Isekai: Slow Life launched in July 2023 as an Idle RPG with simulation and town-building mechanics, and has since built a sizable base of paying users. As the game matured, a growing portion of those high-value players gradually went inactive, creating a significant pool of dormant paying users whose lifetime value remained largely untapped. With this in mind, Mars Games partnered with RZR in June 2025 to launch a retargeting campaign with a clear mandate.
Mars Games started with a test in the US market reactivating two precisely defined dormant-payer segments. Within a year, that single test became a global retargeting engine spanning 5 Tier 1 (T1) markets and a 180-day dormant window.
Play 1: Precision Audience Segmentation
Target: Re-engage high-value dormant payers while maintaining acquisition cost from the outset.
Execution highlights:
CPI came in well below target from day one, with Day 7 (D7) ROI meeting the KPI at launch, validating that RZR's retargeting model was the right fit for dormant-payer re-engagement.
Play 2: Continuous Model Iteration
Target: Hold performance steady across the game calendar.
Execution highlights:
D7 ROI peaked at 95-100% during monthly collaboration events. Even in off-event periods with no promotional tailwind, performance held steadily above the KPI floor — demonstrating model robustness independent of the product calendar.
Play 3: Audience & Market Expansion
Target: Scale audience depth and geographic reach without sacrificing efficiency.
Execution highlights:
RZR's sustained performance cemented its position as Mars Games' primary retargeting partner.
RZR's proprietary machine learning models predicted the latent lifetime value of dormant users, integrating cross-channel behavioral signals to tier bids by user quality.
Audiences were continuously refreshed to eliminate stale targeting, removing churned users, and adding newly lapsed payers on a rolling basis. This solved audience decay at the infrastructure level, enabling stable delivery month over month and across product update cycles.
By mapping client-side event schedules in advance, RZR concentrated resources during the windows of highest player purchase intent, generating measurable revenue uplift at peak. Model-driven strategies held the baseline during quieter periods so results were not dependent on event timing alone.
Every scale-up followed the same playbook: validate in a single market, extend the audience window, then replicate proven strategies into new geos. No blind spend — every incremental dollar was backed by performance evidence.
RZR built a performance foundation that compounded with every campaign phase:
LTV-driven segmentation bidding built around dormant-payer value, not demographic tags
Dynamic real-time audiences audience decay solved at the infrastructure level for stable month-over-month delivery
Game-calendar alignment resources concentrated at peak purchase intent, baseline held in between
Data-first expansion every scale-up validated in one market before replicating into new geos
Long-term partnership one proven phase became the case for the next
Case Study
Ella Yang, VP @DHgate
Case Study
Masha Astapkovich, Head of UA, Highcore Games
Case Study
— Nick He, Head of Performance Marketing, Mars Games
Case Study
Marie Kim, User Acquisition - Team Lead @ Bagelcode
Case Study
- HYEGANG SEO, HEAD OF UA, ACTIONFIT
Case Study
Aleyna Carrah, User Acquisition Manager, APPS
Case Study
.png)
Ecem Kaya, Growth Manager, Paxie Games
Case Study
- MARTIN BOCCARDI, SENIOR PERFORMANCE MARKETING MANAGER @ ROVIO
Case Study
HUOJIAN, PERFORMANCE MARKETING LEAD@TALKIE
Get cutting-edge insights on retention-led growth straight to your inbox.
