Handling an Engineer's Proposal for a Product Recommendations Module in Online Retail

Quick Overview

An execution question about how a product manager should respond when an engineer proposes launching a related-products module on a shopping website. It tests problem framing, opportunity sizing against the roadmap, recommendation design choices, A/B test metrics with guardrails, and a staged rollout decision.

Handling an Engineer's Proposal for a Product Recommendations Module in Online Retail

Company: Asana

Role: Product Manager

Category: Product / Decision Making

Difficulty: medium

Interview Round: Other

You are the product manager for an online shopping website. An engineer on your team wants to launch a "Related products" feature that suggests other items to shoppers. Explain how you would handle the proposal: how you would decide whether it is worth doing, how the feature should work, and how you would launch it and judge whether it succeeded. ```hint Begin with the shopper problem An engineer's enthusiasm is a useful signal, not a decision. Pin down which shopper problem the module would solve, and what else on the roadmap it competes with, before you discuss algorithms or launch plans. ``` ### Clarifying Questions - Is the feature already built, prototyped, or only an idea, and how much engineering time would it take to ship and maintain? - Where would the related products appear: on product pages, in the cart, at checkout, or after purchase? - Does "related" mean similar alternatives to the item being viewed, items commonly bought alongside it, or both? - What are the site's current priorities, such as conversion, basket size or retention, and what already planned work would this delay? ### What a Strong Answer Covers - Understanding the engineer's motivation and the shopper problem before deciding, while keeping the engineer engaged whatever the outcome. - Sizing the opportunity against current priorities and the cost to build, run and maintain the feature. - Concrete product choices: placement, what counts as related, the signals used to pick items, products with no history, and quality rules such as excluding out-of-stock items. - A controlled experiment with a primary metric tied to revenue or conversion per visitor, and guardrails such as page speed and distraction from the item being viewed. - A staged rollout with a clear decision rule, a way to switch the feature off, and named owners. ### Follow-up Questions - The test shows many clicks on the module but no change in revenue per visitor. What do you conclude, and what do you do next? - How would you choose related items for a product that was added yesterday and has no browsing or purchase history? - The merchandising team wants to sell sponsored slots in the module. How do you decide?

Overview: An execution question about how a product manager should respond when an engineer proposes launching a related-products module on a shopping website. It tests problem framing, opportunity sizing against the roadmap, recommendation design choices, A/B test metrics with guardrails, and a staged rollout decision.

Asana logo
Asana
Jul 16, 2025
mediumProduct ManagerOtherProduct / Decision Making
0
0

You are the product manager for an online shopping website. An engineer on your team wants to launch a "Related products" feature that suggests other items to shoppers. Explain how you would handle the proposal: how you would decide whether it is worth doing, how the feature should work, and how you would launch it and judge whether it succeeded.

Clarifying Questions Guidance

  • Is the feature already built, prototyped, or only an idea, and how much engineering time would it take to ship and maintain?
  • Where would the related products appear: on product pages, in the cart, at checkout, or after purchase?
  • Does "related" mean similar alternatives to the item being viewed, items commonly bought alongside it, or both?
  • What are the site's current priorities, such as conversion, basket size or retention, and what already planned work would this delay?

What a Strong Answer Covers Guidance

  • Understanding the engineer's motivation and the shopper problem before deciding, while keeping the engineer engaged whatever the outcome.
  • Sizing the opportunity against current priorities and the cost to build, run and maintain the feature.
  • Concrete product choices: placement, what counts as related, the signals used to pick items, products with no history, and quality rules such as excluding out-of-stock items.
  • A controlled experiment with a primary metric tied to revenue or conversion per visitor, and guardrails such as page speed and distraction from the item being viewed.
  • A staged rollout with a clear decision rule, a way to switch the feature off, and named owners.

Follow-up Questions Guidance

  • The test shows many clicks on the module but no change in revenue per visitor. What do you conclude, and what do you do next?
  • How would you choose related items for a product that was added yesterday and has no browsing or purchase history?
  • The merchandising team wants to sell sponsored slots in the module. How do you decide?