Real Cases

How to understand your application performance

We recommend you to watch this webinar. You will know how to better evaluate the quality of your application performance, how to find the problems and the bottlenecks, how to accurately use benchmarks when making decisions.

Segmentation by Custom Event

There was a hypothesis in one project that the tutorial shouldn't be obligatory for new users.

To check it, the analysts created two custom events: PressingTheSkipTutorialButton and FinishingTheTutorial. Then they applied the segmentation by custom event and divided all the new users to two segments (only the users who started the tutorial but did not finish it were excluded from the analysis).

Then they calculated the day-7 retention and ARPU values for both segments. And they found out that both metrics were 50% higher for those users who finished the tutorial.

So, the tutorial was made obligatory for all new users after that.

More cases about segmentations are covered in our educational webinar. 


ARPU (average revenue per user) is the more important KPI than ARPPU (average revenue per paying user). ARPU takes all active users into account, and ARRPU takes only paying ones.

One project was aimed at ARPPU maximization. To grow the ARPPU they increased the prices by 10%.

ARPPU increased proportionally. But this change decreased the paying share for the project significantly, so the ARPU and revenue started falling. And they had to roll the prices back.

Metric Before After
ARPU $0,2 $0,17
ARPPU $10 $11
Paying Share 2% 1,5%
Daily Revenue $200 $165


Tutorial Optimization & Retention Increase

Project A wanted to increase the long-term retention. The easiest way to do this is to increase the short-term retention (for example, the day-1 retention).

The first thing you need to do when you optimize the short-term retention is to check your tutorial and to find the bottlenecks there. So they used the "Tutorial Steps" report to find the places where users could have some problems and decided to leave the application. 

They found two steps where the conversion decreased dramatically. The reasons were only technical, and they fixed it successfully. So the tutorial conversion increased, then the day-1 retention increased, and after this the day-28 retention increased proportionally.

Metric Before After
Tutorial conversion
(% of users who started the tutorial and then finished it)
53% 77%
day-1 retention 18% 23%
day-7 retention 9% 13%
day-28 retention 6% 9%


No next