Why Flurry's Retention Metrics Are Misleading Developers

Retention is the foundational metric of the freemium mobile game economy, serving as the basis for engagement, virality, and monetization. However, a significant discrepancy exists between standard industry expectations and the data provided by Flurry, a leading analytics provider. While the industry relies on Day N Retention—the percentage of users returning specifically on the Nth day after download—Flurry utilizes Rolling Retention. This alternative metric tracks users who return on the Nth day or any day thereafter, effectively measuring the inverse of churn rather than precise daily engagement.

The use of Rolling Retention is problematic because it artificially inflates performance figures, particularly for older titles. For example, some apps may report a 90 percent Day 30 retention rate under this model, a figure that is impossible under standard Day N definitions. Industry benchmarks for healthy freemium games typically require a Day 1 retention of 30 to 40 percent, Day 7 at 15 to 20 percent, and Day 30 at 8 to 10 percent. By conflating these distinct methodologies, developers, publishers, and venture capitalists risk making uninformed investment and development decisions based on misleadingly high data points.

This analysis, authored by Wilhelm Taht of Playground Publishing in early 2014, highlights a critical gap in the mobile app ecosystem's data literacy. The scope focuses on the global mobile and tablet gaming sector during a period of rapid growth for freemium models. The findings emphasize that Rolling Retention offers limited insight into app quality and can lead to wasted development cycles when teams eventually realize the need for more granular analytics. Following these criticisms, Flurry committed to implementing static Day N retention metrics to better serve the developer community.