Building a success scorecard for Facebook's save-for-later bookmarking feature

You are the product manager for Save on Facebook: the option that lets people bookmark posts, links, videos and other items into a personal Saved list so they can return to them later. How would you measure whether this feature is succeeding? Explain what the feature is for, lay out the metrics you would track, and say which one you would hold up as the main measure of success. ```hint Map the steps first List what a person does with Save, from noticing the option to whatever happens after an item lands in their list, and ask which of those steps shows the feature actually helped them. ``` ### Clarifying Questions - Is success judged mainly by value to the people who save, by overall engagement on Facebook, or by effects on ads and revenue, or should I choose? - Are all saveable item types and both the app and the website in scope, or only posts saved from News Feed? - Is this a health check of the feature as it stands, or are we judging the impact of a particular change to it? ### What a Strong Answer Covers - What Save does, the user problem behind it, and an explicit goal for the feature with the reason for choosing it - The journey from saving an item to coming back to it, broken into steps that can be measured - One primary success metric tied to that goal, with a precise definition and the reason it beats the obvious alternatives - Supporting and guardrail metrics, including whether saving adds to or takes away from engagement in the moment - How the numbers would be read: cuts by item type and user, time windows that fit "later", and separating the feature's own effect from who chooses to use it ### Follow-up Questions - Saves have climbed sharply since a redesign, but your primary metric has not moved. How would you work out why? - If the team wanted to remind people about items they saved and never opened, how would you test whether that helps them or just adds notification noise? - How would you tell whether Save creates engagement that would not otherwise happen, or only postpones engagement that would have happened anyway?