Understanding Gamer Retention in Social Games using Aggregate DAU and MAU data: A Bayesian Data Augmentation Approach

With an estimated market size of over $6 billion in 2011, “social” games (games played over social networks such as Facebook or Google+) have become increasingly popular recently. Understanding gamer retention and churn is important for game developers, as retention rate is a key input to gamer lifetime value. However, as individual-level data on gaming behavior are not publicly available, developers generally rely on only aggregate statistics such as DAU (daily active users) and MAU (monthly active users), and compute ad hoc metrics such as the DAU/MAU ratio to assess retention rate, often resulting in very inaccurate estimates. I propose a Bayesian approach to estimate retention rates of social games using only aggregate DAU and MAU data. The proposed method is based on a BG/BB model (Fader et al. 2010) at the individual level in conjunction with a data augmentation approach to estimate the model parameters. After validating the performance of the proposed method through a simulation study, I apply the proposed approach to a sample of 379 social games. I find that the average 1-day and 7-day retention rates for new players across games are 59.0% and 10.5%, respectively. Further, my results suggest that the median break-even acquisition cost per gamer is about 13.1 cents. In addition, giving out a “daily bonus” or limiting the amount of time that gamers can play each day may increase 1-day retention rate by 6.3% and 6.9%, respectively.