Android Users Are Stickier Than iPhone Users


No, no – that headline wasn’t intended as commentary on the hygiene of Android users (though if a good chunk of the Android devotees I know are any indication, it very well could be. Zing!) Earlier this morning, mobile analytics group Flurry gave us an exclusive sneak peek at their Smart Phone Industry Pulse report for June. Flurry’s June report harvests data from 1,100 applications running across 4 platforms (iPhone OS, BlackBerry, J2ME, and Android) on over 40 million handsets, and sheds a bit of light on the usage habits (stickiness included) of smart phone users over the past few months.

Though Flurry offers statistical analysis for the four aforementioned platforms, adoption of Flurry’s statistical analysis agent (the bit of code that developers add to their app to enable Flurry to track usage) has proven to be substantially more popular on the iPhone and Android platforms. As such, much of the most interesting data from this report comes from these two operating systems.


Before we dive in, a quick look at Flurry’s sample group: of all the applications using Flurry’s stats agent, 79% are iPhone applications. The next largest group is Android, at 16%. BlackBerry and J2ME tail far behind, at 4% and 3% respectively.

Interestingly, the number of Flurry-enabled J2ME apps has seen an incredible drop over the 3 month period, while the iPhone has seen an equally huge growth. Don’t take this as if all of the J2ME developers are switching to the iPhone, however; the iPhone has seen a ton of growth from developers new to the mobile realm.


Nothing too surprising in the consumer usage numbers. The app numbers above already indicated that Flurry tracks far more iPhone apps than any they do for any other platform. More apps tends to work out to more users, so the iPhone absolutely dominates here. Here too, J2ME has seen dramatic shrinkage over the past 3 months.

Here’s where it gets interesting: by taking the two most heavily represented platforms (again, Android and iPhone) and comparing user loyalty on an App-by-app basis, Android wins hands down. As time increases, Android users continue to stick with apps longer.


At 30 days, the proportion of Android-vs-iPhone user retention is roughly 31% higher. By 90 days, it’s roughly 42% higher. That’s huge.

Here’s what Flurry chalks it up to:

One reason we believe retention rates vary is that Android offers far fewer applications compared to iPhone. With applications coming out on iPhone at a faster rate, iPhone users move onto other apps more quickly. For Android users, they make more use of what’s available, with less temptation to move to the next application. Other factors that could also play a role:

  • The Android base tends to be “older,” have less time and interest to try new applications. Once they find an application they like, they stick with it.
  • The Android base is more tolerant, tend to be more tech savvy and find ways to appreciate what they have, even if their applications aren’t perfect.

While I can’t speak for the latter two, I’d imagine that the first theory is spot on for what’s going on here. Apple’s selling the App Store to people under the idea that for any given need or interest, “There’s an App for that.” It ought to be “There are 14 apps for that”. For nearly every app, there’s an alternative or ten waiting to drag people away. It’s rough for developers, but great for the end user.

Not only are Android users using their applications longer, they’re using them more often:


37% of iPhone users use the applications they downloaded less than 5 times per month, while only 11% of Android users were in the same category. Flipping that around, roughly 35% of Android users are tapping into their apps more than 50 times per month, while only 15% of iPhone users are doing so.

Again, the reasoning behind this is likely the same as above; fewer alternatives means more usage of the apps that are already there.

If you’re a developer looking to determine which platform to port your application to next, take these numbers into mind. By the end of 2009, Android will be on enough handsets around the world that the cumulative user count may make the platform much more viable for development. If there’s an Android Market gold rush even half that of the App Store’s, you don’t want to be 25 thousandth app to go live.