Most mobile marketing startups hedging their bets by developing apps for multiple platforms

Moconews has an interesting article about the challenges of multi-platform mobile development (iPhone, Android, Java, Blackberry, Symbian, Windows Mobile, etc) today.

The trade-off between delivering an amazing user experience to 2-3% of mobile subscribers (e.g. iPhone and Android) or a mass market application for a wide range of devices (50% or more) is a difficult one for everyone. Our general experience is that Java applications delivered across Symbian, Blackberry, Windows Mobile and proprietary operating systems get up to 10x as many downloads as an iPhone app but achieving a similar user experience is very time consuming and costly.

The average cost of an advanced iPhone or Android game/app is in the region 25-50k euro depending on complexity whereas as a mass market app/game with 80% phone penetration (across multiple platforms) is about 200.000-400.000 euro when fully optimized across the devices.

So what about automatic porting services then? There are several companies that claim that they can deliver cross-platform porting at costs starting around 10.000-20.000 USD. Our experience however is that the tweaking and optimisation to make the ported application to look and work nicely on the individual platforms make the cost almost as high as doing proprietary development.

More comments on this to follow...

iPhone SDK, Android and Symbian creating more fragmentation

With over 300 million iphone applications downloaded to date and 200m+ applications downloaded every month across all mobile platforms the mobile application market has never been more interesting. But it's not all good. Despite all the positive impact from the iPhone and Android in the last year this is also causing majors headaches for the mobile services industry. Fragmentation is constantly increasing with more OS (iPhone, Symbian, Android, Windows Mobile, Moblin Linux, Blackberry, etc), mobile browsers (Opera, Safari, Chrome, etc), application standards (Java, iPhone SDK, Symbian, Android SDK, Brew, 5+ different widget standards, etc). This will increase technical complexity, time to market, costs and potentially kill the chances of the industry really taking off. Imagine having to developing different versions of every software program for Dell, HP, Lenova, Toshiba, Asus, etc.

Developing 4 different OS versions of an application is possible although costly for most mobile services companies. Mobile game developers have dealt with this issue for some time with porting and testing costs making up as much as 80% of the total budgets which is bearable but certainly not profitable. The even bigger challenge however is maintaining, upgrading and supporting 5 different OS versions of an application that is in need of constant change. Unless you are Google, Facebook, MySpace or another business with 100m+ users this is simple not an option if you want to have a profitable business.

What are the options then unless you have unlimited resources for mobile application development?

A) Browser based solution only
Stick to a browser based solution and do everything you can to optimise the service over time and leverage new functionality such as script languages on the devices where this is possible. The negative side of this is that the user experience is always a little bit slow and the design and interaction capabilities very limited. It will very seldom give the user a WOW experience.

B) Automatic porting tools to support all platforms
There are a various porting tools available for porting from Java to Brew, iPhone, Windows Mobile, Android etc. These reduce the development efforts but not the optimising and Q&A work. However, they also substantially limit the use of native APIs and functionality across the platforms which means that the ported version is usually based on the most common denominator between the platforms, i.e. a bad compromise.

C) Java and iPhone versions
The only application development standard that works on a majority of handsets is Java Mobile Edition (J2ME). Java is currently available on over 90% of all devices in Europe, 80% in North America (includes packaging for Brew) and about 75% worldwide according to Strategy Analytics. The only multimedia enabled device that does not support Java today is the iPhone. Java definitely has its limitations but in terms of cost efficiency it is the only platform of choice.

In conclusion although the new platforms provide great new capabilities it is very unlikely that the development community will be able to support all of them. The decision on which platforms are used for development must be made on a case by case basis but in most instances Java is the only viable solution for downloadable applications in combination with standard XHTML for browser based services. Despite the competition from new and exciting platforms Java has a good chance to continue to be the platform of choice in the future.

We look forward to further debate on this subject!