Mapendo has recently partnered with one of the most downloaded block-crush puzzle games, featuring vibrant cubes exploding in color when players solve a puzzle. Offering daily puzzle variety and rewards, an endless high score hunt and the possibility of playing offline are the reasons behind the app’s high ranking on the US app store. The client’s goal was to acquire large volumes of new high-quality users, achieving a D1 ROAS above 5% and D7 ROAS above 30%.
This operation was not without challenges. First things first, Mapendo had to satisfy the client’s high-performance expectations. Since the app ranked among the top titles on the US app store in the puzzle and games category, the client requested very high volumes of new users daily.
Another challenge emerged during campaign optimization. More precisely, we needed to optimize those ROAS objectives without being able to count on immediate and indicative revenue postback, such as in-app purchases used for IAP-driven mobile games.

By coordinating with the client, we identified an effective in-app event tailored to our strategy. This in-app event was sent via postbacks and reflected new users quality in terms of ads seen within a specific event, which was activated each time a user saw 12 ads in the first 24 hours of app use. This event proved to be highly indicative of the quality of the newly acquired users.
Once we identified the main monetization event and the ROAS, we were able to optimize the campaign in order to reach the client’s KPIs. Our platform created a matrix containing millions of collected data points. After obtaining a sufficient number of newly acquired users, we were able to individuate the pockets, meaning the matrix cells, with the best ROAS results.
This was possible thanks to our machine learning algorithms, which assigned to each pocket a reward, meaning a value linked to the in-app event we wanted to optimize. By doing that, we were able to allocate our budget towards acquiring users from the best-performing pockets, meaning users with a higher LTV, who ultimately contributed to improving the campaign ROAS.
By the end of the campaign Mapendo obtained significant results, proving the strategy to be highly effective in scaling high-quality user acquisition. In fact, installs had grown by 240% and events by 130%, while an installs-to-event rate of over 40% was recorded. All of this, by reaching and overcoming the client’s ROAS D1 and D7 targets.
This campaign demonstrates how a data-driven optimization strategy can successfully scale user acquisition even for apps monetizing primarily through advertising. By identifying a strong engagement signal and leveraging Mapendo’s machine learning technology, we were able to efficiently acquire high-quality users while exceeding the client’s ROAS targets.
- +240% installs
- +130% events
- +40% installs-to-event rate



.png)



