Audiences
What the Audiences domain tests
Audiences are dynamic member sets built from user profile attributes and event behavior. The exam tests criteria logic, time windows, list-versus-audience choice, and how audience membership feeds campaigns and journeys.
Core rules
- Audience vs List: an Audience recalculates from criteria; a List is a static set managed manually, by CSV, or by API. Use an Audience when membership should follow changing behavior; use a List for a fixed, approved group.
- Criteria types: profile attributes (plan, country, signup date), event filters (“has done” or “has not done” an event within a window), event counts, and combined logic with AND/OR. Nested groups change the result materially.
- Time windows: “purchased in last 30 days” is relative to evaluation time, not campaign send time. Verify the window and the event’s date field.
- Refresh and use: audiences can be referenced by campaigns and journeys. Exclusions and suppression apply at send time, not at audience definition.
Worked example
Build “high-value lapsed”: total purchases ≥ 3 AND last purchase more than 60 days ago AND has not unsubscribed. Reasoning: the count identifies high value; the recency window identifies lapse; the unsubscribe exclusion prevents illegal or unwanted sends. If you used OR between purchase count and recency, nearly every recent buyer would qualify-wrong set.
Traps
Mixing AND/OR without grouping produces oversized or empty audiences. Using a List for behavior that changes daily makes stale targeting. Forgetting that suppression is applied later can make audience counts differ from send counts. Testing criteria on a small sample without checking event timestamps can hide timezone or date-field errors.