Questions to Ask a VP of Engineering
Questions to ask a VP of Engineering, for an engineer or engineering manager who gets time with one: as a candidate in an interview loop, in a skip-level, or at an all-hands. The 53 questions follow the order the conversation tends to take, from how engineering is organized and where the VP sits beside the CTO, through the roadmap, budget and hiring, to reliability and technical debt, how managers and promotions are judged, and the VP's own job. The note under each says what a good or a worrying answer sounds like, or what to do with it.
Want questions from the whole vault instead? Try the random question generator.
The questions
Each question, and why to ask it
The org
How is the engineering organization structured, and why is it set up that way?
Why ask it
The VP either drew this chart or chose to keep it, so the reasoning is theirs to give. A good answer ties each choice to a problem, such as a platform group formed because five teams were building the same thing. 'It has always been this way' suggests nobody has looked at it since the company was half the size.
Where does your job end and the CTO's begin?
Why ask it
No two companies split these jobs the same way, so do not assume. You want to hear who owns technical direction, who owns people and delivery, and who an engineer goes to when the two disagree. Where there is no CTO, find out who holds the long-range technical view, because it may be nobody.
What does engineering have to deliver this year for the company to hit its plan?
Why ask it
From this seat the answer should connect to revenue, customers or cost, in words the CEO would recognize. Two or three commitments is a plan. A list of ten usually means the trade-offs have been left for the teams to find out later in the year.
What is the biggest problem in engineering that you have not solved yet?
Why ask it
Asking for the unsolved one gets past the challenges that have already been turned into slides. Listen for whether it is something the VP can act on, such as thin middle management or a slow release process, or something that sits above them, such as funding. As a candidate, ask how close the role you are discussing sits to it.
Which numbers about engineering go to the CEO or the board, and which do you watch only for yourself?
Why ask it
What gets reported upward is what the org will be steered by, whether that is delivery dates, uptime, cost or hiring against plan. The private list is often more telling: attrition on one team, how long reviews wait, how many pages went out at night. A VP with no private list is managing from the board deck.
Where do you want engineering to be in two or three years, and what has to change to get there?
Why ask it
The second half keeps this from turning into a speech. A VP who has worked it through names one or two things that must be different, such as releases that no longer need a freeze week or managers able to run several teams, and says which is already under way. A picture that is today's org with more people in it is a plan for growth and nothing else.
Who do you report to, and is engineering in the room when company strategy is set?
Why ask it
To the CEO, to a CTO, sometimes to a product or operations executive: each puts engineering a different distance from the decisions. What matters is whether engineering hears about a commitment while it can still shape it. The test is the last strategic decision and the point at which the VP was brought in.
What did you last reorganize, and what was it meant to fix?
Why ask it
A reorganization is the heaviest tool a VP has, and the reason given shows what they think structure can solve. A second one aimed at the same problem as the first is the worrying version. A VP who can say how they will know this one worked has thought past the announcement.
How would you describe the engineering culture, and which part of it are you trying to change?
Why ask it
Changing a culture is one of the few jobs that cannot be handed down from this seat, which is why the second half matters more than the first. A VP with an answer names a behavior, for instance teams that wait for permission or design reviews that turn into arguments, and what they have tried so far. No answer means either a rare org or a leader who is not looking.
Which decisions do you keep for yourself, and which belong to your directors and the teams?
Why ask it
An example of each tells you more than the principle. Senior hires, vendor contracts and priorities that cross teams commonly stay at the top. A VP who still approves a single team's architecture has a queue outside the door, and one who keeps nothing leaves open who settles a dispute between two directors.
Roadmap
How is the roadmap decided between you, the head of product and the CEO?
Why ask it
There are companies where engineering is handed a list and companies where it helps write one, and the VP knows which this is. Ask them to walk through the last planning cycle: who proposed, who cut, who had the final word. An account with no disagreement in it has probably been tidied up.
When you and your product counterpart cannot agree, who breaks the tie?
Why ask it
A director can send a dispute upward. At this level there may be only the CEO left, so an honest answer says how often that happens. Regular trips to the CEO suggest the two functions have not agreed on what they are optimizing for.
What did engineering last say no to, and how did the rest of the company take it?
Why ask it
A VP who protects the teams can name the request, who made it and what was offered in its place. If no example comes, either everything gets accepted or the refusing is done lower down without cover. The second half tells you whether saying no carries a cost here.
What happens when the CEO or a large customer wants something that is not on the plan?
Why ask it
It will happen, so what you are checking is whether there is a way to absorb it. Good answers describe a trade: a named item comes off, or a small standing team takes the escalations. 'We find a way' tends to mean the teams find evenings.
What share of engineering time is reserved for work that product did not ask for?
Why ask it
Platform work, upgrades and cleanup need capacity that somebody defended in a planning meeting, and that somebody is often the VP. A rough percentage plus the name of whoever agreed to it is a strong answer. A share that exists only as an intention is the first thing to go in a tight quarter.
Who from engineering has to agree before a date is promised to a customer or the board?
Why ask it
Trace the last outside commitment back to its start. Either a team sized the work first, or the date arrived and the scope was fitted to it afterward. The second can be lived with if scope is allowed to move, so ask whether it was.
Which project would you stop tomorrow if it were only your call?
Why ask it
Plenty of orgs carry one initiative kept alive by a sponsor outside engineering. Naming it takes some candor and shows where the VP's authority runs out. If it is the project you are being hired onto, better to hear it now.
When a priority changes mid-quarter, how do engineers find out why?
Why ask it
The what arrives fast, through a reassigned ticket or a cancelled meeting. The why needs someone senior to write it down or say it out loud. Ask how the last change was explained, then check that against what an engineer on the team remembers hearing.
What is the riskiest bet on this year's roadmap, and what is the fallback if it fails?
Why ask it
A roadmap with no risk on it is either very safe or not being described straight. The full answer has three parts: the bet, the date by which they will know, and what the teams would do instead. A missing fallback is not damning, but follow it up.
Budget and hiring
How is engineering's budget set, and what does finance compare it against?
Why ask it
Some companies peg engineering to a share of revenue, some fund it project by project, and some hand over a headcount number. Ask which applies there. It tells you what kind of argument wins money: a business case for each team, or one negotiation a year that the VP carries alone.
What is the hiring plan for the next twelve months, and how much of it is already approved?
Why ask it
Planned and approved are different words, and the gap between them is where promised teammates disappear. Ask which roles are open today. If you are a candidate joining a team that is supposed to double, write this answer down.
Which roles are hardest for you to fill, and what have you changed to fill them?
Why ask it
The roles named are where the current people are stretched. What changed, whether the pay band, the location, the interview loop or the level, shows how much room the VP has to move. Six months with nothing changed means the team is absorbing the gap.
How much of engineering's work goes to contractors or an outside firm, and is that share growing?
Why ask it
Whether to staff with employees, contractors or an agency is settled at this level, usually with finance. Notice which work goes outside: a bounded project with an end date, or the core product with a changing cast. A growing share can mean flexibility or a headcount cap under another name, and how contractors are engaged depends on the employer and the country, so let the VP describe their own setup.
Where are your engineers based, and will new roles open in the same places?
Why ask it
Where engineering hires is decided at this level, and it moves: a new hub, a push toward one time zone, a changed remote policy. Get the plan as well as the present. At a site that is not where the growth is, the follow-up is what that has meant for projects and promotions among the people already there.
What is your rule for building in-house versus buying, and when did you last break it?
Why ask it
Most leaders can recite the rule: build what sets the product apart and buy the rest. The exception is where you learn something, such as a tool built because a vendor took too long to approve, or bought because nobody wanted to maintain it. For anything built in-house, find out which team is funded to keep it running.
Who owns the cloud and tooling bill, and what happened the last time it ran over?
Why ask it
A cost with an owner in each team gets managed, and one that lands only on the VP's desk gets managed by decree. The story of the last overrun shows which it is: a targeted fix, or a blanket rule that slowed everyone down.
What is your position on AI coding tools across the org, and has it changed your hiring plan?
Why ask it
Policies differ widely between employers and are still moving, so ask what is written down and which code or data may be sent to an outside service. The second half is the part only this seat can answer. A VP who expects the tools to stand in for hires should be able to say what they have measured so far.
Has engineering been through layoffs or a hiring freeze while you have been here, and how were the choices made?
Why ask it
A VP may be limited in what they can say, and that is fine. What you can still hear is the method: whole projects ended, a percentage taken from every team, or individuals chosen on performance. Each leaves a different mood behind and tells you what to expect if it happens again.
Reliability and debt
Who is accountable for reliability: each team, a central group or you?
Why ask it
All three arrangements exist, and each has its weak point. Teams that own their uptime need time set aside for it. A central group needs authority over services it did not write. If the answer is 'everyone', ask who gets the call at three in the morning.
What reliability targets has engineering committed to, and what happens when one is missed?
Why ask it
Targets may be written into customer contracts, set internally or not set at all, so ask which. The consequence is the part to listen to. Feature work pausing on a service until it recovers is a real policy. A review meeting and nothing else is a ritual.
Which system would hurt the company most if it failed tomorrow, and who is funded to work on it?
Why ask it
A VP should be able to name it without thinking. The funding half separates a known risk with a team on it from a known risk everyone is hoping about. One expert and no team behind them is the usual weak spot, so find out what happens when that person goes on vacation.
How do you make the case for technical debt to people who do not write code?
Why ask it
A director schedules the cleanup. The VP has to win the time for it from a CEO and a finance lead who cannot see the problem. Strong answers translate it: slower launches, an outage that cost a customer, engineers who left out of frustration. 'They trust us' is a thin answer, so ask what happened in the last tight quarter.
What is the largest migration or rewrite under way, and when was it first meant to finish?
Why ask it
Long migrations are ordinary, and so is slipping. The distance between the first date and the current one shows how the org estimates large work and whether anyone revisits the decision. While two systems run side by side, every team pays for both.
After a major incident, what reaches your desk, and what do you do with it?
Why ask it
Reading the written review and asking one hard question is a good sign. So is going back a quarter later to see which of the promised fixes were made. Incidents that reach the VP only when a customer calls the CEO are being handled as a reputation problem.
What is standardized across all teams, and what is each team free to choose?
Why ask it
Languages, deploy tooling, cloud services and on-call practice can each be fixed centrally or left open, and neither end is wrong. Ask what happened the last time a team wanted an exception. That shows whether the standard is a default or a rule.
Who decides when security or compliance work outranks a feature?
Why ask it
What is required depends on the industry, the customers and the country, so let them say what applies to them. An org where an audit finding automatically outranks the roadmap is predictable. One where it is negotiated every time will hand you surprises near a deadline.
Managers and growth
How do you judge whether an engineering manager is doing a good job?
Why ask it
Delivery is the easy measure, and it can hide a team that is burning out. See whether the VP also names retention, the quality of promotion cases, what engineers say in skip-levels and whether other teams like working with that one. If you manage, or plan to, this is your scorecard.
How are promotions decided across the org, and what part do you play?
Why ask it
It may be a committee, a calibration meeting or a manager's call with sign-off above, and the VP may chair it, approve the budget or see only the top levels. The more useful half is what made a case fail in the last cycle, which describes the bar better than the ladder document does.
How do you keep the bar for a level the same from one team to the next?
Why ask it
Without some comparison across teams, a senior engineer on one can be a mid-level engineer on another, and people notice. Good answers describe shared calibration, reviewers from other teams or written examples for each level. Where managers are simply left to judge, a mismatch has no way of coming to light.
How many staff and principal engineers do you have, and what do they get to decide?
Why ask it
The count shows whether the technical track is in use. The decisions show whether the titles carry weight: setting a standard, blocking a design, owning a problem that crosses teams. A title with no say in anything keeps people around for a while but is not a career path.
When you last needed a director or senior manager, did you promote from inside or hire from outside?
Why ask it
One appointment does not make a pattern, so ask about the last two or three. Promotions from inside tell managers there is somewhere to go. A run of outside hires, especially people the VP has worked with before, changes who gets heard in the leadership meeting.
Which departure in the last year did you most want to prevent, and what did it teach you?
Why ask it
Attrition figures are easy to quote and easy to forget. One named loss with its real cause shows whether the VP hears why people leave or only that they did. Listen for something that changed afterward: pay adjusted, a manager moved, a project restaffed.
How is engineering pay set here, and how often is it checked against the market?
Why ask it
Bands, equity and review timing differ by employer and by country, and a VP may not share figures. They can say who sets the bands, when they were last revised, and whether someone who has fallen behind the market can be adjusted without bringing in an outside offer. That last point outlasts any starting salary.
What do you do when you hear, from below, that a manager is failing their team?
Why ask it
Worth asking in a skip-level, because it tells you what would follow if you were the one speaking up. A careful answer covers how the report is checked, how the source is protected and how long a fix is given. If the first step is telling the manager who said it, be cautious about what you raise.
The VP
How would you describe your leadership style, and how has it had to change as the org grew?
Why ask it
Leading thirty engineers directly and leading three hundred through directors are different jobs. A VP who has made the shift can say what they gave up: reviewing designs, sitting in standups, knowing every name. A style that has never changed belongs to someone who joined at this size or is still running the smaller company.
What were you brought in to do, and how far along is it?
Why ask it
Few VPs are hired or promoted to keep things as they are. There was a brief: ship faster, grow the org to a certain size, steady it after a hard year. Knowing it explains decisions that otherwise look arbitrary, and a brief that is nearly complete, or has quietly been dropped, is a reason to wonder how long this VP will be the one you work for.
Where does your week go, and how much of it is spent outside engineering?
Why ask it
A large share of the week with sales, finance, customers and the board is common at this level, and it is how engineering gets represented. What you need to hear is that some time still goes to the org itself: one-on-ones with directors, incident reviews, interviewing. Ask which of those gets cancelled first.
When did an engineer last tell you something your managers had not?
Why ask it
A recent example means there is a working channel around the hierarchy, such as office hours, skip-levels or a habit of reading design documents. Ask what happened next, and whether the engineer's manager took it badly. No example at all means the layers are doing all the talking.
What have you changed your mind about since you took this job?
Why ask it
It is a gentler route to the mistakes question and often gets a fuller answer. A reversed position on remote work, team size or a platform bet shows the VP updates on evidence. Notice what changed their mind, since that is how you would persuade them too.
What is the CEO asking of you this year that the teams have not felt yet?
Why ask it
Cost targets, a new market, a sale or a funding round reach the VP months before they reach a sprint. You may get only the outline, and the outline is enough to know which way the pressure will run. At an all-hands, ask it as 'what is coming' so it does not sound like a hunt for secrets.
What did you inherit here that you would not have built yourself?
Why ask it
Unless they started the org, the VP arrived to someone else's structure, stack and managers. What they name, and whether they have changed it or made peace with it, shows how patient they are and how free a hand they were given. It is also a fair guide to what changes next.
What is the most useful thing an engineer can bring to a conversation like this one?
Why ask it
Meant for a skip-level or office hours. The reply is an instruction for next time: a problem that crosses teams, a customer story, a disagreement with the plan. It also shows whether the VP treats these meetings as listening or as broadcasting.
A year from now, what would make you say this hire changed something for the org?
Why ask it
For candidates, near the end of the slot. A VP approves headcount against a need bigger than one team's backlog, so they can say what the hire is meant to move. Something concrete, such as a second team able to ship without waiting on the first, gives you a target. If the VP does not know the role in detail, that is normal in a large org, so ask who does.
How to use your time with a VP of Engineering
Practical guidance for the conversation itself
Before you meet
Know what only this seat can answer
The VP is the person who argues for engineering's budget, agrees the roadmap with product and the CEO, and judges the managers. An engineering manager can tell you about the sprint, and a director about delivery across a few teams. Spend VP time on the organization, the money and the trade-offs, and take questions about code review or the deploy pipeline to the people who live with them.
Find out what the title means there
VP of Engineering is not one job. At a startup it can be the top engineering leader with a dozen people and no CTO above them. At a large company it can be one of several VPs, each with a few hundred engineers, under a CTO or a senior VP. Check the leadership page or ask the recruiter before you choose questions, and open with 'Where does your job end and the CTO's begin?' if you still cannot tell.
Match the questions to the setting
In an interview loop you may get ten minutes at the end, so pick three or four: one from The org, one from Roadmap, and whichever of Budget and hiring or Managers and growth bears on your decision. A skip-level has more room and suits Reliability and debt, Managers and growth and the later questions under The VP. At an all-hands you get one question in front of everyone, so choose something many people want answered, such as the hiring plan or how a change of priority gets explained.
Start from what the VP has already said
An engineering blog, a conference talk, the open job postings and, if you work there, the notes from the last all-hands all carry part of the answer. 'You said in the spring that platform work would get its own teams. How has that gone?' gets further than asking how the org is structured from nothing, and it tells the VP which version of the answer you have already heard.
In the conversation
Put both sides of the trade-off in the question
A question about priorities in general can be answered with 'we balance it'. One that names the tension is harder to smooth over: features against cleanup, a customer's date against the plan, a contractor against a hire, an outside director against a promoted one. The answer then has to say which side usually wins, and 'What did engineering last say no to?' gets you a request, a person and an outcome.
In a skip-level, bring something as well as asking
The VP is using the meeting to learn what the layers in between have not told them. Arrive with one observation from the ground, such as a dependency that keeps blocking two teams or a tool everyone works around, then ask your questions. You get better answers from someone who has just learned something from you.
At an all-hands, keep it to one sentence
No preamble and no two-part question. Pick one that helps the whole room, such as how much of the hiring plan is approved or what the CEO is asking of engineering this year. Keep questions about your own level, pay or manager for a smaller setting.
Say what you are trying to decide
A VP hears 'how were the layoffs handled?' from a candidate weighing two offers quite differently from the same words at an all-hands. Open with the decision in front of you: whether to join, whether to move into management, whether your team's project will survive the next planning round. You get a fuller answer, and when part of it is that they cannot say, take that as an answer too.
Reading the answers
Listen for the mechanism
A good answer at this level has machinery in it: a meeting that happens on a schedule, a named owner, a percentage, a date. 'We take reliability very seriously' has none. When an answer is all conviction, ask once what would happen on a Monday morning because of it.
Notice who is we and who is they
How a VP talks about product, finance and the CEO shows how the roadmap really gets made. 'We decided' and 'they wanted' describe two different seats at the table. A VP who speaks of other executives as obstacles will pass that weather down to every team.
Check it against the layers below
Put one of the same questions to a director, a manager or an engineer: what the top priorities are, or what share of time goes to cleanup. Close answers mean the message travels. Wide gaps do not mean anyone is lying. They show how much is lost between the VP's intention and a team's week, and you would be working in that gap.
Allow for what they cannot say
Financial results, planned reorganizations, an acquisition and individual departures may all be off limits, and the rules differ by company. 'I cannot talk about that yet' is a straight answer. A long reply that never arrives anywhere is the one to weigh, and it is fair to ask when the topic can be discussed.
Mistakes to avoid
Asking delivery questions at the wrong level
How long a pull request waits or how the on-call rotation runs are good questions for a manager or a director. A VP will answer them from memory or from a dashboard, and you will have spent a rare slot on something a future teammate knows better.
Using a skip-level to complain about your manager
If there is a real problem, raise it plainly and once, with an example. A meeting that is all grievance teaches the VP little about the org and something unhelpful about you. 'What do you do when you hear, from below, that a manager is failing their team?' lets you learn how a report would be handled before you make one.
Asking to sound senior
A question built from framework names and metrics you do not use yourself invites an equally polished non-answer. Ask about the thing you actually need to decide: whether to join, what to work on next, whether to move into management.
Taking the pitch for the plan
In an interview the VP is also recruiting you. Growth, new teams and a bigger platform budget may all be sincere and still unapproved. Ask which parts are funded today, and treat the rest as hopes until someone shows you an open role or a staffed team.