Project manager interviews focus on estimation, risk, and the uncomfortable conversations — with the team and with the client. Below are the most common questions with model answers. Middle: deeper understanding, optimization, and real-world situations.
Cut scope, add time, add people, or lower quality — and the fourth is a loan at a bad rate. Adding people to a late project usually slows it further, because they consume the time of those who know the system. Cutting scope is almost always the honest answer, and the work is deciding what with the person who owns the outcome.
By decomposing until pieces resemble something you have done, then estimating a range rather than a number — a three-point estimate makes the uncertainty explicit instead of hiding it in a single figure. A spike to reduce the unknown is often cheaper than a better guess. A single-number estimate on novel work is a commitment nobody should accept.
Make the trade visible rather than refusing: this goes in and that comes out, or the date moves. Refusal makes you an obstacle; accepting silently makes the team's commitment meaningless. If it happens every sprint, the problem is the intake process rather than the person.
Review it on a cadence and retire risks that no longer apply, otherwise it becomes a document written once and never read. Each risk needs an owner and a trigger — the observable event that means it is happening — because a mitigation nobody starts is not a mitigation. Most useful registers are short.
It is a planning input for one team's own forecasting — how much they historically complete in a sprint. It is misused the moment it becomes a target or a comparison between teams, because both are trivially gamed by inflating estimates. A rising velocity chart tells you nothing about whether more value shipped.
Come out with a small number of specific actions that have owners and appear in the next sprint's work, not a list of observations. The failure mode is the same three complaints every time — which means the retro is surfacing them and nothing downstream is acting. Then the thing to fix is the follow-through, not the format.
In a real interview you speak under pressure. Cam asks these same questions, scores every answer, and shows exactly what to fix.
Practice a Middle Project Manager interview →