Metrics question
Around 40% of reviews in Amazon are fake. As a PM in Amazon, how will you tackle the problem? Break your answer into 3 parts — 1. How will you identify a review is fake? 2. What action will you take? 3. If the number of fake reviews are decreased, what will be the impact?
Practice this question out loud. An AI interviewer asks it, follows up like a real interviewer would, and scores your answer. Type or speak.
Start a mock interview on this question · Mock interview from a job description
What this question tests
Tests a full structured answer across detection, action, and impact measurement for a trust and content-integrity problem, following the question's own three-part structure.
How to approach it
- Detection: use signals like reviewer account age, review velocity spikes, review-text similarity across products, and purchase verification status to flag likely fake reviews with a trained model.
- Detection refinement: cross-reference with known review-farm patterns, like clusters of 5-star reviews posted within minutes across unrelated products from the same account cluster.
- Action: for high-confidence fake reviews, remove them and flag the associated seller account; for medium-confidence cases, route to human review rather than auto-removal to avoid false positives.
- Action escalation: apply seller-level penalties (reduced visibility, account suspension) for sellers with repeated fake-review patterns, not just removing individual reviews.
- Impact: model the expected lift in genuine purchase conversion and reduced return rate as trust improves, since buyers currently discount ratings knowing fakes exist.
What a strong answer includes
- Follows the question's exact three-part structure explicitly, showing you can organize a complex answer the way it was asked.
- Names concrete detection signals (account age, review velocity, text similarity) instead of a vague 'use machine learning'.
- Distinguishes high-confidence auto-action from lower-confidence human review, showing awareness of false-positive risk in content moderation.
- Quantifies expected impact with an illustrative number, e.g. assume fixing this recovers 3-5% of conversion lost to review distrust, tying detection work back to business value.
Common mistakes
- Ignoring the question's explicit three-part structure and giving one blended answer instead.
- Proposing auto-removal for all flagged reviews without considering false-positive risk to legitimate sellers.
Likely follow-up questions
- How would you handle a legitimate seller wrongly flagged by the detection model?
- How would you measure whether removing fake reviews actually changed buyer trust, not just review counts?
More metrics questions
- How do you define success for Yelp reviews?Google · Metrics · Medium
- Utilization went down by 45% on app XYZ in Italy for the month of August. Give a reason why and draft a plan to fix it.Spotify · Metrics · Medium
- You launched a new signup flow to encourage new users to add more profile information. A/B test results indicate that the % of people that added more information increased by 8%. However, 7 day retention decreased by 2%. What do you do?Google · Metrics · Hard
- Define the metrics for YouTube search.Google · Metrics · Medium
- You walk in to your office and find that Google cloud subscription has gone down by 20%. What do you do as a product manager?Google · Metrics · Medium
- What would be the top six metrics for WhatsApp? Now pick one from the six. Tell me three things you would do to improve that metric.Google · Metrics · Medium
More questions from these companies
- Google PM interview questions
- Uber PM interview questions
- Walmart PM interview questions
- Flipkart PM interview questions
Learn the skill behind it
Chapters of the AI PM course that teach what this question tests.
- Chapter 9: Prove it paid off: outcomes, economics, and pricing
- Chapter 2: Data fluency: SQL, logs, and reading the truth yourself
- Chapter 14: Get the job: the AI PM interview loop