Product Management

Update and perfect your app by getting a better understanding of your user base

Understanding Your User's Retention

Sensor Tower's Retention feature within Usage Intelligence, you can understand the percentage of users that are using an app on a regular basis. Furthermore, you can also gain insights into the relative health of an app by understanding how to identify a good retention curve vs. a bad retention curve.

Looking at Half-Life Retention

One heuristic approach in understanding retention is to look at the half-life period of a retention curve. The half-life retention can be understood as comparing half the period you're looking at versus the end of the period and looking at the drop off (difference) between the two. Apps with good half-life retention show very minimal dropoff, while those with bad half-life retention curves will indicate more significant dropoffs. Regardless of the date range picked - whether you're looking at 15, 30, 60, 90 days, the same concept applies across all ranges.

In the example below, I have compared the Retention curves for three different apps - Fire Emblem Heroes, Lucky Day, and Solitaire.


Looking at the Day 0 retention rates (day of installation), all three apps are hovering around a 60% retention rate. However, their behavior drastically changes when you look at 15 and 30-day retention. Which of these three curves do you think has the best 30-day retention? 


Now looking at the 15 Day retention, it's pretty evident that Solitaire exhibits lower retention as the curve sharply dropped from 61% to 4.3% in just two weeks! In other words, they lose about 93% of their users who installed the app on a given day. This curve is an example of what a bad retention curve might look like.

Note: Due to the nature of some game categories (i.e Card) where the general player base is more casual, this behavior can be expected for those genres. It does not necessarily mean that the game itself exhibits poor retention.


On the other hand, a good retention curve would look similar to Fire Emblem Heroes where there is minimal dropoff between the 15 and 30 days (or even 30 and 60 days) retention curves, as shown below.


From Day 15 until Day 30, Fire Emblem Heroes exhibited a dropoff of only 5.75% and even less between the 30 and 60-day mark. They are better at retaining their users as they've only lost about 20% of their userbase between the 2nd - 4th week of use. "In an ideal world, and overall retention rate of 30% or higher is desired for all date ranges" according to Hiten Shah, and Fire Emblem Heroes comes closest to achieving that goal after 30 days.

These are just some examples to understand half-life retention. It's a heuristic method to quickly obtain a general sense of an app's retention and longevity.


For more information or if you have any questions, please email

Was this article helpful?

Have more questions? Submit a request