Tell me about a time a stalled debate was decided against you and you had to deliver it.
answer
- where you stood, stated once
- who made the call and why
- the sentence where you committed
- delivered fully, no hedging
- outcome plus what you watched
basics
~20 sTests whether you can commit to a decision you argued against without sulking or sandbagging. Answer with your position stated once, the moment you got behind the call publicly, and evidence you delivered it well.
how to answer
5 beats- the call and where you stoodOne or two sentences on what was being decided and what you argued for, with the reason compressed to a clause. Keep this short — roughly fifteen percent of the answer — because the story is about what came after, not about who was right.
- how the decision got madeSay who owned the call and, in their words, why they chose what they chose. Stating the winning argument fairly is the strongest single proof that you engaged with it rather than waiting it out.
- the sentence you said out loud, and to whomDescribe what you said and who heard it. Commitment that only ever happened in your head is invisible to an interviewer; a sentence spoken in front of the team or the decision owner is the evidence they are listening for.
- how you delivered itThis is the bulk of the answer, around half your airtime. Show pace and quality on an approach you did not pick, and mention any safeguard you attached — a rollback, a budget, a checkpoint — that took your concern seriously without reopening the decision.
- outcome and honest reflectionGive the result with one number, then a short reflection that neither gloats nor grovels. If you were right, say what the checkpoint caught; if you were wrong, say what you had underweighted.
your answer
5 story prompts- Pick a decision you genuinely argued against, then implemented — not one you were neutral on.
- Be able to state the winning argument fairly, in one sentence, on their behalf.
- Recall what you said out loud when you committed, and who was in the room.
- Bring one number showing the work shipped at full quality, not grudgingly.
- This can be your escalation story continued past the moment the call was made.
draft and rehearse your own answer in a learn session
go deeper
This probes commitment and professional maturity after a decision goes against you. Interviewers want evidence that you separate advocacy from execution: that you argued honestly, accepted a legitimate owner's call, delivered the disfavoured approach at full quality, and kept your dissent from leaking into the team's morale or the work itself.
Our reports section needed charts, and I argued for extending the handful of chart components we already had rather than pulling in a full charting package. The other engineer wanted the package. We went round for most of a week, the reports work sat still, and we took it to the platform lead, who owned that call. She chose the package — her reasoning was that we had two people on the whole surface and no business being in the charting business. I made my case one last time in that meeting, which was that the package was 214 KB gzipped and our per-route budget was tighter than that. When the call went the other way I said, in front of both of them, that I thought it was the wrong trade but that I would own making it work. Then I made it work properly. I route-split the reports section so the package never loads anywhere else, which kept the entry bundle flat and put the whole weight on the one route that needs it. I wrote the wrapper so swapping the library later is one file rather than a rewrite. The part I am least proud of is that the teammate who had agreed with me kept reopening it in standup, and I let that run for two days before asking him to stop. We shipped reports six days behind the original date, and the budget held.
The commitment is visible because it was said out loud, and the follow-through is provable: route splitting and a swappable wrapper are what a person does when they intend the disfavoured option to succeed. The self-critical note about letting the relitigation run reads as honest rather than damaging.
We were replacing the customer dashboard and I pushed hard for shipping it route by route behind a flag. After a week of circling we escalated to our engineering lead, and the call went the other way: one cutover, because carrying two shells for a quarter was judged worse than the risk of a single switch. I said my piece once, which was that a big-bang gives you no way to trade scope for date when the date slips. Then I committed, and the part that actually mattered was how I did that in front of my own team. Everybody had heard me argue the other side; if I had hedged even slightly, three engineers would have hedged with me. So I explained the reasoning as if it were mine, because from that morning it was. What I did negotiate was two conditions that made the chosen option safer: a rollback we rehearsed rather than assumed, and a bundle-size gate in the pipeline that failed the build over budget. We needed both. The cutover slipped twelve days, and the entry bundle drifted from 690 KB to 1.02 MB before that gate forced us back down to 738 KB in the fortnight after launch. I still think phased would have been calmer. But it was a real owner's call with the trade-off stated, and my job from there was to make it the best version of itself.
The senior move is the audience: naming that hedging would have spread through the team, and then explaining the decision as his own. Negotiating a rehearsed rollback and a build gate turns dissent into a safeguard rather than a grudge. Losing that and the answer collapses to compliance.
It is enough to show you voiced your view once, understood the reasoning behind the call, and then did the work properly. Asking why the decision went the other way — genuinely, to learn the trade-off — is a strong move at this level.
Show the commitment as an action, not a feeling: what you did in the first day after the call, and how you kept the implementation quality high on an approach you did not choose. Naming a condition or safeguard you negotiated into the losing option is what lifts this answer.
People watch how you take a loss, so describe the audience: what you said to your own team about a decision you had argued against, and how you stopped the debate from leaking into standups. Add the checkpoint you set so a wrong call would surface as data instead of as an argument.
Frame it as modelling the norm: how you commit publicly, keep the dissent on record without keeping it alive, and make it safe for people to lose a call and stay invested. Talk about the review trigger you built so decisions get revisited on evidence, not on persistence.
saying these in an interview costs you the question
- Quiet sabotage — building it half-heartedly to prove the point
- Using the interview to relitigate and show you were right all along
- Telling your own team you are only doing this because you were told to
- Agreeing so instantly that you clearly never held a position
- No evidence the work actually shipped or that you owned its quality
- Blaming the decision maker for how the outcome turned out
- Did the decision turn out to be right?Answer honestly either way, briefly, and without scoring. If it turned out well, say what you had underweighted — that is the growth signal. If it went badly, describe how it surfaced and what you did next, and resist the told-you-so framing entirely; the interviewer is watching for whether you can be wrong or right gracefully.
- How did you talk about that decision with more junior people on the team?Show that you carried the decision as your own in front of them: explaining the reasoning and the trade-off rather than distancing yourself from it. Say explicitly that you did not tell them it was somebody else's idea, and mention how you handled someone who kept reopening it.
- When would you push back again instead of committing?Give a bar rather than a mood: new information the decision maker did not have, a safety, security, legal or data-loss risk, or a one-way door being treated as reversible. Say how you would raise it — to the same owner, once, in writing, with the new evidence — and note that repeating the original argument is not new information.
## Wordings and entry points This prompt comes in several close wordings: - "tell me about a time you disagreed with a decision but had to implement it" - "a time you had to support a decision you did not agree with" - and the explicit "give me an example of disagree and commit". Interviewers also arrive here as a follow-up to an escalation story — you describe taking a stalled debate to a decision owner, and they ask what happened when the call went the other way. Prepare one story that works from any of those entry points. ## The axis is not agreeableness The axis being measured is not agreeableness. Teams do not need people who fold; they need people whose **commitment is not conditional on winning**. The failure mode this question hunts for is the engineer who says the right words in the meeting and then delivers the disfavoured approach slowly, sloppily, or with a running commentary. That behaviour is corrosive precisely because it is deniable, so interviewers look for positive evidence of the opposite: speed after the call, quality of the implementation, and how you spoke about the decision to people who were not in the room. ## A strong answer has a distinct shape 1. **First, a position**, stated once and briefly — if you cannot say what you argued for in a sentence, the story has no tension and the commitment costs nothing. 2. **Second, the decision**: who made it, and, importantly, the reasoning you can now articulate on their behalf. Being able to state the winning argument fairly is the single clearest signal that you engaged rather than sulked. 3. **Third, the moment of commitment**, ideally something you said out loud in front of other people. 4. **Fourth, the delivery** — what you built, how well, how fast. 5. **Fifth, an honest note on cost or risk**, and any safeguard you attached to the losing option. ## The safeguard That safeguard is the most underused move in this answer. Committing does not mean pretending the risk you saw disappeared. Negotiating a rollback, a budget gate, a checkpoint date, or a metric you would watch converts your dissent into an **early-warning system** instead of a grudge. It also gives the story a natural ending, because you can say what the checkpoint showed. ## Beware two traps 1. The first is **relitigating**: candidates who were eventually proved right often cannot resist landing that punch, and it costs them more than the vindication is worth. Deliver the outcome flatly, in one sentence, whichever way it went. 2. The second is the **empty version**, where the candidate had no real view — "I did not mind either way, so I just did it". That answers a different question and reads as low investment. The commitment only counts if something was given up. ## The level shift is about audience Mid-level answers are judged on your own execution. Senior and above are judged on the people watching you: - whether the team learned that losing a call is survivable, - whether you shut down the litigation in standups, - and whether you made it possible to revisit the decision on evidence later. If you lead people, say what you said to them; that sentence is usually the whole answer.