How to analyze paying users. Part 2, Structure of income by time

Last time we talked about how you can segment paying users, recalled RFM analysis, as well as whales and dolphins.
This time, we will also use segmentation, but on a completely different principle. Have you thought about the structure of your income? Who brings more - beginners or oldies? What is the ratio of income from new and old users, how does it change over time? That's about it and talk.

Audience structure in general
First, we divide our entire audience (both paying and non-paying) into several segments in time from the time they are registered. How many segments to allocate is solely your business and depends on the specifics of your business and the period of involvement in the project.
In any case, we would recommend limited to 5-7 segments.
For example:

• 1 segment - less than 14 days from the moment of registration;
• 2 segment - from 14 to 30 days;
• 3 segment - from 1 to 2 months;
• 4th segment - from 2 to 6 months;
• 5th segment - from 6 months to 1 year;
• Segment 6 - more than a year from the date of registration.

By highlighting custom segments, you can build a report on the structure of your audience at the time of analysis.

What you can see in this report:

• If newbies predominantly prevail, you have trouble holding. The project can not keep users for a long time. So, either work on holding or think how to monetize the beginners (for example, to make the application paid).
• If old men noticeably prevail - this is not good either. Are you all right with new registrations? Maybe worth buying some traffic? Remember that the more users, the more users. And on some oldies you won't get far, sooner or later the application will start to go down in the rating.

The next stage can be considered not only the structure of your audience, but its dynamics - how this structure has changed over time. Usually it is at this stage that the most interesting is found out.

Structure of paying audience
Let's do the same manipulations, but now exclusively for the paying audience. For example, refer to the “ Users & Gross structure ” report in the devtodev system.

This example shows how the stability of the size of your paying audience hides the pitfalls, and the growth of one segment is offset by a decrease in the other segment.
We see that the percentage of newbies (up to 30 days from the moment of registration increases), and the percentage of old people (from 6 to 12 months from the moment of registration) falls. And without considering the structure, we could not notice.
A sign of a healthy application is that the segment of old people must at least slowly, but grow - more and more users must reach the most adult segment and remain in it.
')
Income structure
Finally, income can be analyzed in the same way, cutting it into segments according to the time from the moment of registration of users making payments.
In the income structure report, all distortions in favor of old people or beginners are usually more pronounced. The fact is that usually (in projects based on long-term retention), beginners have a small average check, and old people have a big one.

As we see, the income in our example has a downward trend (remember that the size of the paying audience was stable at the same time). And the decrease in this trend is primarily due to a decrease in income from old people. Until the green segment, inclusive, there is some stability, then begins to fall.
Our verdict on the considered project is that the project has problems with payments from users registered 3 months ago and earlier. It is necessary to optimize the long-term retention of the project so that the natural influx of users in the last segment exceeds the natural outflow.

Math modeling
Having the above reports, you can make a mathematical model for predicting your income for several months in advance.
What you need for this:

1. estimate the size of each of the selected segments;
2. for all segments, calculate the probability of transition from segment N to segment N + 1 (what is the probability that the user, active during the month, will become active in the second month?);
3. calculate the average revenue per user (ARPU) of each segment.

By combining all calculated values ​​into one model, you can model how your audience structure and income will change in a month, two, three, six.
Moreover, this model will allow you to calculate various experiments with traffic and monetization.
Examples of questions she can answer:

1. What if I disable paid traffic and stay only on viral? How will this affect my income in 12 months?
2. If I optimize retention (for example, 30 days retention) by 2%, how will this affect the structure of the audience and income?
3. I'm going to make changes in the balance of the game and thereby raise the average check of users of the eightieth level (which is achieved on average after six months of the game) by 10%. How much interest will my income change?

And so on.

With this article, we wanted to convey to you one simple idea: it is important to study the structure of your audience and income over time from the moment users register. This will help you make more informed and effective decisions, be it marketing, monetization or game design.
Soon expect another article on the analysis of paying users.