A product management interview gives you a chance to explain how you identify a user problem, choose an action, work with a team and assess whether it helped. Good preparation includes both past evidence and practice with unfamiliar cases. Reciting a framework without clarifying the problem can make an answer less useful.
The 14 questions and worked exercises below are original practice material. They are not leaked employer questions or records of successful candidates. Interview rounds differ: Atlassian's product interview handbook, for example, describes its own expectations around leadership, craft and outcomes. That process should not be treated as the format used by every company.
Product sense and research questions
1. How would you improve an application you use?
Clarify the business goal and choose a user segment. Describe the current journey, identify a specific problem, compare possible solutions and explain the tradeoff. End with measurement and a safe initial rollout. Avoid proposing a feature that already exists without first checking the current product.
Fictional case: For a job-alert app, choose a user who receives too many irrelevant alerts. Investigate why matching fails before proposing more notifications. A useful outcome might be more relevant saved opportunities, with unwanted-alert reports as a guardrail. Define how both events would be measured.
2. How would you discover the most important user problem?
Combine observed behaviour with conversations. Look for where users struggle, what they try instead and how often the problem occurs. Ask neutral questions about recent experiences rather than asking users to endorse your preferred feature. Explain which segments your research covers and which remain unrepresented.
3. What is an MVP in your proposed solution?
Identify the smallest usable intervention that tests the important assumption. State what you would include, omit and learn. “Small” does not mean ignoring privacy, accessibility or essential reliability. Explain why a prototype, manual service or limited release is appropriate for this particular risk.
4. How would you estimate market opportunity?
Define who might buy, the relevant geography and a time period. Build a transparent estimate from stated assumptions, then identify which inputs need verification. Distinguish the broad addressable opportunity from the portion you can realistically serve. An estimate is not a revenue forecast merely because it uses large numbers.
Prioritisation and execution questions
5. How would you prioritise two proposed features?
Start with strategy, user impact, constraints, dependencies and risk. A scoring framework can organise estimates, but it does not make uncertain assumptions true. Compare options on a consistent basis and explain what evidence would change your choice.
A worked RICE exercise
Intercom's RICE framework combines reach, impact, confidence and effort. Its score is (reach × impact × confidence) ÷ effort. Use a consistent time window for reach, a shared impact scale and total person-effort rather than calendar duration alone.
The following fictional estimates use quarterly reach and person-months of effort:
| Proposal | Reach | Impact | Confidence | Effort | Score |
|---|---|---|---|---|---|
| Improve alert filters | 1,000 users | 2 | 0.8 | 2 person-months | 800 |
| Add a new profile badge | 600 users | 1 | 0.5 | 1 person-month | 300 |
For filters, 1,000 × 2 × 0.8 ÷ 2 = 800. The estimates favour filters under these assumptions. A required security fix, contractual obligation or critical dependency could still take precedence. Explain how you would validate the reach and effort estimates before committing.
6. How do you handle a stakeholder's urgent request?
Clarify the underlying need and deadline. Compare it with existing commitments and involve the relevant team in assessing impact. Offer options, such as a smaller solution or a documented schedule change. Explain the decision to affected people without pretending you can promise engineering capacity unilaterally.
7. What would you do if engineering estimates double?
Understand what changed: uncertainty, dependencies, quality requirements or scope. Revisit the objective and options with engineering and design. You might reduce scope, change sequencing or move the date. Communicate the tradeoff and update assumptions rather than simply asking the team to work faster.
8. How would you plan a launch?
Describe acceptance criteria, instrumentation, support preparation, rollout scope and rollback conditions. Check that users can complete the core task, not merely that a feature is deployed. Decide who monitors the launch and how issues will be escalated. The plan should fit the product's risk and scale.
Metrics and diagnosis questions
9. How would you define success?
Connect the metric to value delivered. Specify who is eligible, the action that counts, the time window and the denominator. Include guardrails for harmful side effects. A measure such as “active users multiplied by order value” does not automatically represent revenue or a useful North Star metric.
For the fictional alert app, you might assess the percentage of eligible users who save a relevant opportunity in a week. This requires a clear definition of relevance; self-reported usefulness, later application behaviour and unwanted-alert rates can provide additional context.
10. What would you do if conversion dropped?
Check measurement and reporting first. Locate the affected step, cohort, device, geography or release. Review recent changes and external conditions. Form hypotheses and gather evidence before choosing a remedy. A coincidental release date does not by itself establish the cause.
11. How do you calculate retention or churn?
State the population and definition. Suppose a fictional cohort contains 200 accounts at the start of a month and 20 of those accounts cancel during the month. Under that stated definition, logo churn is 20 ÷ 200 × 100 = 10%. Revenue churn is a different measure. New accounts should not silently change the denominator in this exercise.
For activity retention, specify what counts as returning and the observation window. Different definitions can produce different numbers, so ask before comparing dashboards.
12. How would you assess an experiment?
Define the hypothesis, primary metric, guardrails, eligible population and allocation method. Discuss adequate evidence, duration, instrumentation and possible interference. Do not stop a test just because an early chart looks positive. If an experiment is impractical, explain what observational or qualitative evidence can and cannot establish.
Leadership and past experience questions
13. Tell me about a disagreement with a teammate.
Choose a true example. Explain the shared goal, disagreement, evidence considered, your own actions and the outcome. Show that you understood the other person's concern. Do not turn a collaborative decision into a claim that you personally controlled every function.
14. Tell me about something you shipped that did not work.
Describe the initial assumptions, signals you observed and what changed afterward. A useful answer can include an unresolved limitation. Avoid manufacturing a dramatic recovery or business result you cannot support. Freshers can use a college or volunteer project with its scope clearly labelled.
The STAR guide can help organise these stories. Keep hypothetical case reasoning separate from statements about your employment history.
A case worksheet to use during practice
Goal and constraints:
Target user and current journey:
Evidence available / assumptions needing validation:
Problem selected and why:
Options and tradeoffs:
Chosen first intervention:
Primary metric and exact definition:
Guardrails and rollout conditions:
What would change my decision:
Practise filling this in for one familiar product and one unfamiliar problem. Ask a partner to challenge the assumptions or change the budget halfway through. The purpose is to develop reasoning you can explain, not to memorise a supposedly perfect answer.
Frequently asked questions
Do I need to know every prioritisation framework?
No. Understand a few tools well enough to explain their inputs and limitations. Choose the tool that helps the decision rather than forcing every problem into the same formula.
Can I prepare without a product manager job title?
Use genuine examples of understanding a user need, coordinating work or evaluating an outcome. Explain your role accurately. A side project is not automatically equivalent to managing a commercial product.
What should I ask the interviewer?
Ask about the product's users, team responsibilities, decision process and how success in the role is assessed. Use the HR question guide for broader interview preparation and the application email guide when submitting your profile.
Editorially revised on 8 October 2026. All calculations and product scenarios are fictional exercises; no interview outcome is guaranteed.