How We Handle Trust & Safety

Set clear policy and acceptance standards before campaigns run. Prevention reduces the volume and severity of issues that detection and response must handle.

• Document eligibility policies and prohibited behavior patterns (e.g. multi-accounting, fake proof, commitment breach) and make them visible in product and support.
• Embed campaign-level quality requirements into setup guidance so creators know what proof and behavior are required.
• Train reviewers on consistent interpretation of acceptance criteria; use a shared rejection-reason taxonomy and examples.

Detection combines system signals and human review. Use signals to prioritize so manual effort goes to the highest-risk or highest-impact cases first.

• Use anomaly cues (e.g. spike in rejections, duplicate proof patterns, velocity changes) to prioritize manual checks efficiently.
• Tag rejection reasons in structured categories for trend analysis and so prevention can be updated when patterns emerge.
• Escalate repeat incidents with clear severity thresholds (e.g. same user rejected 3+ times, or same campaign with >X% rejection rate) so response is consistent.

Fast, consistent response protects trust for both sides. Define who owns what, how to communicate, and how to close the loop.

• Define response ownership and communication templates (e.g. rejection reason text, appeal response) so users and creators get clear, consistent messages.
• Log corrective actions and follow-up outcomes in a shared timeline so patterns and repeat offenders are visible.
• Review incident patterns monthly to harden preventive controls (e.g. update acceptance criteria or product checks when the same issue recurs).