Facebook Groups Engagement Falls 20%: Building the Root-Cause Investigation

Facebook Groups usage has fallen 20%. As the PM for Groups, how would you work out why? ```hint Shape before causes Before listing causes, pin down which usage metric fell and whether it dropped overnight or slid over weeks. Then ask whether Facebook as a whole moved too. ``` ### Clarifying Questions - Which metric fell 20%: daily active users of Groups, time spent in Groups, posts and comments, or new group joins? - Over what period is the drop measured, and did it happen overnight or gradually? - Is the drop on every surface (the Facebook app, web, Facebook Lite) and in every country, or concentrated somewhere? - Is overall Facebook usage down by a similar amount, or is the decline specific to Groups? ### What a Strong Answer Covers - Validating the drop: metric definition, logging or pipeline changes, the comparison baseline and seasonality - Separating a Groups-specific decline from a Facebook-wide one, then localizing it by platform, app version, region, user cohort, group type and entry point - Decomposing usage into its drivers, including creators versus readers, to see where the chain broke - An organized set of internal causes (releases, ranking, notifications, navigation, integrity enforcement) and external ones (OS changes, competitors, events), each with a test, checked in a sensible order - A conclusion with next steps, including whether some of the lost activity was worth losing ### Follow-up Questions - The whole drop is in visits that start from notifications. What would you look at next? - A spam-enforcement change removed a lot of low-quality group activity. Is a 20% drop acceptable in that case, and how would you decide? - How would you set up monitoring so the team catches a drop like this within a day?