.png?width=200&height=52&name=Adikteev-H-White%20(1).png)
“Adikteev didn't just run our campaigns, they validated our 2-day inactivity window against real return behaviour. That confidence in the setup let us scale Rock N' Cash without second-guessing efficiency.”
SEO KYUNG LEE
Growth Marketing Manager
background
Rock N' Cash is one of Playlinks' flagship social casino titles, already running retargeting across both iOS and Android with a 2-day inactivity window (IW). Before scaling further, Playlinks wanted to confirm whether that IW setting was truly optimal and to improve overall campaign efficiency.
objective
The campaign set out to:
- Validate whether the existing 2-day IW was the most effective retargeting window.
- Achieve a lower cost per reattribution (CPA), and increase D7 ROAS while maintaining scalable volume.
campaign strategy
The strategy combined three moves:
- Comeback analysis to validate the IW: We studied organic users who saw no ads to understand natural return behaviour. Around 90% of inactive users did not come back on their own after two days, confirming that the 2-day IW struck the best balance between scale and user freshness while minimising wasted impressions.
- Value-based audience segmentation: Using Playlinks' postback data, purchasers were split into the top 20% and the bottom 80%, with distinct machine learning objectives for each. The top 20% were optimised for CPA reattribution to reactivate high-value users at the lowest cost. In comparison, the bottom 80% were optimised for ROAS to surface the highest revenue potential in the wider base.
- Weekly budget reallocation: Performance was reviewed across segments every week, with the budget continuously shifted toward the audiences delivering the best CPA, ROAS, and scale.
results
We scaled campaign spend by 2.5x while hitting both goals at once. It confirmed the 2-day IW as the optimal setting, achieved Playlinks' CPA target through audience-specific optimisation, and increased D7 ROAS by applying tailored machine learning strategies across purchaser segments
- For social casino and other purchase-driven games, a single optimisation strategy rarely fits everyone. This campaign showed that validating the inactivity window against real return behaviour, then tailoring machine learning objectives to different purchaser segments, lets one campaign deliver low CPA and high D7 ROAS at the same time.
- Treating the top purchasers differently from the rest, and reallocating budget weekly toward the strongest audiences, is what makes both goals achievable while scaling.
