Video Game Analytics 101: Basic Definitions

Posted by Dmitri Williams

Find me on:

Oct 4, 2013 7:02:00 AM

Video Game Analytics 101-2 Ninja MetricsHere are the basic metrics any basic dashboard should have, along with definitions. These are the kinds of metrics that you would hand to your CEO or board to give a high-level view of how things are going. They’re simple, and best seen in the context of time, so the smartest use is to see their period-over-period change in a table or on a graph. More complex metrics are for developers, marketing and community managers who will need to dig deeper into changes. All metrics should be chosen based on the questions your business really needs answers to - see our Practical Metrics blog post on the subject.

Also, all of these metrics are based on aggregates, meaning that they are averages across all players. That means they will always hide variations. So, they’re fine for benchmarks and reports, but they are never actionable at the individual user level.

DAU: Daily Active Users

This is how many users logged in on a given day, but make sure you understand what both “active” and “day” mean. A basic definition of active would mean the user logged in to the application at least one time. A more conservative definition could mean more than once or logged in and took some additional action. We use the basic definition. A basic definition of “day” will mean a 24-hour period, which is fairly obvious, but which 24-hour period? Standardized to the time zone of the developer, the player, or some fixed value? Because games are truly global in both development and play, we recommend and use GMT.

MAU: Monthly Active Users

“Active” can vary here too, and again we recommend any one login, but you could make the case for a stricter definition. Which month is straightforward: it should be 30 days leading up to the reference date. If the MAU is for today, that’s the last 30 days prior to midnight GMT yesterday. If the MAU is for the 15th, it’s the 30 days that ended at midnight GMT on the 14th. So, watching MAU over time is a rolling number. Some firms will average over a time frame.

ARPU: Average Revenue Per User

This is the total amount of spending divided by the total number of users. If your app is subscription-based, it’s going to be the same value as the subscription price. If your app is a free-to-play upsell model, then likely only a small fraction of users will be paying. In that case, the ARPU number is going to seem relatively small since a small number of dollars will be divided by a large number of users. Keep in mind that ARPU gives you no real sense of the distribution of that spending. It could be one person spending a lot, or many people spending a little. To dig deeper, you’ll need a gini index (see next post in the seires next week).

ARPPU: Average Revenue Per Paying User

This is the total amount of spending divided by the total number of users who spent. Again, if you are subscription-based, this is the same as ARPU. If you have an upsell model, ARPPU is always going to be larger than ARPU. It also tells you nothing about the distribution of spending--one person could have spent $20,000 and the rest averaged $2, and your ARPPU of $5 is going to give you a false sense of what’s typical.


This is a % value that tells you how well your application keeps users coming back. It needs to have a time range attached to it, so you will see 1-day Retention, 7-day Retention, 30-day Retention, etc., etc. Different companies handle time zones differently here. For example, does the 1-day retention reset for everyone at midnight, or is it custom for every player? Again, we recommend sticking with midnight GMT for simplicity’s sake. So long as you have the same definition every day, this is a good benchmarking metric.


There is an analytics company that markets K-factor as a super powerful tool. It’s useful, but like all of these statistics, it’s an aggregate and can get you into trouble. How should you use it? It’s an indicator of whether your users successfully invite their friends. A value over 1 means you’re probably growing. It’s just the number of invitations sent times their % success rate. That’s it. If you thought it was some black magic, go see the Wikipedia entry. A more powerful approach is to understand the social effects taking place among your users, which requires deeper metrics like Social Value and virality measures that require social network analysis.

DAU/MAU: Daily Active Users/Monthly Active Users

Daily Active Users divided by Monthly Active Users. This is one of the few basic statistics that starts to give you a sense of proportion. Imagine you have 100k MAU. If your DAU yesterday was 10k, then you had about 1 in 10 of your regulars online. If your DAU was 50k, you had about 1 out of 2 of them online (which would be unusually high). This metric gives you a small sense of the intensity and repetition of use.

Session Metrics

Average Session Length and Average Number of Sessions Per User. These are both self-explanatory. All you want to be aware of is what the time period is, and how these metrics are trending. If your average session length or number of sessions are going up, that’s usually good since it means there is more engagement. But what is going on with spending at the same time frame? If it’s also increasing, great. If it’s holding steady, you have no gain and possibly higher costs. If it’s decreasing, you may have a problem.

OK, those are the basics, and again, they are really only there for benchmarks and big-picture thinking. You can’t take action or diagnose with these. For those decisions, you will need more advanced tools which aren’t actually that complex. If you can wrap your head around these, you’ll have no problem with the next set. Read the next post on part 3 on Video Game Analytics.


Topics: Game Analytics, Video Game Analytics, Setting Up Game Analytics, Game Metric Definitions


Subscribe to Email Updates

Ninja Metrics Katana Demo

Ready for a live Demo of Katana?

If you have a social game or application, we can help you learn who matters most in your network and what features are affecting the health of your community