Problem
- Basic frequency capping treated everyone the same
- Risk of mid-content ads hurting experience for retention-critical users
- Needed revenue upside without burning high-ARPU cohorts
Eligibility — when ads can fire
- Only at a genuine session break — never mid-content
- App backgrounded and reopened within 30 minutes, or
- 60+ seconds of pure inactivity within a session
Segmentation by ARPU
-
Low-ARPU — no daily cap; up to 2 per session (yield opportunity outweighed
marginal experience cost)
-
High-ARPU — capped at 2 per day, 1 per session, with a minimum session gap
(protect retention-critical users)
- Rolled out starting with the low-ARPU segment first
Outcome
- Significant revenue jump from the low-ARPU segment
- Highest-value users left untouched by the aggressive path
- Monetization + UX tradeoff encoded in policy, not one-size-fits-all caps