The Engineering Manager Interview: A Complete Guide

An engineering manager interview is hard to prepare for because you do the job all day and still cannot easily show it. An engineer can solve a problem in front of an interviewer. A manager's work happened in one-to-ones, planning meetings and awkward conversations months ago, and all you can bring into the room is an account of it. The loop is a test of whether that account holds up when four or five people question it from different directions.
It also asks for range that few managers have ready on any given week. Engineers moving into management prepare the technical round and arrive at the people round with opinions where stories should be. Experienced managers polish their stories and then cannot defend a data model. This guide covers the loop, how to prepare for each round, and a six-week plan. The questions themselves live in a separate post, 50 engineering manager interview questions by round, so they are not repeated here.
How the loop is put together
No two companies run the same loop, and few publish theirs, so treat this as the common shape and ask your recruiter for the real one. After a recruiter call there is usually a conversation with the hiring manager, then a loop of four to six interviews of 45 to 60 minutes:
- People management. Performance, feedback, growth, hiring, conflict, people leaving.
- Execution. Planning, prioritisation, the project that slipped, reporting upward.
- Technical. A system design discussion or a deep dive into a system your team owned. Coding is uncommon for managers and not unknown.
- Cross-functional. Working with product, design, other engineering teams and people senior to you.
- Organisation design or leadership. More common for senior manager roles: team structure, managing managers, strategy.
- Hiring manager, sometimes a skip-level. Whether your way of managing fits this team's actual problems.
Ask the recruiter three things before you prepare anything: which rounds are in the loop, whether there is coding, and what level the role is scoped at (a first-line manager of one team, or a manager of managers). The answers change the plan. Amazon's manager loop differs enough from this shape to have its own guide.
One point holds across companies. The interviewers write their feedback separately and the decision is made afterwards, so a strong people round rarely covers for a technical round where you bluffed. Treat each round as one that has to stand on its own.
Preparing for the people round
This round carries the most weight in most loops, and it is the one where preparation changes the result most.
Build the case list first. Go through the last three years and list every real case under six headings: a low performer, a high performer you grew, someone who left, a hire you made, a conflict inside the team, and a piece of feedback that was hard to give. Most managers find they have far more cases than the few they are proud of. The uncomfortable ones are the ones you will be asked about.
Write each case as a timeline, with dates. When you first noticed, what you said and when, what you agreed, what happened next, how long each step took, how it ended. Interviewers in this round have all managed people. They tell a real case from a tidy one by the middle of it: the week the feedback did not land, the plan that was missed for a reason you had not expected.
Prepare the outcome you did not want. Someone you could not keep. A performance case that ended in an exit. A promotion you argued for and lost. A manager with no such stories has either not managed for long or is choosing not to tell them, and both readings cost you.
Say what you actually said. "I gave him direct feedback" is a summary. "I told him the last two designs had come back from review with the same gap, and that at his level I needed him to catch it himself" is evidence. Rehearse the words.
Each answer should run about two minutes on the first pass, longer only when the interviewer asks for depth, leaving time for the follow-ups, which are where the round is graded. The STAR method gives the outline; spend most of the time on what you did and why.
Preparing for the execution round
The interviewer wants to know whether you deliver through a team and whether your manager would hear bad news from you early.
Bring three projects, written up on one page each: one that went well, one that slipped, and one where you cut scope or stopped. For each, know the date you committed to, the date you delivered, when you first knew there was a problem, what you told your stakeholders and when, and which mechanism you changed afterwards. The slipped project is the one that gets asked about. An answer that passes shows early detection ("I knew in week three, from the integration tests, and told my director that Friday") and a change that outlived the project.
Then prepare your numbers. Write a single page with the size of your team and how it changed, attrition and why people left, hires made, promotions, what the team shipped in the last year, on-call load and incident counts, and any delivery measure you actually use. You will not be asked for all of them. Having them to hand changes how you sound in every answer, and a manager who cannot say how many people left last year sounds like a passenger.
Have a view on process as well. How does your team decide what to work on? What do you measure, and what did you stop measuring? There is no correct method. The interviewer is checking that you chose yours on purpose and can say what it costs.
Preparing for the technical round
The bar is credibility. Can you question a design, see the risk in a migration, and represent your team's work to senior engineers without a translator? Nobody expects you to out-design a staff engineer.
Two formats are common, and you should ask which one you will get.
A system design discussion. The same kind of prompt an engineer gets, graded more on judgment and less on detail. Structure matters most: requirements and scale first, a simple design, then the trade-offs. Managers who have been away from design for a few years should do six to eight timed prompts aloud. The system design framework and one or two worked examples such as a rate limiter are enough to get the fluency back.
A deep dive into a system your team built. You explain the architecture, why it is that way, what went wrong in production and what you would change. Prepare by drawing it from memory and then asking your most senior engineer to find the holes in your explanation. If you cannot explain why your own system uses the datastore it uses, find out before the loop.
In both formats, the failure is bluffing. "I would want to see the read and write ratio before choosing; my instinct is this, and here is where I would ask my staff engineer to challenge me" is a good answer from a manager. A confident guess that falls apart under one follow-up is not.
If the recruiter says there is coding, take it seriously: two weeks of timed, medium-difficulty problems in one language, spoken aloud.
Cross-functional and organisation design rounds
The cross-functional round is about the people you cannot direct. Prepare four stories: a disagreement with a product manager about priorities, a dependency on another team that was late, a decision from above that you disagreed with and still had to carry out, and a time you said no to a senior stakeholder. The interviewer listens for two things: whether you can disagree and stay on good terms, and whether you told your team the truth about a decision you did not like without undermining it.
Organisation design appears mostly at senior manager level and above. Typical prompts, given here as examples: how would you split a team of twenty-five; when do you create a platform team; how do you handle a reorganisation you would not have chosen. Answer from something you did. "When my team reached fourteen people I split it along the checkout and payments boundary, and here is what went wrong in the first quarter" is worth more than a general principle about team size. If you have not done it, say what you watched happen and what you would do, and be clear about which is which.
A six-week plan
The plan assumes you have a job and can give five to seven hours a week. If the loop is in three weeks, do weeks 1, 2 and 6 and take the technical work at whatever depth the time allows.
| Week | Focus | What you produce |
|---|---|---|
| 1 | Inventory | Case list under six headings, the numbers page, questions sent to recruiters |
| 2 | People stories | Eight cases written as timelines and told aloud once each |
| 3 | Execution stories | Three project pages; your account of how the team plans and what it measures |
| 4 | Technical | Six timed design prompts aloud, or your own system drawn and challenged |
| 5 | Cross-functional and org | Four stakeholder stories; two organisation design answers from experience |
| 6 | Mock loop | One mock people round, one mock technical round, feedback applied, then rest |
A few notes on running it.
Start telling stories aloud in week 2. Do not wait for week 6. A written case and a spoken one are different things, and the first spoken version of every story is too long.
Get a second person in by week 4. A peer manager from another company is the best free option, because they will ask the follow-ups a colleague is too polite to ask. Where that is not available, the engineering management coaches on techinterview.coach include engineering managers and a VP of Engineering, and one session in week 4 leaves time to act on what you hear.
Keep the job in view. Week 6 should also include one hour per company: read the engineering blog and the job description again, work out what the team's problem probably is (growth, delivery, quality, a rebuilt team), and choose the two cases that speak to it.
First management role. If you are a tech lead interviewing for your first team, use the same plan with honest substitutions: mentoring in place of performance management, leading a project in place of running a team. Say plainly that it is the closest thing you have done. The guide to moving from IC to manager covers the transition itself.
Mistakes that cost offers
- Philosophy in place of cases. "I believe in radical candour" followed by no example. Every claim about how you manage needs a case behind it.
- Every story ends well. No exits, no lost promotions, no failed hire. Interviewers stop believing the successes.
- "We" in the people round. Performance and feedback stories are yours alone. Say what you did.
- Bluffing in the technical round. One invented detail found out costs more than three honest "I would check".
- Speaking ill of your current company or manager. Describe the situation and your part in it. The interviewer is imagining being your next manager.
- Not knowing the level. A manager of one team answering as though the role were director level, or the reverse. Scope your stories to the role.
Where coaching fits
A management loop is graded on delivery as much as content: whether a performance story sounds like care or like evasion is something you cannot hear in your own voice. Ten of the 40 coaches on techinterview.coach list engineering management as a specialty as of November 2026, and the engineering management page shows their backgrounds. A session with a coach who has hired managers can run the people round against your real cases and tell you how each one would be written up. For solo repetition, the practice interview runs a behavioural round by voice with an AI interviewer and gives a written recap; it is free, sign in to start.
Related guides
- 50 engineering manager interview questions by round: the question bank this guide points to, with what interviewers listen for.
- Amazon engineering manager interview: the SDM loop round by round, with practice questions.
- Staff engineer interviews: the individual contributor route at the same level, if you are weighing both.
- Director of engineering interview: the loop one level up, for managers of managers.
- Building your story bank: the method behind the case list in week 1.
- Mock interviews: how to brief the person running your week 6 rehearsal.