What language for your mobile app?

5 years ago the choice for a company creating mobile applications was easy.  The clear global leader in mobile was Java (J2ME).  If you wanted reach you needed to create your app in Java, and port to the different platforms and devices.

In some regions there were specific leaders - BREW, led by Verizon in the US, with support in Korea and Latin America.

Smartphones were a segment of their own - with their own solutions, and a self selecting group of high value consumers who could afford handsets, and would expect appropriately high quality applications that could be achieved with access to native SDKs

The mobile market going into 2010 is starting to look a little different, and challenges the accepted wisdom that Java is best for reach.

As we enter 2010, we have even more fragmentation in the mobile platform with more devices coming out with different platforms, platforms supporting multiple development environments.

I expect that by the end of 2010 a mobile developer who is creating "mass market" applications will need to consider the platforms below that will all have greater than 10% market share in countries in Europe

• J2ME - for "feature phones" from Sony Ericsson, Samsung, LG and Nokia
• Symbian - for "smartphones" from Nokia
• Android - for "smartphones" from HTC, Sony Ericsson, Motorola
• LiMo - for "smartphones" from manufacturers to be revealed
• OS X - for "smartphones" from Apple
• Blackberry - for "smartphones" from RIM

There are also going to be some "also ran" platforms - Windows Moble, Palm's WebOS - but at least 60% of the market share is going to be split between 6 platforms, some of which force use of Java, some force use of native code, and some allow choices of the two.

So any application developer is going to need to consider creating at least one native application, one java application, and might choose to take advantage of other technologies (Web Runtime, Flash) that provide an advantage for a particular application.

This significantly increases the cost for any application developer who aims to reach the mass consumer market on mobile, compared to using desktop & web technologies where there are fewer competing platforms and fewer differences between the platforms.

The installed base of PC's globally is around 1 billion, by comparison the number of mobile phone users globally is around 4 billion.

If you want to deliver a new service to the users of those devices you need to look deeper however.

The global number of unique internet users is estimated as 1.6 billion - each PC worldwide is shared between 1.6 people.

In much of europe mobile penetration rates are typically 120% - so by comparison the average phone is used by 0.83 people.

So a developer targetting the "middle class" economies, where mobile, PC and internet usage is growing, and there is disposable income to spend on new applications and services has a tough time - 1.6bn people can be reached using web, managing fragmentation across browsers.

To achieve the same in mobile the developer is going to have to manage at least 6 platforms, where each platform has it's own fragmentation similar to supporting the PC browser market.  While the market is still 2.5 times the size, it's 6 times the cost.  This is compounded by the fact that price expectations are lower in the mobile market than in the PC market.  The most common price point in Apple's App Store for the highest grossing applications is $4.99 - a fraction of what is achieved by the massive revenue generating applications such as Microsoft Office in the PC arena.  

The economics for mobile developers are already tough - and will continue to get tougher in 2010, but it's not all bad news - as the smart developers in mobile will be able to achieve scale beyond the PC, and will be able to reach emerging markets faster than the PC will.  The simplicity, convenience and low cost of mobile devices create new opportunities that the PC can't reach - but 2010 will continue to be a year of investment for mobile application developers.

Comments

[Andrew Braun]

I agree with your numbers but not your conclusion. The big changes that the new operating systems have brought is access for the small software developer. A small team can build a application that now has prominent "shelf space" and sell to the masses. A small team can create a niche application that takes advantage of open APIs and sell for high price. You are correct the pricing will not be the same as MS Office, but be honest only microsoft has that massive revenue. That market was closed long ago.

The economics are different but I believe it is getting better not tougher. More at http://bit.ly/BZDsZ

[Matt]

The long tail does indeed create profitable niches - unfortunately the mobile market fragmentation breaks the economics too. If you can create a "long tail" book that you can sell on amazon in English, then you can be confident that everyone will be able to read it. In the mobile market in 2010 you'll have to write the book 6 times - this cost impact limits the scale of the overall market opportunities.

I do agree it's possible for niche businesses to be profitable - but the market scale limits the number and size of these, and there are few very profitable mobile application businesses.