Amazon product strategy: cutting avoidable returns through root-cause analysis

You are a product manager on Amazon's quality and logistics team, and your goal is to reduce the number of customer returns. Explain the process you would follow to attack the problem, and then describe what you would build. ```hint Not all returns are alike Split returns by why they happen before proposing anything. Then ask which causes this team actually controls and which it can only influence through other teams. ``` ### Clarifying Questions - Is the target the return rate, the cost of returns, or both, and do returns a customer is entitled to (such as a change of mind) count as a problem at all? - Which goods are in scope: everything sold on the site, or only items Amazon stores and ships itself rather than third-party sellers who fulfill their own orders? - Must the current return policy stay as it is, or is making returns harder on the table? - What data does the team already have: return reason codes, free-text comments, inspection results on returned units, damage reports from warehouses and carriers? ### What a Strong Answer Covers - A clear goal with a primary metric and guardrails that protect customer trust and sales. - A structured breakdown of returns by root cause, sized by volume and cost, and mapped to what a quality and logistics team owns. - A prioritized product or system to build, including how it detects problems and what action it triggers. - An MVP and a measurement plan that handles the delay between a purchase and its return. - Risks such as noisy or gamed reason codes and shifting cost onto sellers or customers. ### Follow-up Questions - Returns marked as defective spike for one product, yet inspection finds most returned units work fine. What is happening, and what would you do? - How would you test whether your changes reduced returns when a return can arrive weeks after the purchase? - Your change lowers returns in a category but also lowers its sales. How do you decide whether to keep it?