Project Manager Interview Questions

Not your career. One project, current or last, in ninety seconds: what it delivers, who pays for it, how many people, where it stands against plan.

The size numbers do more work than the narrative. A recruiter screening for a forty-million programme is checking whether you have handled a budget with that many zeros, and "a large enterprise project" tells them nothing at all. Say the contract value, the headcount and the duration.

If the honest answer is that it is late, say it is late. You will get to explain, and a candidate whose projects have all been healthy is describing a career that does not exist.

The trap here is the true believer. Say what you actually run, and if what you actually run is a hybrid — which is what delivery tends to look like once it meets a procurement department — say that instead of picking a camp.

Then name an artefact. "Two-week sprints, a scope baseline that only moves through a signed change, and a burn-up the sponsor reads" is an answer. "Agile" is a word.

If yes, say which and when. If no, don't apologise — give one project and its numbers instead. Certification gets a CV past a filter; it has never once finished a project.

The number you controlled, not the number on the programme. Inflating it is the fastest way to be caught, because the follow-up is about your approval limit and the two do not come apart cleanly.

The real date. The second half is a free point: saying you would not walk out mid-cutover costs you a fortnight and buys you a reputation before you have started.

They are not asking whether you have delivered late; take that as read. They are asking when you told somebody.

The shape: what the plan said, when the first signal appeared, what you did in the week after that signal, who you told and on what day, and what the project delivered in the end. The middle of that list is the entire answer, and it is the part candidates skip on their way to the recovery story.

Here are two versions of the same six-week overrun, because the first one is what people give and it sounds perfectly reasonable.

The wrong answer is the one where you win. A senior interviewer has met the executive who could not be managed and has worked around them; that executive does not get charmed into compliance, and claiming you did reads as fiction.

What they want: you found where the decision actually sat, you got the objection said out loud instead of absorbing it, and you either got a decision or made the cost of not deciding visible. Putting that cost in days and money is the whole move.

A test of whether you are a project manager or a project secretary.

Don't stage it as a confrontation. The professional version is arithmetic: you took the request, priced it in days and money against the current baseline, and handed back the trade. "Yes, and here is what moves" is a no nobody can argue with.

If you have genuinely never said no to a sponsor, say so and say why — a twelve-person company does not have the governance to produce the situation. Inventing the story is worse than not having it.

The pull here is to answer as though you were the person's line manager. In a matrix you usually are not, and saying so is the first correct move.

Then the sequence: you worked out whether it was capability, capacity or clarity, because those are three different problems with three different fixes; you fixed what was yours to fix; and you went to their manager with a specific ask and a date rather than a complaint.

Pick a regret with a cost attached. "I'd have done more stakeholder engagement" is not a regret, it is a phrase; "I'd have killed it at the first gate and saved eleven weeks" is one.

The bar for a senior candidate is higher: the regret should be about a decision you owned, not about something someone else did to your project.

Name mechanics, not culture: the one synchronous hour you protect across time zones, what you moved into writing, how you find out somebody is stuck before stand-up rather than during it.

Re-baseline the expectations before you re-baseline the plan.

The first meeting is not a status update. It is finding out what the new sponsor thinks they are buying, which is rarely what the old one bought, and then writing down what you agreed the same day — because what is unwritten in week one is disputed in month four.

It is easy to define it and stop there. The definition is the setup, not the answer.

The longest chain of dependent activities, and therefore the chain with no float: every day lost on it is a day lost on the project. What you do with it is the point — where you put your strongest people, which delays you can absorb without telling anyone, and which ones you escalate the same afternoon.

Then get ahead of the follow-up, which is about near-critical paths. That is where projects actually get hurt: a path with four days of float and a six-day slip was a critical path that nobody was watching.

This separates people who have run the technique from people who memorised the formulas the night before, and the separation happens on the second follow-up rather than the first.

Planned value, earned value, actual cost. Cost performance index is earned over actual: under one and you are spending more than you are getting. Schedule performance index is earned over planned: under one and you are behind the plan you set.

Then say the thing the textbook leaves out. Schedule performance index lies at the end of a project — as you approach completion it drifts toward 1.0 whether or not you are late, because earned value converges on planned value by definition. Anyone who has run earned value on a long job knows this, and saying it is worth more than reciting both formulas perfectly.

Nobody wants your number. They want your method and your range.

Say which technique and why — analogous from a comparable job, parametric from a rate, bottom-up from the work breakdown, three-point where the uncertainty is genuine. Then say where the contingency sits and who owns it, because contingency hidden inside task durations is contingency that will be spent.

If you pad, admit it. A delivery manager who has been doing this a decade knows the team pads, and would rather you managed it in the open.

These three arrive as one and you will usually get one of them, not all: how you handle scope creep, what your change process looks like, and what you do when someone asks for a small addition. Prepare them together — they have one answer with three doors.

The answer is not that you say no. It is that every change gets priced in days and money against the baseline and goes to whoever owns the budget, and that the small ones are the dangerous ones: a change log full of two-day items is how a project loses a month without a single meeting about it. Carry a real threshold — what you absorb, what you log, what stops work until it is signed.

They want your sequence, not your feelings.

Confirm it is real and get the new date from the person doing the work rather than from the plan. Work out what it hits: the critical path, an external commitment, a contractual milestone. Build two or three options and price them. Then tell the sponsor, the same day, with the options attached.

Say that last part out loud. Candidates describe the analysis beautifully and forget to mention that they told anybody, which is the only part of the hour being graded.

Start honestly: a risk register is often a spreadsheet that gets opened before a steering committee and closed afterwards.

What makes one live is three things. Every risk has a named owner who is not you. Every risk has a trigger that says when it stops being a risk and becomes an issue. And the review removes items — a register that only grows is a register nobody believes.

Bring one risk that materialised and one you retired. The retirement is the more interesting story, and it is rarely the one that gets told.

Risks, assumptions, issues, dependencies. The neglected one is assumptions, and it is the one that kills projects: an unwritten assumption becomes a dependency you discover in month four.

Get the date in writing from the person who owns it, put it on your critical path, and agree what happens if it moves before it moves. A dependency you have not negotiated is a risk you have not logged.

Name them and be precise about depth. Someone who claims to have run a resource-levelled master schedule and cannot describe how they handled over-allocation is found out in a single follow-up.

If your last three projects ran on a spreadsheet and a shared board, say so. A great deal of successful delivery does, and the interviewer is trying to find out whether you understand scheduling, not whether you hold a licence.

A status report is a political document and pretending otherwise is a tell.

Say what yours contains — the decisions you need from them, the exceptions, the two numbers that matter — and how long it is. One page is a good answer. Then give your rule for going red, because a project that moves from green to red in a single step means somebody was managing the colour rather than the project.

Start with the contract. The instinct is to start with the relationship, and the contract is what will decide it.

What is the obligation, what evidence do you hold, and did you put the failure in writing at the time it happened? Then the escalation ladder inside their organisation, then the commercial lever you actually hold. If you have never held one, say the relationship was the lever and say what that cost you.

Benefits, cost, and the number you would have been held to. If somebody else wrote it, say what you would have challenged in it.

The follow-up worth preparing: what happened to the benefits after go-live. Benefits are rarely measured after the fact, and having tried — even if your honest answer is that nobody ever checked — separates you in a senior interview more than any framework will.

Handover to whoever operates it now, contract and budget closed, a named owner for the benefits, and a lessons session that produces one change somebody adopts.

"We held a retrospective" with no adopted change describes a meeting, not a closure — and interviewers who have inherited a badly closed project can tell the two apart from the way you say it.

Something specific and mildly unflattering, then the compensating strength. "Too deep in the detail for the first month, and they were right" is believable; "they'd say I'm a great communicator" is the eleventh time the interviewer has heard it this month.

Do not re-plan first. It is the reflex, and the plan is rarely the problem.

Two weeks: find out what is actually true by talking to the people doing the work rather than reading the status pack; establish who the sponsor really is and what they believe they are getting; and separate the problems that are real from the reporting that has drifted away from them.

Then re-baseline once, publicly, and never again. A manager who re-baselines twice has lost the room, and the second one is always harder than the first.