Rocket Studio, a Vietnamese game studio, has been operating since 2015, focusing primarily on developing arcade action games, including the highly popular and well-reviewed “Galaxy Attack: Space Shooter.”
As one of Vietnam’s earliest hyper-casual game studios, Rocket Studio initially concentrated on exporting games to the US, which remains their largest and most valuable market. Since 2017, they have been utilizing third-party app measurement suites. Beyond measuring the effectiveness of user acquisition campaigns, they focus on acquiring players with high lifetime value (LTV), achieving high return on ad spend (ROAS), and reducing cost per install (CPI).

Challenges Faced by Rocket Studio
When Apple introduced the SKAdNetwork (SKAN) protocol and the App Tracking Transparency (ATT) framework, Rocket Studio, like many advertisers, faced numerous new measurement challenges in their iOS advertising campaigns. With iOS accounting for nearly one-third of Rocket Studio’s total traffic, the team actively sought a solution to help them avoid significant revenue loss.
The 24-hour postback window mandated by SKAN proved particularly complex for Rocket Studio’s longer payment window, making it especially challenging to identify potential high-value players and accurately measure the effectiveness of their ad campaigns. Additionally, the limited measurement and data visibility made it difficult to balance ROAS and CPI, as well as to scale UA marketing campaigns on iOS.
Solution

Rocket Studio collaborated closely with third-party tools and Meta’s Gaming division to design a comprehensive solution that enhances their measurement capabilities and improves ROAS, resulting in tangible revenue outcomes.
“The third-party Predict solution is a key component of our iOS marketing campaign optimization and retention strategy, helping us acquire more valuable users with a high likelihood of making payments.”
The team adopted a two-pronged approach. First, they viewed the predictive analytics solution as an ideal starting point to overcome SKAN limitations. Second, they utilized Meta Gaming’s value optimization features to translate predictive insights into acquiring high-value players.
Predictive insights enabled Rocket Studio to measure SKAN campaign performance from the initial 24 hours to projected 30-day performance metrics, including 30-day ROAS, average revenue per user (ARPU), predicted percentage of paying users, and rolling retention rates.
These insights were provided through a unique predictive model designed to adapt to each app’s business logic, correlating initial user engagement measurements with their final outcomes based on advanced analysis of Rocket Studio’s historical data.
Predict directly transmitted these values to Apple’s SKAN protocol, converting them into SKAN conversion values, encapsulating the entire measurement within the 24-hour window, and providing direct feedback to integrated media partners.
With the third-party predictive data in hand, it was time to test it. The two teams jointly designed a geographically based holistic framework to measure incremental ad revenue lift, offering Rocket Studio better ad placement and measurement while enhancing their iOS campaign optimization capabilities.
The third-party incremental solution relied on a time-based regression (TBR) method, estimating the incremental causal impact of ad interventions on response metrics, such as ad revenue in Rocket Studio’s case. The geographically comprehensive experimental design ensured proper balance and high correlation between test and control groups, maximizing the likelihood that test results were attributable to the incremental effect of the campaign rather than group differences or other variables.
The team optimized their campaigns based on predicted revenue metrics, considering the predicted lifetime value of players. These predictive values were delivered within just 24 hours, providing Rocket Studio with actionable insights into their campaign’s expected performance within the Apple SKAN framework. The predictive values were seamlessly transmitted to Meta Gaming’s campaign optimization engine through the SKAN framework itself.
Marketing Results
This framework has proven to be highly effective.
“Meta Gaming has always been one of our preferred UA channels. Post-ATT, we believe that optimizing Facebook gaming ad campaigns with Predict and Incrementality to drive revenue is a reliable way to attract high-quality users and improve ROAS.”
Compared to the control group, Meta Gaming’s value optimization campaigns, which relied on third-party predictive insights, increased Rocket Studio’s revenue by 6 times. Additionally, the average ROAS on day 7 improved by 42% compared to typical app install campaigns, solidifying the success of this strategy.