Decide Whether to Replace Facebook Friend Suggestions With an Unbuilt Shopping Unit
Company: Meta
Role: Product Manager
Category: Product / Decision Making
Difficulty: medium
Interview Round: Other
Facebook's People You May Know unit suggests people a user might want to add as friends. Imagine a proposal to give its placement to a new unit, Products You May Buy, that would recommend things to purchase instead. The new unit has not been built. As the product manager, how would you decide whether to make this swap without first building the new feature?
```hint Two halves of one decision
The swap has a cost and a benefit. Ask which of the two you could observe without writing the new feature, and what cheap stand-in could give you a read on the other.
```
### Clarifying Questions
- Does "swap" mean removing People You May Know from the product entirely, or only moving it out of this placement while friend suggestions stay available elsewhere?
- What goal is driving the proposal: growth of a commerce surface such as Marketplace, ads revenue, or overall engagement?
- Would the new unit recommend items listed by people and shops on the platform, or products from any retailer?
- Is the decision for all users, or for a particular market or user group?
### What a Strong Answer Covers
- Framing the decision as the opportunity cost of a scarce placement, judged against the company's mission and the stated goal.
- A way to measure what is lost by removing friend suggestions, including long-term and network effects on new and sparsely connected users.
- Ways to estimate the new unit's value without building it, with the bias of each method.
- A common currency and an explicit decision rule for comparing the two sides.
- Options beyond a yes-or-no swap, and the risks of a more commercial feed.
### Follow-up Questions
- Early signals suggest people would click shopping recommendations far more often than friend suggestions. Why might clicks still be the wrong basis for the decision?
- Two weeks after a change, friend requests look unchanged. What longer-term effects could still make the swap a mistake?
- Leadership asks for a compromise instead of a yes or no. How would you decide which users see which unit?
Overview: A product decision prompt about replacing Facebook's friend-suggestion unit with a shopping recommendation unit that has not been built yet. It tests how a product manager weighs the measurable cost of removing an existing feature against the estimated value of a new one, using cheap experiments, opportunity sizing and a decision rule.