{"id":810,"date":"2026-10-14T17:50:00","date_gmt":"2026-10-14T15:50:00","guid":{"rendered":"https:\/\/grindloop.ai\/blog\/?p=810"},"modified":"2026-10-11T09:28:16","modified_gmt":"2026-10-11T07:28:16","slug":"disagreed-with-your-manager-interview","status":"publish","type":"post","link":"https:\/\/grindloop.ai\/blog\/disagreed-with-your-manager-interview\/","title":{"rendered":"Why Your &#8220;Disagreed With Your Manager&#8221; Answer Stops at the Decision"},"content":{"rendered":"<p>Asked about a time you disagreed with your manager, many engineers stop at the decision. The manager chose. The candidate went along. The story ends there. That leaves out the part the question is built to score. Interviewers already assume you can state an opinion. They want to know what you did after you lost. A strong answer covers four things. Say how you disagreed, once and with evidence. Say what you built to make the decision succeed, even though you argued against it. Say whether you reopened the decision. Name the new facts that justified it. Then answer honestly who turned out to be right. &#8220;Partly&#8221; is often the true answer. It scores well when you can show both halves. This post follows one Android release decision through each of those steps. It ends with a full answer of about two minutes.<\/p>\n<h2>When you disagreed with your manager, the scoring starts after the decision<\/h2>\n<p>Amazon writes the expectation into its <a href=\"https:\/\/www.amazon.jobs\/content\/en\/our-workplace\/leadership-principles\" target=\"_blank\" rel=\"noopener\">leadership principles<\/a>. That page carries no date. &#8220;Have Backbone; Disagree and Commit&#8221; asks leaders to challenge decisions they disagree with. It then ends on one line: &#8220;Once a decision is determined, they commit wholly.&#8221; The principle has two halves. A typical answer only tells the first.<\/p>\n<p><a href=\"https:\/\/handbook.gitlab.com\/handbook\/values\/\" target=\"_blank\" rel=\"noopener\">GitLab&#8217;s public handbook<\/a> spells out the second half more concretely. Its section on this pairs disagreeing with committing and advocating. The page was last modified in August 2026. It says you have to deliver results on a decision while it stands. That holds even while you&#8217;re trying to get it changed. So commitment shows up as work. An interviewer can only score it if you describe that work.<\/p>\n<p>The Muse&#8217;s <a href=\"https:\/\/www.themuse.com\/advice\/conflict-with-boss-interview-question-answer-examples\" target=\"_blank\" rel=\"noopener\">guide to the conflict-with-your-boss question<\/a> was last updated in September 2022. It advises picking a conflict that ended well. It says nothing about reopening a decision. A positive ending is fine. But say the ending is that you were right. Then the interviewer still hasn&#8217;t seen you commit to a call you opposed.<\/p>\n<h2>A payment flow that had to launch on a partner&#8217;s date<\/h2>\n<p>Lena is an Android engineer on a food delivery app. Her team built a new checkout flow. It supports a card-linked discount from a partner bank. The bank&#8217;s campaign starts on a fixed Monday, with emails already scheduled.<\/p>\n<p>Lena wanted a staged rollout. Google Play lets you <a href=\"https:\/\/support.google.com\/googleplay\/android-developer\/answer\/6346149\" target=\"_blank\" rel=\"noopener\">release an update to a percentage of users<\/a> and raise it by hand. Her plan was 5% on Monday, then wider over the week. Her manager chose a full rollout on Monday. Users seeing the bank&#8217;s email had to find the discount in the app. A 5% rollout meant most of them wouldn&#8217;t.<\/p>\n<p>Both positions were reasonable. That makes it a good story. A disagreement where your manager was plainly wrong tells the interviewer little about you.<\/p>\n<h2>You disagreed once, with evidence and a condition<\/h2>\n<p>Start the answer with how you raised it. Lena wrote one short message before the decision. It named the risk. The new flow touched payment code that had no production traffic yet. It named the cost of a bad crash. Checkout is where the app earns money. She also said what would change her mind. If the team could switch the new flow off without a release, she&#8217;d accept a full rollout.<\/p>\n<p>In the interview, that condition shows your argument was about a risk. A plan you were defending would have no exit. The condition also gave the manager something to agree to. Our post on <a href=\"https:\/\/grindloop.ai\/blog\/why-a-good-star-answer-still-gets-you-a-silent-rejection\/\">silent rejections after a good STAR answer<\/a> covers this kind of up-front condition. Say in your answer that you raised it once, in writing, before the call. Repeated pushback after a decision reads very differently.<\/p>\n<h2>To commit, build what the losing side was worried about<\/h2>\n<p>The manager kept the full rollout and agreed to the switch. Lena then had four days. She put the new checkout behind a flag in <a href=\"https:\/\/firebase.google.com\/docs\/remote-config\" target=\"_blank\" rel=\"noopener\">Firebase Remote Config<\/a>. Firebase&#8217;s docs describe it as a way to change app behavior without publishing an update. A staged rollout would not have covered the same ground. The Play help page says halting a rollout only stops new users from getting the version. People who already updated keep it.<\/p>\n<p>She also wrote a one-page rollback note. She set a crash alert on checkout and took launch-day watch. None of that was her task. It made the decision she argued against more likely to work. That is the evidence GitLab&#8217;s line asks for. Describe it in concrete terms, because &#8220;I fully supported the decision&#8221; can&#8217;t be checked.<\/p>\n<p>Leave out anything that hedges against the decision behind your manager&#8217;s back. Shipping the flag quietly, or holding the release on your own, would undo the story. Lena&#8217;s manager approved the flag before she built it.<\/p>\n<h2>Reopen the decision only with something new<\/h2>\n<p>On launch day, the new flow crashed for users on one older Android version. About 3% of checkout sessions on that version failed. Lena&#8217;s alert fired. The on-call lead turned the flow off for that version within forty minutes. Everyone else kept the discount. The fix shipped three days later.<\/p>\n<p>At the retro, Lena raised staged rollouts again. This time she had launch data. GitLab&#8217;s handbook asks for this when you reopen a decision. Show the argument builds on the earlier conversation. Take it to the person who can change it. She proposed one rule. Payment changes get a flag and a staged rollout by default, unless a partner date applies. The manager agreed.<\/p>\n<p>Tell the interviewer what was new. Without new facts, the interviewer hears the same argument twice. If nothing new happened in your story, say you didn&#8217;t reopen it. Then say why.<\/p>\n<h2>&#8220;So who was right?&#8221; deserves a split answer<\/h2>\n<p>Expect this follow-up. Picking one side, yours or the manager&#8217;s, drops half the evidence. Lena&#8217;s honest answer is &#8220;both of us, about different things.&#8221; The date mattered. Promotion emails went out Monday. The discount worked for nearly every user. The risk was also real. The crash happened. The flag contained it.<\/p>\n<p>Name what you learned about your own position. Lena learned to put a price on her risk. &#8220;Something might break&#8221; lost to a fixed partner date. &#8220;We can&#8217;t turn it off without a release&#8221; won her the flag.<\/p>\n<h2>Lena&#8217;s answer, start to finish<\/h2>\n<p>Lena, her team and the numbers are made up for this post. Use the order and replace the details with your own decision.<\/p>\n<blockquote>\n<p>I was an Android engineer on a food delivery app. We built a new checkout flow for a discount from a partner bank. The bank&#8217;s campaign started on a fixed Monday. I wanted a staged rollout, starting at 5%. My manager wanted everyone on Monday, because the bank&#8217;s emails would send users looking for the discount.<\/p>\n<p>I sent one message before the decision. The new flow touched payment code with no production traffic yet. I offered to accept a full rollout if we could turn the flow off without a release. My manager kept the full rollout and agreed to that switch.<\/p>\n<p>So I put the flow behind a remote flag. I put together a rollback note, set a crash alert on checkout and took launch-day watch. On Monday, the flow crashed for about 3% of checkout sessions on one older Android version. The alert fired. We turned the flow off for that version in forty minutes. Everyone else kept the discount. The fix shipped three days later.<\/p>\n<p>At the retro, I brought the launch data and proposed one rule. Payment changes get a flag and a staged rollout unless a partner date applies. My manager agreed. Looking back, we were both right about different things. The date mattered. So did the risk. I also learned that pricing the risk persuades better than warning about it.<\/p>\n<\/blockquote>\n<p>The second paragraph sets the condition before the decision. That makes the third paragraph read as Lena keeping her word.<\/p>\n<h2>When the decision isn&#8217;t one you should commit to<\/h2>\n<p>Treat disagree and commit as a rule for judgment calls like Lena&#8217;s. Release timing, scope and architecture trade-offs all qualify. It shouldn&#8217;t cover a decision that breaks a law, a privacy promise or a safety rule. Say your manager tells you to ship analytics that send precise location data before users consent. Committing there is the wrong answer. Raise it with your manager first. If they hold, take it to whoever owns privacy or legal. Tell the interviewer you did both, in that order. That story belongs under a different question, about escalation. Our post on <a href=\"https:\/\/grindloop.ai\/blog\/influence-without-authority-interview-escalation\/\">influence without authority when it ends in escalation<\/a> covers how to tell it. The &#8220;disagreed with your manager&#8221; question asks something else. It asks whether you can lose a reasonable argument and still deliver. So if your only manager disagreement is an ethics case, save it for an escalation question. Use a smaller call you lost for this one. A sprint scope call or a library choice is enough to show the commit half.<\/p>\n","protected":false},"excerpt":{"rendered":"<p>Disagreed with your manager and lost? Show how you disagreed, what you built to make their call work and when you reopened it. A full Android sample answer.<\/p>\n","protected":false},"author":4,"featured_media":812,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"rank_math_title":"Why Your \"Disagreed With Your Manager\" Answer Stops at the Decision","rank_math_description":"Disagreed with your manager and lost? Show how you disagreed, what you built to make their call work and when you reopened it. A full Android sample answer.","rank_math_focus_keyword":"disagreed with your manager","footnotes":""},"categories":[10],"tags":[15,35,75,16,94,19],"class_list":["post-810","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-behavioral","tag-android","tag-behavioral-interview","tag-conflict-resolution","tag-interview-prep","tag-managing-up","tag-soft-skills"],"_links":{"self":[{"href":"https:\/\/grindloop.ai\/blog\/wp-json\/wp\/v2\/posts\/810","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/grindloop.ai\/blog\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/grindloop.ai\/blog\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/grindloop.ai\/blog\/wp-json\/wp\/v2\/users\/4"}],"replies":[{"embeddable":true,"href":"https:\/\/grindloop.ai\/blog\/wp-json\/wp\/v2\/comments?post=810"}],"version-history":[{"count":1,"href":"https:\/\/grindloop.ai\/blog\/wp-json\/wp\/v2\/posts\/810\/revisions"}],"predecessor-version":[{"id":811,"href":"https:\/\/grindloop.ai\/blog\/wp-json\/wp\/v2\/posts\/810\/revisions\/811"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/grindloop.ai\/blog\/wp-json\/wp\/v2\/media\/812"}],"wp:attachment":[{"href":"https:\/\/grindloop.ai\/blog\/wp-json\/wp\/v2\/media?parent=810"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/grindloop.ai\/blog\/wp-json\/wp\/v2\/categories?post=810"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/grindloop.ai\/blog\/wp-json\/wp\/v2\/tags?post=810"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}