Engagement, Churn, and Dormancy: A Better Prism

Written by Jeff Day, Principal Research Scientist

Video game players exhibit different patterns of play. Some engage in bursts of a few days, some over prolonged time periods, others come and go often, while still others play at predictable time intervals. Whether these engagement patterns are driven by game level completion, freshness of new content, competitor releases or seasonal factors like holidays and the school calendar, understanding these patterns matters greatly in terms of how you interact and manage your player base, regardless of whether your title has an evergreen player base or a finite decay curve. Yet the industry standard for tracking engagement is woefully inadequate, as it simply lumps players into two buckets: Retained or Churned. Great for a dashboard of daily or monthly active users (DAU or MAU), but little utility beyond that.For all intents and purposes, in many studios all inactive players are treated the same and no distinction

For all intents and purposes, in many studios all inactive players are treated the same and no distinction is made as to who is likely gone for good and who may be likely to re-engage. The same holds true for engaged players and their likelihood of continued engagement. Typically, no adjustments are made for where a given player resides in the game’s typical life cycle, and when they are likely to take a break or churn entirely. We have found that even the most engaged and tenured players are likely to have at least some periods of inactivity. And these players are often those who purchase content or monetize in-game. Failure to understand and actively manage to this dynamic leads to lost opportunities. You need a better prism. A prism that allows you to better understand, interact, and service your customers in ways that benefit both the player and you.At NBI, we categorize player activity into three broad categories: Engaged, Churned, and Dormant.

At NBI, we categorize player activity into three broad categories: Engaged, Churned, and Dormant. Current, and likely, membership across these categories can be determined statistically. We overlay derived playing patterns with where a player is in the game’s lifecycle, and then combine it with player statistics and in-game behavior to not only classify a player as either likely to remain engaged, become dormant, or churn altogether, but also assign a probability of continued or future engagement. Knowing this information before a player leaves, or when they are potentially likely to come back, can help maximize the effectiveness of retention and reactivation efforts.Specifically, this information can be fed into an in-game recommendation engine to help mitigate churn

Specifically, this information can be fed into an in-game recommendation engine to help mitigate churn in the first place as well as allow you to calibrate the timing of content releases, events, and promotions around optimal re- acquisition windows for dormant players. In addition, by separating Churned and Dormant players, and being more certain that the Churned status really does mean highly unlikely to return, adjustments to your installed base calculations can be made, thereby enabling better projections of future key performance indicators. There may be multiple types or “clusters” of dormancy that a player may experience (school schedules, time after finishing content, leaving temporarily to play another game, etc.) that should be treated separately in ongoing reporting, and re-acquisition efforts. Simply put, Churned players have for whatever reason moved on… and will be very difficult, and expensive, to reacquire. Dormant players, however, represent a pool of potential players whose eventual re-engagement is essential to hitting your KPIs, and likely represents a larger pool than the number of brand new players acquired each month. The Engaged are your life blood, keeping them engaged and knowing when to provide that extra incentive can mean the world to the health of your game. Managing to these three constructs… Retained, Churned, and Dormant, is a simple concept but one that yields powerful and actionable insights.