How Falcon Games scaled without sacrificing quality across their portfolio

Falcon Games is a Hanoi-based mobile game studio and publisher, primarily developing mobile puzzle titles. The studio’s portfolio of popular puzzle games includes hits like Goods Puzzle: Sort Challenge, Cake Sort - Color Puzzle Game, Bird Sort 2: Color Puzzle, and Block Match: Detective Puzzle.

The goal:

Falcon Games looked to further scale their portfolio with players who monetize through a hybrid of in-app purchases (IAP) and ads over the long-term. Importantly, they had established a performant D7 ROAS campaign foundation on Unity Ads. They sought a solution to grow with new high-quality users without disrupting what was already working.

The solution:

The studio layered longer-horizon optimization windows using D28 Hybrid ROAS and D28 Ad Revenue campaign types on top of their D0 and D7 campaign types to unlock new cohorts of valuable players across their portfolio.

How Falcon Games grew total scale with D7 and D28 ROAS campaigns

Starting in mid-March 2026, Falcon Games began running with D28 Hybrid and D28 Ad Revenue ROAS campaigns on top of their existing D0 and D7 campaigns across their leading titles.

The approach was to use the longer window campaign types as an additive to their user acquisition stack. Instead of replacing their D0 and D7 campaigns, the D28 campaigns were intended to capture high-quality users that showed no conversion signals within the first week.Without this addition, these users would have gone unaddressed.

Their campaign strategy stacked D28 budgets on top of D0 and D7 spend, leaving existing campaign budgets intact. Rather than reallocating budget across optimization windows, this let campaigns capture the full value of user cohorts converting at D7 or D28**.

The result was an 84% increase in total ROAS campaign spend at the account level. Account-level spend trend data shows D28 budgets as a new layer sitting above the prior baseline, demonstrating new incremental scale for Falcon Games’ portfolio.

How Falcon Games’ D28 ROAS campaigns drove quality and volume improvements

While the scale improvement alone from Falcon Games’ D28 campaign was substantial, the greater win was the quality uplift that these campaigns drove and maintained. Comparing four-week averages before D28 campaigns were launched (January-February) against completed cohorts after the launches (April-May), user quality improved.

In same the time period, account-level D28 ARPU grew 1.5× illustrating how users acquired through D28 campaigns were generating materially higher revenue over the 28-day cohort window. Similarly, retention was also stronger for users acquired through these campaigns. Account-level D28 Retention improved by 22%, reflecting the capture of higher-intent users who invested long-term in the titles***.

How D7 and D28 work in tandem to cover the full LTV spectrum for Goods Puzzle: Sort Challenge

Taking a closer look at Falcon Games’ campaign architecture for Goods Puzzle: Sort Challenge for iOS in the US reveals how the multi-campaign strategy operates in practice for their portfolio****. For the title, Falcon tested every available ROAS campaign type and converged on the combination that best covers their full LTV window:

Each campaign type has a specific function in the campaign mix that enabled Falcon Games’ results. D7 Hybrid ROAS campaigns optimize for efficiency across both IAP and ad revenue signals within the first week, generating strong early returns at scale. D28 Hybrid ROAS extends that same combined signal across a full 28-day window, targeting users with highest-value behaviors. These include larger purchases, deeper ad engagement, and stronger retention - all behaviors that don't fully materialize until later. Running both these campaign types meant the studio was able to optimize for both volume and quality.

At the same time, D7 Ad Revenue ROAS campaigns cast a wide net early, helping to identify users who engage with ads quickly and acquiring them at a lower CPI. But, it only sees the first seven days of a user's behavior. D28 Ad Revenue ROAS captures slower converters who show signals later.

All together, these varied ROAS campaign types form a stack that optimizes across most meaningful points in the user lifecycle from D0 through D28.

How the right data signals empower Unity Ads campaign performance

Falcon Games’ scale and quality results did not happen in isolation. A significant predictor of their campaign success was the strong data foundation they supplied to Unity Vector, the machine-learning algorithm that powers Unity Ads.

By integrating the latest Unity Ads SDK and aligning with Vector's data integration requirements, Falcon Games helped to provide signals for Unity Vector’s models to understand the value across the 28-day cohort window.

*Source: Internal Unity Data, based on Falcon Game Studio campaigns. Account-level spend growth measured based on a data comparison of Jan 2026 to Feb 2026 vs May 2026 to June 2026. Account-level D28 ARPU growth measured based on a data comparison of Jan 2026 to Feb 2026 vs April 2026 to May 2026. Account-level D28 retention growth measured based on a data comparison of Jan 2026 to Feb 2026 vs April 2026 to May 2026.

**Source: Internal Unity Data, based on Falcon Games campaigns. Account-level spend growth measured based on a data comparison of Jan 2026 to Feb 2026 vs May 2026 to June 2026.

***Source: Internal Unity Data, based on Falcon Games campaigns. Account-level D28 retention and D28 ARPU growth measured based on a data comparison of Jan 2026 to Feb 2026 vs April 2026 to May 2026.

****Source: Internal Unity Data, based on Falcon Games’ Goods Puzzle: Sort Challenge iOS US campaigns.