Increase posts receiving comments via experimentation

Goal: raise the count of posts that get at least one substantive comment. a) State the success metric exactly as the share of newly created posts that obtain at least one comment that is not removed within 24 hours of posting; enumerate guardrail metrics such as DAU, creator session duration, abuse reports, and comment quality/profanity. b) Compute the current baseline and minimum detectable effect from the most recent 28 days; call out any seasonality factors. c) List at least 10 ideas spanning supply (lower the effort to comment), demand (boost the desire to comment), matching (show posts to probable commenters), and notifications/feeds; for each, note the risks and anticipated effect size. d) Identify the data to gather: post and comment creation timestamps, creator and commenter network relationships, content type, notifications delivered and opened, and dwell time. e) Pick your two strongest ideas and lay out the experiment design, including randomization unit, how to handle network interference (for example, cluster randomization at the poster or community level), ramp schedule, power assumptions, guardrails, and spillover checks. f) Describe how you will attribute any lift (new commenters versus comments moved from elsewhere), watch for abuse and quality issues, and make the final ship/no-ship call.

Overview: This prompt tests a data scientist's ability in experiment and metric design, power and MDE estimation, causal inference under network interference, product analytics, and generating ideas across supply, demand, matching, and notification levers.