. Your actions should help the decision succeed rather than wait for it to fail.
Step 6: Track results with metrics
Define success and stop conditions in advance, such as conversion at baseline, errors below 5%, and zero high-risk customer complaints. Report business and quality outcomes plus external factors you did not control; do not select only favorable numbers.
Step 7: Own the result when it is poor
If the decision causes harm, protect users and the business first, then review the timeline, monitoring, and decision record. Identify which signals were underweighted, where your communication could have been earlier, and what you would change. “I told you so” is not a conclusion.
Step 8: Extract a transferable principle
Close with a principle: seek facts and clear disagreement before the decision, act as one team after it, and reopen discussion only at agreed evidence thresholds. This raises the story to a repeatable behavior while keeping the concrete experience central.
Trade-offs and boundaries
Trade-off 1: Persist or converge
Persist when evidence can change the risk decision and ask for a validation step; converge when major evidence is covered and decision rights are clear. Repeating the same point drains trust, while silence loses the chance to surface risk.
Trade-off 2: Public challenge or private conversation
Facts affecting a group decision belong in a shared record; personal performance or sensitive information can start privately. Either way, make the decision, owner, and next action traceable.
Trade-off 3: When to reopen the decision
Re-convene only on pre-defined signals, such as errors above 5%, a compliance change, or impact to a key customer. New evidence should change action, not become a contest over who was right.
Failure drills and evolution plan
Drill 1: The objection is heard as obstruction
Ask a colleague to restate your view and recommended action. Check that they heard risk, options, evidence, and next steps rather than only “I oppose.” Send a one-page decision record afterward.
Drill 2: Execution slips after commitment
Break your commitment into an owner, date, and metric. Check monitoring, rollout, support wording, and rollback before release. If resources are missing, state the gap immediately instead of making a cosmetic promise.
Drill 3: Results exceed expected risk
Prepare stop conditions and an incident contact. When a threshold is crossed, protect users first and then review whether the decision should change; do not wait for the next formal review.
Common mistakes and follow-ups
Mistake 1: Making the story “I proved I was right”
The interviewer evaluates judgment and collaboration, not a winner. Even when the result supports your forecast, explain how you supported the final decision.
Mistake 2: Saying only “I respected the decision”
Respect is not passive waiting. Name the monitoring, rollout, communication, or rollback work you owned.
Mistake 3: Raising an objection without evidence
“I felt risk” cannot guide a decision. Add the data source, assumptions, impact range, and validation cost.
Mistake 4: Treating a safety red line as a preference
For compliance, user safety, or integrity concerns, follow the escalation path and explain why disagreement-then-execution is insufficient.
Mistake 5: Reporting only team numbers
Use “I” for your actions and “we” for the team result. MIT’s STAR guidance emphasizes individual action and measurable results.
Mistake 6: Not saying when you would change course
Add a trigger threshold and explain how new evidence would reopen the discussion, showing both conviction and learning.