Amazon: Turn Single-Product Checkouts into Multi-Category Baskets

You are a product manager at Amazon working on the shopping experience. Today many customers arrive for one particular kind of product, buy it and check out. The goal is to grow their baskets with a wider mix of different products. Explain how you would do it, possibly by combining upselling, cross-selling and other levers placed across the stages before, during and after a purchase. Then describe how you would roll the work out and how you would measure success. ### Clarifying Questions - Does "basket" mean the contents of a single order, or the range of products a customer buys over a period such as a month? - Which outcome matters most: more distinct product types per customer, higher order value, or profit? Does buying more units of the same item count? - Which marketplace, platforms (app or website) and customer segments are in scope, for example Prime members versus other shoppers? - Are there limits on discounts, or on how much can be added to the checkout flow? ### Part 1 — Choose the levers Pick the customers you would target and explain why they leave with a single kind of product today. Then propose levers for the pre-purchase, purchase and post-purchase stages, and prioritize them. ```hint Why stop at one? A shopper leaves with one item because nothing else came to mind, nothing suggested felt relevant or trustworthy, or there was no reason to buy it now. Tie each lever to one of those reasons at a specific moment in the journey. ``` #### What This Part Should Cover - A target segment and evidence-based reasons for single-product checkouts. - Levers placed at specific pre-purchase, purchase and post-purchase moments, more concrete than "show more recommendations". - Prioritization by impact, effort and risk to the core buying experience. ### Part 2 — Plan the rollout Explain how you would launch the levers you chose: what goes first, who sees it first, how exposure expands, and what would make you stop. ```hint Checkout is the riskiest page Anything placed near the buy button can lose orders as well as add items. Separate the levers you can test cheaply on a small share of traffic from the ones that need more caution. ``` #### What This Part Should Cover - Sequencing by risk, effort and dependency. - Controlled experiments with holdouts, and how exposure grows across segments and categories. - Explicit criteria to scale, iterate or roll back. ### Part 3 — Measure success Define the metrics that show customers are buying a wider range of products, and the ones that show you have not hurt customers or the business along the way. ```hint Breadth is not volume Items per order rises if someone buys three of the same thing. Find a measure of variety, and think about what could move the wrong way as baskets grow. ``` #### What This Part Should Cover - A primary metric for product variety, with supporting metrics. - How to show the lift is incremental rather than purchases pulled forward or shifted between orders. - Guardrails on conversion, returns, customer trust and margin. ### What a Strong Answer Covers - A clear interpretation of basket size as product variety, tied to a business goal. - Customer insight that explains single-product checkouts, not just a feature list. - Levers that span all three stages of the journey and fit together. - A rollout and measurement plan that proves incremental lift and protects the core experience. ### Follow-up Questions - Your cross-sell module raises distinct items per order, but orders per customer fall. What might be happening, and what do you do? - How would you keep suggestions from annoying customers who just want to buy one thing quickly? - How would the plan change for groceries compared with electronics?