How to Answer "Tell me about a time you disagreed with your manager"

Mid-career software engineer

"My manager wanted to build our new search feature on a managed service; I argued for self-hosting on our existing cluster, and I made the case the way I believe it should be done — privately, in writing, with a cost model, once. He heard me out, disagreed, and chose the managed service. So I committed fully: I built the integration properly and didn't keep a quiet told-you-so folder. Eighteen months later, the honest scorecard is mixed. The managed service cost more than my model predicted — but the on-call burden I had breezily estimated at 'a few hours a month' turned out, on a sister team that did self-host, to be a part-time job. He was righter than I was, and the part I was right about mattered less than the part I was wrong about. I told him exactly that in my review. What it taught me: my technical estimates are solid; my operational-burden estimates were naive, and I weight them differently now."

Marketing manager

"My manager wanted our Q3 campaign to claim '3x productivity gains' based on one customer's best week. I disagreed — not on ambition, on evidence: one anecdote marketed as a statistic is how brands end up in the screenshot hall of shame, and in our space, buyers' engineering teams verify claims. I raised it once, directly, with the alternative ready: three customers on the record with ranges, which I could get in a week. He overruled me — launch deadline — and here's the part that matters: I shipped his version with full effort, because disagree-and-commit only counts when you lose. Six weeks later a prospect's procurement team asked for substantiation and we nearly lost a $200K deal over it. He pulled the claim himself and asked for my three-customer version. I didn't say anything; I just sent it. We're now the team that substantiates everything — and he proposes it as his own rule, which is exactly how I want it."