Disagreement and Conflict

Nearly every loop has this question, usually as "tell me about a time you disagreed with a colleague" or "with your manager". The interviewer is testing two things at once: whether you hold a position under pressure, and whether disagreeing with you is expensive.

Pick well. A disagreement with a real technical or product cost — architecture, scope, quality bar, a launch decision. Not a personality clash, not a story where the other person is stupid. Interpersonal-conflict stories rarely land well and often say more about you than intended. A disagreement with someone senior to you is stronger than one with a peer, because it shows you'll speak up when there's a cost to doing so.

The structure that works: what the disagreement was and why it mattered, what their position was stated fairly — this is the part interviewers listen hardest for — what you did to resolve it rather than escalate it, how it ended, and what happened afterwards.

Two endings are strong. One: you were right, and you got there with evidence rather than volume — a prototype, a benchmark, a shared doc — and the relationship survived. Two: you were wrong, you changed your mind on the evidence, and you say so plainly. Candidates avoid the second and it's often the better story, because "I updated" is rarer and more valuable than "I prevailed".

A third ending is fine and underused: you disagreed, lost, and committed anyway — executed the other approach properly rather than half-heartedly, and said what you'd watch for. Companies with a "disagree and commit" norm are explicitly looking for this, and it's the most senior of the three when the decision genuinely wasn't yours.

Watch the register. Describing the other person dismissively — "they didn't understand the system" — is the fastest way to fail this question even when it's true.

At a small company you'll disagree with the founder about scope within a month. They're deciding whether that will be a productive conversation or a tax. It's also the question where candidates most often reveal how they talk about ex-colleagues, which is heavily weighted and rarely mentioned.

• Choose a disagreement with a technical or product cost, not an interpersonal one.
• Disagreeing upward is a stronger story than disagreeing sideways.
• State the other position fairly and strongly; a strawman fails the question outright.
• Resolve with evidence — prototype, benchmark, data, a written comparison — rather than escalation or repetition.
• "I was wrong and changed my mind" is a strong ending, not a weak one.
• "I lost, committed fully, and named what I'd watch for" is the most senior ending when the call wasn't yours.
• Never describe the other person as incompetent; how you talk about them is itself the signal.
• Say what the relationship looked like afterwards — that's the part that predicts working with you.