If you want a successful game, you’ll need to create long-term relationships with your users. A few logins aren’t enough. You want them to play your consistently so that they’ll see it as a worthy investment.
If a player doesn’t return to the game, the relationship is over. They won’t get to experience your elder content or the exciting events you’ve added to your live ops calendar. So, of course, it makes it harder for them to convert into paying customers.
Along with conversion rate, retention is one of the most important metrics for a F2P game.
You already know retention is important, but perhaps you aren’t aware of exactly how massive an impact it can make in your business. But before you can boost your retention, you have to know how to measure it. In this article, I’ll explain how to measure retention in F2P games.
A Quick Primer on Retention
Simply put, retention measures the game’s ability to keep players returning. It tells you how many new players become recurring players.
Retention is a lagging metric. That means it reports on what happened, but it doesn’t tell you what you can do to increase or decrease it. The key method to get value out of your retention data is to measure it at different points so you can see how your changes affected it.
It’s similar to your weight. The number on the scale doesn’t tell you much unless you compare it to yesterday’s number or tomorrow’s number.
Retention does not measure what the player does in the game. It can’t tell you which part of your game players like most, or if they’re likely to become customers. You’ll need other metrics to gather that information.
Retention is sometimes used to describe an entire phase in the player lifecycle. You need to solve retention before your players are going to become reliable customers.
Before we dive into percentages, you need to understand how to conduct a cohort analysis. It’s a helpful way to understand retention’s role in your F2P game.
A cohort analysis is a method of breaking down data into related groups for examination and comparison. In this case, we want to break down users based on when they downloaded the app. This will help you track user engagement over time. (You can create cohorts around other characteristics, but we’re talking in the context of retention.) If you don’t perform this kind of cohort analysis, new users could pollute your understanding of your retention and you won’t see clear retention trends.
Calculating retention across all of your users isn’t granular enough to inform your decisions. It’s important to break your users into cohorts based on acquisition date and track their retention every day. Or, at the very least, you should calculate retention on Day 1, Day 3, Day 7, and Day 30.
Download day is “Day 0.”
Did you catch that? Because this is something fundamental many experienced folks aren’t clear about. We start counting at 0, not at 1!
Day 1 is the game’s first opportunity to retain the player, even though it’s technically the player’s second day with the app. (And if that’s confusing you, please stop and reread above until it makes sense!) If a player opens the app on Day 1, they were retained on Day 1. If they don’t open the app on Day 1, they won’t show up as retained on that day.
Check out the following retention table. This tracks different cohorts every day for 10 days after first launching the app.
Each row represents a cohort based on the people who downloaded the app in a particular day. On January 27, for instance, 1,257 people opened the app. Each column after that represents daily retention: The percentage of users who used the app from that group each day for 10 days. The bottom row refers to all users. It averages the retention rates of each day.
This analysis helps us track user behavior over time. If you look horizontally, you’ll see that fewer people from each cohort play each day. This gives you insight into the quality of your game loop, your operations, and your new content.
If you look vertically, you can see how the game performs over time. For instance, you might ask why more people who downloaded the app on January 25 stuck around than people who downloaded it on January 31. What changed between those days?
Why do we break it down by cohort? Since a cohort is here defined by acquisition date, we can measure how game changes affect retention. If you make a change to the onboarding sequence, for instance, you can see how it influences whether people keep playing.
Why do we track retention at different intervals? Because they speak to different phases of the user experience. Day 1 retention is really just a measure of your onboarding and first time user experience. Day 30 retention, however, speaks to your players’ engagement and enjoyment.
Now that you understand how to analyze retention over time, let’s dive into how to get those retention percentages. There are a few ways you can calculate retention, and it can make a BIG difference. Plus, you may want to compare your retention with other benchmarks, and you’ll want to make sure you’re comparing apples to apples— or androids to androids.
1. Classic Retention
Calculating classic retention is pretty straightforward:
Day X Retention Rate = Daily Active Users on Day X / Cohort Size
Let’s say we want to know well we retain users on Day 7. We would divide the number of Daily Active Users who are on their eighth day of play (remember download day = Day 0) by the number of people who downloaded eight days ago.
Let’s use some fake numbers: On January 5, 785 people downloaded the app. On January 12, 164 people opened the game. In this case, the Day 7 retention rate for that cohort is 20.1%.
2. Range Retention
Range retention is classic retention over a given time period. It’s a useful way to smooth out the noise of day-to-day retention figures and look at retention over long periods of time.
To calculate range retention, you simply determine how many people from a given interval return during a subsequent interval. You can use any interval for range retention, but 7-day and 30-day ranges make the most sense.
For example, let’s say 125 people download your app during the first 7-day week of March and 24 of them play again during the second 7-day period (which is the first retention period). In this case, your range retention rate is 19.2% for the first retention period. If 8 people return during the third week of March, your range retention is 6.4% during the second retention period.
There are two limitations here. First, weekly or monthly granularity means you don't know if retention happens at the beginning or end of your interval. Second, you have to wait longer to get data.
3. Rolling Retention
One problem with the classic retention calculation is that it only measures players who return on the day you measure. If someone played on Day 8, they wouldn’t be measured in the Day 7 retention metric. But that doesn’t mean they aren’t loyal players.
Rolling retention is a method of calculating retention over a period of time. It’s similar to classic retention, but it includes people who play on your interval day or later. So if you’re measuring Day 7 retention, a rolling calculation includes users who return to the game seven or more days after downloading.
This means that rolling retention will always be equal to or greater than your classic retention. This example shows how they’re usually related.
What is “Good” Retention?
Generally speaking, a 40-20-10 retention profile is considered successful. That’s when your Day 1 retention is 40%, your Day 7 retention is 20%, and your Day 30 retention is 10%.
That said, even though I hear this question a lot, it somewhat misses the point. If your game is at scale, the goals will be different than if you’re launched to a small group of super-fans. If your game is casual, it’ll have a different retention profile than something more hardcore, etc…
Retention is one of those metrics you should always be looking to improve. You may have the highest retention rate in the industry, but that doesn’t mean you should be happy with it.
Like your conversion rate, small tweaks to your retention rate can have dramatic impacts on your game. But first you have to measure it and analyze it correctly to do so. Hopefully this guide will help you gain a clear picture of how well you’re retaining customers.