Overview
Problem
CAC remained stable, but user quality varied significantly across campaigns.
Observed:
• same CAC produced 2–3x LTV differences
• low-cost users often churned early
• high-quality users came from specific creative angles
What we did
We rebuilt creative strategy around user intent segmentation.
Actions:
• categorized creatives by intent level (high vs low intent)
• tested multiple psychological hooks per segment
• separated traffic flows by expected LTV profile
• aligned creatives with downstream monetization behavior
• reallocated spend toward higher-value cohorts
• removed low-quality acquisition angles despite strong CAC performance
Result
• +20–35% improvement in blended LTV
• −25–40% reduction in low-quality user share
• CAC remained stable
Key insight
Creative is not just a performance lever. It is a filtering system for user quality.



