I was recently interviewed by InfoWorld for a segment on Android Fragmentation. Seeing that I get asked about this a lot I figured I’d post my thoughts about why a lot of Android fragmentation occurs in the Android ecosystem. The content below is an edited version of an email conversation I had with an editor at InfoWorld.
Question – Does Android Fragmentation exist? If so, why do you think it exists?
Yes, it definitely exists. We have a large fragmentation problem currently. What I mean by that is that the large fragmentation issue largely stems from manufacturers failing to upgrade existing devices most of the time. A lot of devices can handle the upgrades without issue (CPU and memory/etc and device properties can handle the upgrade fine) but the manufacturer won’t update them. Therefore we have a large variance of devices in the wild.
Why won’t manufacturers update existing devices? Well, I have to speculate (as I have never heard it from them directly) and deduce that the reason lies in profits and revenue. In my opinion, a manufacturer has no interest in updating the devices to the new version of the software unless is quite trivial. Unfortunately its not trivial. The reason for this is because a lot of manufacturers like to take android and apply their own skin (TouchWiz, HTC Sense, AT&T Whatchamacallit, etc). The manufacturer have to DEVELOP this theme/skin on their own dime, Google does not do this for them. So when a new version of Android comes out that completely changes a lot of UI paradigms, the manufacturers have to look at the business and determine how much it will cost to upgrade their theme so they can push out an update to their existing Android customers. My guess is that the cost of the development time to upgrade their themes is quite large and the return is small. Think about it – What additional revenue will a new updated OS bring? Probably something very very very small (if any at all – most likely none at all). So the manufacturers develop new devices and market those as the “new best thing” because the new OS is present on them. If the user wants the new OS, they need to upgrade. They need to buy a new device. New devices are revenue generators. The cost is already built into the new device development and marketing and profit margins are higher when someone buys a new device vs the manufacturer updating an OS and sending an OTA (Over the Air) update.
Now… if the device is a Nexus style device (Galaxy Nexus, Nexus 7, Nexus 4, etc) … these are all what I like to call Vanilla Android devices. There have been no modifications (other than getting it set up for the carrier) to the theme/etc. So the cost to push new updates is radically different than those of other customized devices like the SII or SIII, HTC One, Hero, etc. Since the first Android device hit the market I’ve owned an array of Android devices (over 25). Of all the devices I’ve owned, the “Nexus” models out perform the other manufacturers devices and always receive updates months in advance of others and usually continue to receive updates after other capable devices are not supported anymore (because of the speculation noted above).
So … at the end of the day we have a fragmentation problem because of business revenue and business decisions, not because the devices can’t handle them. Sure there are exceptions – not all devices can be upgraded – but you get the point that I’m getting at.
At the end of the day it comes down to this: Money In is not equaling money out. Result: Fragmentation
Note: This is not the only reason fragmentation exists or the only variance. We also have various screen sizes, resolutions etc to deal with. This article mainly talks about the variance in OS versions in regards to fragmentation.