Product and Decision-Making Onsite Case Prompt Set
You are a Product Manager candidate preparing for a mixed onsite loop covering strategy, data, design, architecture, analytics, execution, and learning from failure. For each prompt, state assumptions, structure the problem, justify trade-offs, and explain how you would measure success.
Constraints & Assumptions
-
Treat this as a set of independent PM interview prompts, not one combined product.
-
Use clear frameworks, but adapt them to the facts of each question.
-
Avoid unsupported market claims or fake precision; use illustrative assumptions only when clearly labeled.
-
Show decision quality: objective, users, options, trade-offs, risks, metrics, and recommendation.
Clarifying Questions to Ask
-
Which prompt should I answer fully first?
-
What company, geography, user segment, or business objective should I assume?
-
Is the interviewer expecting a product design answer, strategy answer, analytics debugging answer, or behavioral story?
-
Should I optimize for breadth across prompts or depth on one prompt?
Part 1 - Strategy and Business Analysis
Prompt: Should Google acquire iRobot, the maker of Roomba?
What This Part Should Cover
-
Strategic objective, such as smart home, robotics, data, hardware ecosystem, or distribution.
-
Market attractiveness and competitive landscape.
-
Build-versus-buy analysis.
-
Synergies, integration risks, privacy or regulatory risks, and financial logic.
-
A recommendation with conditions rather than a generic yes or no.
Part 2 - Data-Driven Decision Making
Prompt: Share a past example where you used large-scale experiments and big-data analysis to inform a critical product decision.
What This Part Should Cover
-
A STAR story with experiment design, metrics, guardrails, and decision impact.
-
How you handled noisy data, segmentation, novelty effects, and statistical uncertainty.
-
The business or customer result of the decision.
Part 3 - Product Design and User Experience
Prompt: Design a bookshelf for young children.
What This Part Should Cover
-
User segmentation: child, parent, caregiver, teacher, and buyer.
-
Safety, accessibility, durability, independence, and delight.
-
MVP product features, trade-offs, and success metrics.
-
Research and testing plan with children and caregivers.
Part 4 - End-to-End Product and Architecture Design
Prompt: Create a dog-walking marketplace app.
What This Part Should Cover
-
User needs for dog owners, walkers, support, and trust/safety teams.
-
Core marketplace flows: booking, matching, payment, tracking, messaging, reviews, and support.
-
Trust, safety, identity, insurance, and dispute handling.
-
Metrics for liquidity, reliability, retention, quality, and unit economics.
Part 5 - Technical Leadership and Execution
Prompts:
-
Describe the most technically complex project you have led and why you are proud of it.
-
For Google Maps' real-time traffic layer, what data should be collected, and how would you enhance accuracy and latency?
What This Part Should Cover
-
Your role in translating technical complexity into product decisions.
-
Trade-offs across latency, accuracy, privacy, reliability, and cost.
-
Systems thinking, cross-functional leadership, and measurable outcomes.
-
A concrete data and product plan for traffic accuracy and freshness.
Part 6 - Analytics and Performance Management
Prompts:
-
Which core metrics would you track to evaluate the health and growth of Facebook Live?
-
Messenger's daily engagement drops sharply overnight. How would you debug the issue and prioritize next steps?
What This Part Should Cover
-
Metric trees with acquisition, activation, engagement, retention, creator supply, viewer demand, quality, safety, and monetization where relevant.
-
Debugging steps that separate instrumentation issues from real user behavior changes.
-
Segmentation by platform, geography, app version, cohort, channel, and feature surface.
-
Prioritization based on severity, reversibility, user impact, and confidence.
Part 7 - Learning and Growth
Prompt: Describe the most memorable product failure you experienced, what went wrong, and what you learned.
What This Part Should Cover
-
A real failure with stakes and your contribution.
-
What assumptions were wrong and how you discovered them.
-
What changed in your process, instrumentation, or product judgment afterward.
-
Accountability without blame.
What a Strong Answer Covers
A strong answer chooses the right framework for each prompt, names assumptions, reasons from users and business goals, makes trade-offs explicit, and ends with measurable success criteria. It should show practical PM judgment rather than reciting generic frameworks.
Follow-up Questions
-
Which prompt would you prioritize if the interviewer gives only 20 minutes?
-
What assumptions most change your recommendation?
-
How would you measure whether the solution worked?
-
What would you cut from the MVP and why?
-
What is the riskiest part of your answer?