For most of the ten years I have been a VC device fragmentation has be-deviled innovation in the mobile internet, and the iPhone’s success since 2008 came in large part because for the first time there was enough users on a single mobile device to attract a flock of application developers. However as you can see from the table below, the iPhone is only a small part of the installed smartphone base and if mobile is going to deliver on its promise those iPhone developers will need to be able to get their apps on the devices that the majority of the market will be using.
Note that this market share is based on device sales and is different to the much publicised Admob survey which is based on traffic. The iPhone and Android rate much higher on this measure as the devices get used more. I chose to show the device sales data because that is a better indication of how fragmented the market is today.
I’m writing this post after reading about fragmentation within the Android platform, I had heard some mutterings about different flavours of Android for different operators and the challenges of working with multiple operator controlled app stores. Engadget described the problem like this:
the kind of fragmentation that has already left users running not one, not two, not three, but four distinct versions of the little green guy (1.5, 1.6, 2.0, and 2.1) depending on a seemingly arbitrary formula of hardware, carrier, region, software customization, and manufacturers’ ability to push updates in a timely fashion
There is good news for the Android platform though, and that comes in the way Google are talking about resolving the problem (again from Engadget):
We’ve been given reason to believe that the company will start by decoupling many of Android’s standard applications and components from the platform’s core and making them downloadable and updatable through the Market, much the same as they’ve already done with Maps. In all likelihood, this process will take place over two major Android versions, starting with Froyo and continuing through Gingerbread. Notice that we said apps and components, meaning that some core elements of Android — input methods, for instance — should get this treatment. This way, just because Google rolls out an awesome new browser doesn’t mean you need to wait for HTC, Samsung, or whomever made your phone to roll it into a firmware update, and for your carrier to approve it — almost all of the juicy user-facing stuff will happen through the Market.
I like the way this removes operators and device manufacturers from the upgrade path. I can also see consumers getting prompted to download relevant components when they need to much like a web page asks if you want to download a new plugin if it is required to render the page properly or show a video. This model works extremely well on the PC and could be equally powerful on the mobile, particularly as individual developers could contribute to Android to make their apps work better, should they want to.
I guess I like this story because I’m looking for reasons to believe the device fragmentation problem will go away, both from technology based solutions for developers and from OS/value chain based solutions like the one described here.