Amazon Cart Page Success Metrics: North Star, Guardrails and Dashboard Layout

You are the product manager who owns the shopping cart page on Amazon: the page where shoppers review what they have added, change quantities, remove items or save them for later, and move on to checkout. You are asked to build a dashboard that shows whether the cart is succeeding. What would you put on it, and how would you organize it? ```hint Follow one carted item Decide what the cart is for, both for the shopper and for Amazon, before you pick any numbers. Then trace a single item from the moment it is added: it can be bought, parked for later or dropped. Which of those outcomes count as success, and over what time window? ``` ### Clarifying Questions - Does the scope stop at the cart page, or should it also cover add-to-cart actions on product pages and the checkout steps after the shopper leaves the cart? - Which surfaces should be included and broken out: desktop web, mobile web and the shopping app? - Who will use the dashboard most: the cart team watching day-to-day health, or leadership reviewing the business weekly? - Should the dashboard also report on experiments running on the cart page, or only on steady-state performance? ### What a Strong Answer Covers - The cart's purpose for shoppers and for the business, turned into a north-star metric the cart team can actually move. - A metric tree or funnel from adding an item to placing the order, with an input metric for each step. - Guardrails that would catch a gain bought at the shopper's or the business's expense. - A dashboard layout: views for different audiences, segment cuts, time granularity and alerts. - Precise definitions for slippery metrics, and how the team would use the dashboard to diagnose a change. ### Follow-up Questions - Items added to the cart rise while your north-star metric falls in the same week. How would you use the dashboard to find out why? - Many shoppers leave items in the cart for days as a reminder list. How does that change the way you define and read abandonment? - A recommendations module added to the cart page raises items added but lowers checkout starts. Would you call it a success?