Before launching a feature, we can imagine that the first step we’d have to take is analyzing the existing data in our product to make a decision about exactly what to build. This process is what creates the building or change of a feature problem that gets asked in product interviews.

Examples of questions like this usually take the form of:

• How would you analyze the effect of adding a green dot on LinkedIn messaging?
• How would you analyze the effects of the close friends story on Instagram?
• How would you decide whether or not to launch a new personalized recommendation widget on the Amazon home screen?
• How would you analyze the effectiveness of allowing users to trade fractional shares on Robinhood?

For this type of problem, let’s tackle an example problem.

Let’s say that you work for Uber. A product manager wants to add a feature to the app that would increase wait times for a lower price. Is this a good idea?