Do Teens Pull Back When Parents Join Facebook? Causal Analysis and Product Fixes

You are a product manager at Meta. Consider teenagers on Facebook whose parents have joined Facebook too. First, explain how you would evaluate what the parents' arrival does to the way those teenagers use Facebook. Then propose how you would improve the product based on what that evaluation shows. ### Clarifying Questions - Does "parents joined" mean the parent created an account, became Facebook friends with their teen, or started interacting with the teen's posts? - Which ages count as teenagers: only 13 to 17, or older teens as well? - Is the improvement meant to win back teen activity on Facebook itself, or to maximize the family's engagement and well-being across Meta's apps? - Which data can the analysis use: only declared family relationships, or also inferred ones, given the privacy limits on studying minors? ### Part 1 — Evaluate the impact How would you measure what a parent joining Facebook does to their teenager's use of the product? Say what you would treat as the parent joining, which teen behaviors you would track and which method would give you a credible answer. ```hint Not an experiment you can run You cannot randomly assign parents to join. Think about what else changes for a teen around the time a parent arrives, and which teens could serve as a fair comparison. ``` #### What This Part Should Cover - A precise definition of the exposure and of the teen outcomes, going beyond time spent. - A comparison that is credible despite teens getting older, platform-wide trends and families choosing when parents join. - The privacy limits on analyzing minors and family relationships. ### Part 2 — Improve the product Based on what the evaluation could show, how would you improve Facebook for teenagers whose parents are on it? Choose the users and the problem you would focus on, propose and prioritize solutions, and say how you would measure success. ```hint Two audiences on one profile A teen's friends and their parents may now see the same posts. Ask what the teen stops doing because of that, and what the parent actually wants from being there. ``` #### What This Part Should Cover - Improvements that follow from specific findings in Part 1 rather than a generic list of teen features. - The needs of both teens and parents, with protections for minors kept intact. - Prioritization, success metrics and an experiment design that accounts for teens and parents being linked. ### What a Strong Answer Covers - A clear decision about what "better" means for Meta: teen activity on Facebook alone, or the whole family's engagement and well-being across Meta's apps. - A logical chain from the measured effect, to the problem chosen, to the fixes proposed. - Balance between teen privacy and autonomy and parents' legitimate interest in oversight. - Guardrails that would catch a fix that helps teens but drives parents away, or that weakens protections for minors. ### Follow-up Questions - Suppose the analysis shows teens post less after a parent friends them, but send more private messages. Is that a problem worth fixing? - How would you run an experiment on these features when a teen and their parent could otherwise land in different test groups? - A parent wants more visibility into their teen's activity while the teen wants more privacy. How do you decide what to build?