Crunching some numbers

Last year I was reading David’s blog and he mentioned he found some strange spikes and drops in the install base of his apps. These spikes were from reviews or other promotion activities. At that time I did not had any apps in the Android market (yes now it’s the Play Store but then it was still the Android market. But I found it interesting to test some scenario’s and visualize them.

Please take in regards that I did not have any knowledge on how many people would keep an app installed, and so I used a 70% ratio. Now I know it’s more likely to be near the 35%-40% range. But that does not matter that much. These graphs still give a good indication on why and how to read your stats.

Control
This shows a steady increase of users.
It calculates the total active install base, that is 10 new users every day of which 70% will keep the application installed.

Peak by exposure
This shows a steady increase of users but with one peak, this exposure peak will give the application 4000 new users in one day.

Result is a high peak and a smaller dip in the days after the peak, this last 14 days (this is the number of day’s in this calculation it takes for everyone to remove the app that does not want it.) After the cool down the rising trend will resume as normal.

This dip is caused because after the peak the absolute number of people that uninstall in the first day’s is higher than the amount of new users normally installing the app.

New phone
This shows a steady increase of users.
It calculates the total active install base, that is 10 new users every day of which 70% will keep the application installed.

But after two years the phone is replaced by a new phone. This sheet shows the effect if 90% of the user will buy a new phone after two years without re-installing the app on a new phone.

The effect is that after two years of steady growth a critical point is reached. After this point growth will slow down dramatically.

Growing Android market
This graph is the same as the new phone graph but with one key difference. The market for Android apps will be filled with new apps, this will give the app less exposure and lesser new installs every day. We take a decrease of new installs of 0.1% every day.

The result shows that after the critical point of two years has been reached the active install base is declining. This is because the growth is shrinking and is lower than the number of people buying a new phone and not re-installing the app.