Tag Archives: Analytics

What is the best way of understanding and measuring game retention?

Retention provides key information used in the optimization of your game.  A recurring user gives you many opportunities to find the best configuration of game play, engagement, monetization, and more.  High level retention stats, such as stickyness (DAU/MAU = engagement) can only give you a quick snapshot on how you are doing.  While its a good, quick health-check, it is usually not deep enough to be actionable. In order to effectively measure retention and really make it actionable, you need deeper statistic capabilities, an understanding of your gameplay, and a bit of a curious mind.

Of course you should be able to track standard usage metrics like DAU, MAU, and Time Spent (if you have the PlayPhone SDK integrated, you’re already doing this!) but additionally you should be able to track your own game defined KPIs, such as purchasing a certain item, reaching a certain level, connecting with a certain number of friends, etc. From there, you will want to be able to track cohorts, or blocks of users that fall into a certain category like ‘date acquired’ or ‘time since last play’ or ‘paying vs non-paying’, etc.

Setup consistent analysis for each cohort over time, week over week or day over day (IMO a month is a long time in the world of game play and you can miss some important trends), to identify which users exhibit great engagement and retention, which ones do not, and what are the key differences between each cohort.

As an example, you could;
Show me the week over week retention of all users who play in Easy mode vs Hard mode.
Action: Maybe the game is too easy, people like a challenge, make it more difficult and compare a new cohort versus these 2 controls.

To gauge external factors; (ex. Media Buying effort)
Show me the week over week retention of users who came from AdMob vs InMobi.
Action: Pay more per click where your better retention is to acquire more users.

If you’re not capturing data in this format, its actually become somewhat trivial using a company like Mixpanel which tracks defined actions to give you solid funnel and waterfall analysis tools. If you have the inhouse capabilities, start with simple models (ex. DAU/MAU with weekly cohorts) to get your reports ironed out, and build from there based on business questions you want answered.

Beyond that, don’t hesitate to reach out with questions!

E: scott.teger@playphone.com