The Localytics Blog

Smartphone OS Wars: What platforms to develop for? Part I (Business perspective)

As the smartphone wars rage on, mobile application developers regularly face a common dilemma: what platforms should I develop on? After some research and our own analysis we’ve deconstructed the answer into business (Part I) and technology perspectives (Part II).

Part I: The Business Perspective

iPhone, the reigning application king has 50,000+ apps, largely driven by the simple developer user interface and first-to-market App Store.  The iPhone has surpassed the expectations of many with an estimated 15M (UPDATED: Note, does not include ~40M iPod Touch) users worldwide and approximately 14% of shipped smartphones.  Apple will continue to foster the growth of its App Store, not because it’s a profit center on its own, but because Apple is leveraging the App Store to drive iPhone sales (emulating what iTunes did for the iPod).  However, iPhone’s premium positioning and exclusive-carrier approach leaves plenty of room for other players. While many developers are starting out with the iPhone they may quickly find themselves developing on…

…Android. Taking an open source approach, we project Android to grow at least 10M users by this time next year.  The G1 is shipping across multiple carriers in five new markets (France, Germany, Spain, Italy, Australia), and there are 18 upcoming Android handsets from almost all major manufacturers (Acer, HTC, Huawei, LG, Motorola, Samsung, & Sony Ericsson).  Handset makers and carriers need help fighting proprietary smartphone OS’s like iPhone and BlackBerry, and are hoping Android is their answer.  A key concern for developers is the wide variety of hardware expected (not even considering Android on netbooks and alternative devices), particularly in the US where carriers are seeking to differentiate Android handsets to gain an edge. While consumer apps will compete directly with iPhone, hardware tailored to a vertical market may create app competition with…

…BlackBerry. With clear appeal to its 30M strong targeted user base, BlackBerry is limited by its clumsy App World interface (particularly the app buying process), less sexy UI, generally more expensive apps, and the millions of BB users on pre-Pearl (late ‘06) that can’t gain access. Still, for the millions of BlackBerry users unfamiliar with iPhone App Store (and actually aware of the App World), this is a dramatic improvement.  In listening to developers’ frustrations with the iPhone App Store, Blackberry is committed to turning around application submissions in less than 2 weeks, a service found only on BlackBerry and…

…Symbian.  The current worldwide smartphone market leader with an estimated 50M users, Nokia’s Symbian has been losing market share quickly in recent years (62% in 2007, 49% in Q1 2009). Nokia’s Ovi stumbled into the app marketplace in late May, ‘09 touting 20,000 ‘items’ but today offers only about 1000 apps. Developers seeking a global user base, particularly segments within EU and Asia ought to consider Symbian, but others may want to hold out until they are more like iPhone/Android and less like…

 
…Windows Mobile.  Working closely with a core set of app developers now, Windows Mobile is far behind in the app store phenomenon. There are 20,000+ WinMo apps available via web, about 30M users, and the money to make a big splash, but it is too early to project.  Windows Mobile 6.5 is supposed to be greatly improved, and their app store, Skymarket, is promised to launch later this year, but it will be interesting to see if this is enough for them to get back in the game.

This article is continued in Part II, “A Technical Perspective”...

Thanks for sharing this info. Love to see an update down the road to include Pam’s WebOS/

FYI: The iPhone platform has over 40mil devices include the touch devices. For app developers this is a non-trivial additional user base as people who have non-AT&T phones may still get a touch device.

Great point Greg - updated the post to reflect this.  We’re currently trying to get data on the rate of app downloads/usage for iPhone vs. iPod Touch.

Thanks for sharing this info. We are currently developing Apps for various smartphones. This gives us a better picture of where to target first.

Thanks,
antloc Team
http://www.antloc.com

How does something like http://www.phonegap.com tie into being able to make one app and run it across all of them (or atleast make it available across platforms where it otherwise would take a large investment to develop natively on)

Mike - PhoneGap, rhomobile, and all of the other combined frameworks are a great idea.  However, they aren’t yet receiving tremendous adoption because typically for an application to be successful it needs to be tailored for specific behaviors on each platform.  For example, on the iPhone applications are not provided with access to the contact list and BlackBerry apps tend to make heavy use of the scroll wheel.  So if your application uses the contact list it needs to be redesigned to work around this limitation, and if your UI is targeting a BlackBerry it won’t be appropriate on a Windows Mobile device.

As these frameworks mature these problems will be addressed, but currently it feels like using these frameworks causes the developer to end up working on all platforms but not having the amazing experience required to fare well in the application stores. 

Another interesting variation on the theme is Flash.  This is, after all, one of the best selling points of flash and if it becomes ubiquitous across all platforms there could easily be a large amount of flash development.

Great post and good pointers for clarifying the often treacherous landscape out there.

One thing that I’ll share with everyone is that for Symbian, Windows Mobile and Blackberry, the Developer Revenue Share can be substantially lower than the figures in your figure, depending on:

1. type of sales channel utilized; and
2. discounts offered at the channel level

For example, Handango, an early and still popular appstore, has a Rev Share of 50% _before_ discounts. This leaves substantially slimmer margins for end developers.

kl

Hi Kenny - this assessment focused on existing or upcoming app stores only, which is where those rev shares come from.  While some of the app stores still have a long way to go, we believe they will become the primary means of distributing a BlackBerry/Symbian/WinMo app because they make the sales process so much easier.  However you’re right, if dealing with third party sales channels you’ll get significantly less revenue share.