I spent too long on the obvious stuff like report volume and resolution rate before getting to the actual crux of the question.
Start by clarifying the goal of the 'Report Ad' feature: to improve ad quality and user trust by removing bad ads while minimizing harm to legitimate advertisers. Then define success metrics across user experience, advertiser impact, and operational efficiency, explicitly addressing false positives and false negatives. Finally, propose a measurement framework that balances these metrics and suggests experimentation to optimize thresholds.
Pro tip: Emphasize that false positives and false negatives have asymmetric costs—false negatives harm user trust and platform integrity, while false positives hurt advertiser ROI and can chill legitimate advertising. Propose a weighted metric or a guardrail metric to ensure balance.
Define the primary goal of 'Report Ad': to empower users to flag bad ads and improve ad quality. Identify key stakeholders: users, advertisers, and Meta's ad review team.
Measure user trust and satisfaction: report rate, resolution rate, and reduction in bad ad impressions. Track false negatives via user surveys or downstream signals like ad hides.
Measure false positives: incorrectly removed ads, advertiser appeals, and impact on advertiser spend and retention. Use control groups to estimate counterfactual.
Track precision and recall of the ad review system, time to resolution, and cost per review. Use these to optimize the trade-off between false positives and false negatives.
Combine metrics into a balanced scorecard with guardrails. Suggest A/B tests to tune thresholds and measure long-term effects on user trust and advertiser health.
AI-generated suggestions, not part of the candidate's original notes. May be inaccurate — verify before relying on them.