LTV and retention are common metrics used to assess the performance of a game. But the performance – and overall value and potential – of a game is not reduced to the way it performs for new installs. The more a title ages, the more valuable and important the active user base. That’s because on a daily/weekly/monthly basis, your active users are the ones driving your revenue the most. Even in the absence of growth potential (say if your LTV isn’t high enough to sustain mass injection of installs) a game can be managed very profitably because it has a group of dedicated fans who love the game and make it part of their daily routine. And usually, the older the live game, the more dedicated that group of active users.
In that case, the game’s active userbase is an asset even more valuable than its growth potential via UA/installs. And even more than active users, you want to be looking at active customers.
The thing is, while it’s very straightforward to observe fluctuations in the number of active customers playing your game it’s not always as simple to understand what is behind those fluctuations. Retention is not a metric that will provide you with insights into how your active customers are engaging and returning to the game on a regular basis.
There is a good metric you want to be tracking to see how your active customer base is varying. It’s looking at the return rate of your active customers. You can look at it daily/weekly/monthly or any other time period that’s relevant to you. The natural cycle around which users tend to play your game (and basically structure most other aspects of their lives) is on a week. For example, you’ll see more active customers in your game on the weekend, and players tend to return to the game more on a 7 day cycle.
The way you want to look at this is pretty straightforward. You want to be looking at the percent of active customers in a given week that are returning to the game the following week. While this metric is bound to be relatively stable (and pretty high – in the 70-75% range for a weekly basis is relatively standard), small variations or inflexions can have a big impact over time. Keeping an eye on that also helps to identify the often subtle – but very real – impact of gameplay or economy balancing changes.
Because the percentages you will be dealing with are relatively high to begin with – and most of the time you won’t observe huge fluctuations – you probably want to be zooming in to notice some key variations. A 2-3% variation in next week return rate for your customers can have a huge impact.
In addition, if you are observing significant fluctuations in your active customer base, you need to ensure you understand where it’s coming from. Maybe it’s because fewer of your customers are returning to the game – if that’s the case, the previous graph should indicate that. But maybe it’s something else. Maybe you are generating less new customers in the game – either via new installs or conversions of preexisting users. Maybe you are seeing less customers returning after a lapse.
That’s why in addition to looking at the weekly return rate of your customers, it can be useful to have a view that shows the delta – this time in absolute value – of your active customer base. In order to track the delta – and the source of variation – you can look at the following:
- How many active customers from the previous week (or whatever period you decide to look at) did not return this week
- How many active customers who did not play the previous week have played the current week – let’s call those reawakened customers
- How new customers did you generate during that period? Depending on how detailed you want to be, you can divide that in various subsections. For example:
- New customers that installed in that period that are organic installs
- New customers that installed in that period that are paid installs
- New customers that installed in a previous period
You would then be looking at something like this (to keep things simple and readable here, I’m looking at new customers in general):
Alexandre Macmillan 
One comment