Amazon shopping cart: choosing a north-star metric, funnel metrics and guardrails

You are the product manager who owns the shopping cart on Amazon. Which metrics would you use to judge how well the cart is doing, and which of them matter most? ```hint Start from the cart's job The cart sits between a shopper showing intent and paying. Decide what a successful cart visit produces before picking metrics, and ask how each ratio you pick could improve without anyone being better off. ``` ### Clarifying Questions - Does the cart end at "Proceed to checkout", or do the checkout and payment pages that follow belong to the same team? - Are web and the mobile app both in scope, and do neighboring features such as Save for Later and Buy Now (which skips the cart) count? - Is this about tracking the cart's ongoing health, or about choosing the goal metric for a specific planned change? - Should Prime members and other shoppers be treated as separate segments, since shipping costs shape how each group fills a cart? ### What a Strong Answer Covers - A stated purpose for the cart and its place in the shopping funnel, with a clear scope boundary against checkout - One north-star metric with a precise definition, unit and time window, and the reasoning for it over alternatives - Supporting funnel and diagnostic metrics that explain movements in the north star - Guardrail metrics that catch gaming or harm to shoppers, sellers or Amazon - Cart-specific measurement traps such as long-lived carts, the Buy Now bypass and shopping across devices ### Follow-up Questions - Cart abandonment jumps sharply one week. How would you work out why? - A team wants to show more product recommendations inside the cart. Which goal metric and guardrails would you set for the test? - How would your metrics change for a grocery cart that shoppers rebuild every week?