About a month ago I released a new version of the 10,000 app. It contained a few things but nothing ground breaking. Nonetheless I had some assumptions about how fast users would upgrade their app, more precisely I expected that 80 to 90 percent would be running the latest version within 30 days of launch.
At the time of writing a little over a month has passed and I have some numbers to reflect on. With a total of about 10,000 active device installs (totally coincidental :)) I think there’s some basis for statistics, which I find interesting.
EDIT: I made a miscalculation the first time I ran through the numbers. I have adjusted this and some of my conclusions has been re examined and adjusted accordingly.
After 1 day: 15.5 % had upgraded
After 2 days: 28.9 %
After 3 days: 33.7 %
After 4 days: 36.4 %
After 5 days: 38.9 %
After 6 days: 40.0 %
After 1 week: 42.1 %
After 2 weeks: 48.4 %
After 3 weeks: 52.6 %
After 4 weeks: 55.9 %
After 1 month: 56.8 %
What I find interesting is two things.
- The “curve” flattens quicker than I anticipated. After recalculating the numbers it’s not cut in half every day, but the curve still rapidly flattens
- A little over half the devices have been upgraded one month after the release.
I’ll be wary of what to conclude based on this, but for future reference I’ll just note that getting a new version out to all users/devices is going to take a lot longer than I first anticipated. If only 3.3 % upgrades every week from now, which is the difference between week 3 and 4, it’ll take more than 4 months total, for all devices to get upgraded.
Just something to think about 😉