You are the Product Manager for an e-commerce marketplace. You notice a rise in negative comments and reviews from users. How would you investigate the issue, prioritize the response, work with cross-functional teams, and decide what actions to take? ### Constraints & Assumptions - Negative comments may come from app reviews, on-site reviews, support tickets, social media, buyers, sellers, or delivery partners. - Do not assume the root cause before segmenting the problem. - Balance customer trust, marketplace integrity, seller fairness, operational feasibility, legal/comms risk, and business impact. - Include short-term mitigation and long-term prevention. ### Clarifying Questions to Ask - Where are the negative comments appearing, and which users are affected? - Did the rise begin after a product launch, policy change, logistics issue, pricing change, or seasonal event? - Are comments about product quality, delivery, checkout, returns, seller behavior, search relevance, pricing, or trust and safety? - How severe is the issue in volume, revenue impact, retention risk, and public visibility? ### Part 1 - Diagnose The Problem How would you investigate the rise in negative comments? #### What This Part Should Cover - Segmentation by source, user type, topic, product area, geography, cohort, device, seller, and time. - Quantitative analysis of volume, trend, affected GMV/orders, refund rate, contact rate, and retention risk. - Qualitative review of comments, support tickets, user sessions, and seller/operator input. ### Part 2 - Prioritize The Response How would you decide what to handle first? #### What This Part Should Cover - A severity framework such as reach, impact, confidence, urgency, reversibility, and trust risk. - Distinction between urgent incidents, high-volume friction, and lower-priority cosmetic complaints. - Clear guardrails for customer safety, payment issues, fraud, legal exposure, and marketplace trust. ### Part 3 - Work Cross-Functionally Which teams would you involve, and how would you coordinate action? #### What This Part Should Cover - Engineering, analytics, customer support, operations/fulfillment, seller teams, trust and safety, legal/comms, design, and leadership where appropriate. - One source of truth for what is known, owners, timelines, decisions, and customer communication. - A plan for both immediate mitigation and root-cause fixes. ### Part 4 - Define Actions And Metrics What actions would you take, and how would you know whether they worked? #### What This Part Should Cover - Short-term actions such as pausing a broken experiment, fixing a bug, improving communication, issuing credits/refunds, or escalating abusive/fraudulent content. - Long-term fixes based on root cause, such as catalog quality, delivery reliability, returns flow, search relevance, seller enforcement, or product education. - Metrics for negative-comment rate, CSAT/NPS, refunds, support contacts, repeat purchase, seller health, moderation accuracy, and issue-specific outcomes. ### What a Strong Answer Covers - Does not jump to one solution before diagnosing. - Shows prioritization under ambiguity and reputational risk. - Balances customer empathy with operational, seller, legal, and business constraints. - Measures both recovery and prevention. ### Follow-up Questions - What if negative comments are accurate but removing them would improve ratings? - How would you handle a public social-media escalation? - What if the root cause is a third-party seller group? - How would you decide between refunds and product fixes? - What guardrail would prevent over-moderation?