Performance Indicators: Cumulative ARPU

What is a cumulative ARPU, how does it differ from a traditional ARPU, how to calcualte it and why it is needed. Another translated article of Vera Karpova, analyst at devtodev.

In this article, we’ll look at a metric that will be very useful for assessing traffic quality and selecting the optimal CPI metric. This is a cumulative ARPU.

This metric is calculated exactly the same as the usual ARPU, by dividing income by the audience, but it has one very important feature that distinguishes it from usual ARPU and makes it useful for traffic analysis.

The accumulative ARPU is calculated not for the entire audience, but only for a certain cohort of users, and this cohort is formed from users who installed the application in a certain period (it is most convenient to use day or monthly cohorts).

Another feature of this metric is that it increases every day for one cohort (not for nothing that it is called “cumulative”).

cumulative arpu chart

How does this happen?

Let’s say we decided to investigate users who installed the application on 01/01/2017, say, 1000 people. On the first day they made certain purchases, and the revenue from these purchases was $ 800. ARPU of this cohort of the 1st day was $ 800/1000 = $ 0.8

We begin to maintain the table and insert all available data.

cumulative arpu table 1

The next day, some of the same users made payments totaling $ 500. No matter how many of them returned to the project and how many people paid, the size of the cohort to which we will divide the income is always 1000 users.

And now you can count the cumulative ARPU. It will be equal to the sum of ARPU day 0 (installation day) and day 1.

cumulative arpu table 2

In the same way, we calculate the remaining days.

cumulative arpu table 3

Now you can output the final formula, which calculates the cumulative ARPU.

Cumulative ARPU for N days = Revenue (selected cohort for N days) / Number of users (within the selected cohort)

The accumulative ARPU of the Nth day for a particular cohort is equal to the income that this cohort brought for N days, divided by the number of users in this cohort.

This metric can only grow with time, because the more time users spend in the project, the more they pay. And if at some point they stop paying or using the product in general – this chart, having reached a certain level, will go to the horizontal line.

What is interesting about this metric?

Firstly, knowing it, you will know how much money the user will bring for 7 or 30 days in the application, or in 3 months. That is, you can count on when and how much money will bring you a new user.

Hence its second advantage follows. This metric is very useful in assessing the quality of traffic and its payback period. It is important to know when the user starts to bring more money than you spent on attracting him. This will be the moment when the cumulative ARPU becomes equal to CPI (Cost Per Install).

And if you know what a cumulative ARPU cohort, for example on 365 days (or on any other day after that), you can already talk about the average income per user for the lifetime of the project (this is the Lifetime value metric, LTV or CLV).

And one more advantage of this metric: due to the peculiarities of calculation, the cumulative ARPU allows you to compare different cohorts. Which is especially convenient when you made any changes in the product.

In this case, you can compare the dynamics of the cumulative ARPU for the cohort that installed the application before it was modified, with those who installed after to see how these changes affected the payments of users.

Thus, the calculation of the cumulative ARPU is a big step towards cohort analysis, a good indicator of the quality of the purchased traffic and just a useful metric for assessing the nature of user payments and their profitability for the project.

Articles in this series:

  1. Performance Indicators: ARPU
  2. Performance Indicators: ARPPU
  3. Performance Indicators: Cumulative ARPU
  4. Games performance indicators: paying users
  5. Performance indicators of games: paying conversion
  6. Performance Indicators: ROI

Even more interesting info about the gamedev industry:




Site: :


Leave a Reply

Your email address will not be published. Required fields are marked *