Audiences Customer Studio

Operator builds a Customer Studio Parent Model on top of the warehouse but picks an unstable session_id as the identity column instead of customer_id.

Parent Model identity column must be stable across sessions (customer_id, user_uuid). Per Customer Studio docs, every audience traits + related-events join through this identity — instability fragments users and breaks audience membership. Use identity resolution to merge known aliases.

Audience filters users where churn_risk_score >= 0.8. The trait is updated daily by dbt; the audience is refreshed hourly. Operators think the audience reflects today's data when it reflects yesterday's.

Document the trait's source-of-truth freshness in the trait description; set the audience refresh cadence to the trait cadence (or chain via dbt Cloud trigger). Surface the trait's as_of timestamp in the audience UI so analysts see actual recency.

Operator splits a 'high-value lead' audience 50/50 into Treatment and Control. On every refresh, users reshuffle between the splits; experiment results are meaningless.

Use a deterministic hash on the identity column (e.g., mod(hash(user_id), 100) < 50) for split assignment so users stay in their split across refreshes — per Customer Studio docs, splits should be sticky. Persist the assignment in the warehouse for downstream attribution.