Diagnose a News Feed Engagement Dip That Coincides With Growth in Facebook Events

You are a product manager at Meta. Over the same period, engagement with Facebook Events has gone up while engagement with News Feed has gone down. What would you recommend to fix the situation? Walk through how you would investigate it before you make the recommendation. ```hint Linked or coincidental? Before proposing a fix, work out whether the rise and the fall share a cause or merely overlap in time. The right recommendation is very different in each case. ``` ### Clarifying Questions - How is engagement measured on each surface: time spent, interactions such as likes and comments, sessions, or daily active users? - How large are the two changes, and did they begin at the same time? - Did anything ship around then, such as an Events launch, new Events notifications, or a change to Feed ranking? - Is the Feed decline spread across all users, or concentrated in a segment such as the people whose Events usage grew? ### What a Strong Answer Covers - Checking the data and metric definitions before treating either change as real - Testing whether the two changes share a cause or whether Feed fell for reasons of its own, with specific hypotheses for each - Segmenting by user group, region, platform and timing to locate the decline - A recommendation judged on Facebook's overall health, including total engagement, retention and ad revenue, rather than on one surface's metric - A concrete next step, such as a holdout experiment or a rollback, with the metrics that decide it ### Follow-up Questions - The analysis shows people moved time from Feed to Events, and total time on Facebook is unchanged. Is there anything to fix? - Suppose a minute in Feed earns far more ad revenue than a minute in Events. How does that change your recommendation? - How would you design an experiment to measure how much of the Feed decline the Events change caused?