Technical program manager interview questions, with strong answers
12 questions from Compass's mock interviews, the way an interviewer asks them. Each has a strong answer, a common miss, and why one lands and the other doesn't.
Plan the program
You're the TPM for a new mobile payments feature: backend, mobile and compliance teams, 12 weeks. How do you start?
A strong answerKickoff to agree scope and owners, then map work and dependencies across the three teams, find the critical path, and set milestones every two to three weeks.
Scope, owners, dependencies, critical path, cadence. A solid opening.
A common missI'd set up a ticket board right away, so the teams can start working immediately, and track progress with a weekly status meeting.
Tools before plan. Find dependencies and owners first.
Compliance approval takes 8 weeks and gates the launch. What do you do?
A strong answerIt's the long pole, so start it in week one with a draft design, book compliance review slots now, and ask what they need early to avoid a late rejection.
Treats it as the critical path and de-risks it early.
A common missI'd launch first and get compliance approval after, since 8 weeks is too long to wait and we can fix any issues they raise later.
Skipping compliance on payments is a serious risk, not a plan.
In week 6, the backend team says they'll be two weeks late. What do you do?
A strong answerCheck if the work is on the critical path. If so, lay out options: cut scope, add people, or move the date. Get the accountable owner to decide this week.
Diagnose, options, and a fast decision with the right owner.
A common missI'd wait a couple of weeks to see if they catch up on their own before raising it, since it might sort itself out without escalation.
Waiting turns a 2-week slip into a surprise. Raise it now.
How do you tell leadership about the slip?
A strong answerOne page: what slipped and why, impact on the launch date, the options with their trade-offs, my recommendation, and the decision I need from them by when.
Bad news early, with options and a clear ask.
A common missI'd keep it vague in the status report so leadership doesn't panic, and give the full details once we know more about the new date.
Hiding a slip costs trust. Leaders need it early.
Risks and dependencies
You're moving a payments service to a new cloud region in three months. What are the top risks?
A strong answerData loss at cutover, downtime at peak, an unmapped dependency, and team capacity. I'd rank them by likelihood and impact, each with an owner.
Specific risks, ranked, with owners. That's the habit.
A common missThe main risk is that the new region is more expensive, so I'd focus on keeping the cloud bill under control during the whole move.
Cost matters, but data loss and downtime come first.
How would you reduce the risk of data loss at cutover?
A strong answerRun both regions in parallel with replication, verify the data matches, rehearse the cutover, test the rollback, and cut over at the quietest hour.
Parallel run, verification, rehearsal and rollback. Thorough.
A common missI'd trust the engineers to handle it, since data migration is a technical problem and the TPM's job is mainly to track the schedule.
TPMs own the plan for the risk, even if engineers do the work.
A team you don't manage owns a dependency and keeps missing dates. What do you do?
A strong answerAsk their lead why: priorities, capacity or an unclear ask. Agree a date and a checkpoint. If it still slips, escalate with options, telling them first.
Understand, agree, then escalate openly if needed.
A common missI'd ask my engineers to build the piece themselves, since waiting on another team is too risky and we can move faster on our own anyway.
Sometimes right, but it can duplicate work. Talk first.
It's launch day and one test is failing. How do you decide?
A strong answerUse the go/no-go criteria agreed earlier: does it block payments or data? Get the owner's read, check rollback is ready, and the accountable owner calls it.
Pre-agreed criteria and the right owner, not a hunch.
A common missI'd launch anyway, since one failing test out of hundreds is a tiny percentage, and the date matters a great deal to leadership.
One failing test can be the one that loses payments.
Technical depth
Engineering proposes a new notification service for 50 million users. What questions do you ask?
A strong answerPeak sends per second, delivery latency we need, what happens if it's down, retries and duplicates, which channels, and how we roll it out and roll back.
Throughput, latency, failure, rollout. The right risk questions.
A common missI'd let the engineers design it without questions, since they're the experts, and focus on tracking their progress against the plan.
TPMs are expected to probe designs for risk.
The design has one database that every request hits. Any concern?
A strong answerIt's a single point of failure and a bottleneck. I'd ask about replicas and failover, read caching, and what our recovery time is if it goes down.
Names the risk and the standard mitigations.
A common missNo, one database is simpler to manage than several, so it's the best design, and fewer moving parts means fewer things can go wrong.
Simple, yes, but it's a single point of failure.
Should we build our own notification service or buy one from a vendor?
A strong answerBuy unless it sets us apart. Compare cost at our volume, time to launch, reliability, privacy and lock-in, and prototype with a vendor first.
Criteria, a default, and a cheap way to learn.
A common missBuy it, because vendors are always cheaper and more reliable than anything we could build ourselves, so building is never worth it.
"Always" is the tell. Use criteria for this case.
How would you roll it out?
A strong answerShadow traffic to compare with the old system, then a 1% canary behind a flag, ramping while watching delivery and errors, with a tested rollback.
Safe, staged, measured, reversible.
A common missLaunch to all 50 million users at once on a quiet weekend, so any problems affect fewer people and the team can fix them quickly.
Everyone at once means everyone is hit by any bug.
Reading answers helps. Saying them under a little pressure helps more: Compass runs these as mock interviews and scores each turn.
Practice freeMore interview guides
- Product manager interview questions, with strong answersProduct sense, metrics and strategy questions from real product decisions, each with a strong answer, a common miss, and why one works.
- Sales and account executive interview questions, with strong answersAccount executive interview questions and role-plays on discovery, pipeline and closing, each with a strong answer, a common miss, and why.
- Case interview questions, with strong answersConsulting case interview questions on structure, profitability and market sizing, each with a strong answer, a common miss, and why one works.
- Investment banking interview questions, with strong answersInvestment banking technical questions on accounting, valuation and deals, each with a strong answer, a common miss, and why one works.