Java To Save Human Race. Evidently.
By Carlo Longino, Thu May 06 22:00:00 GMT 2004

An analyst says revenue from mobile Java apps will grow to $15.5 billion in five years, from $1.4 billion last year. That's a hell of a lot of Tetris.

ARC Group analyst Malik Saadi is bullish, to say the least, on mobile Java apps. While the number of Java handsets sold last year tripled over 2002 to 95.5 million, Java adpotion and use is still skewed to Japan and Korea. Half of Japanese handsets have Java, and 80 percent of people with them are active users, compared to 10 percent in Europe. The report says Java already accounts for more than 10 percent of data revenues, but only if you ignore those generated by person-to-person messaging. That seems like a weak way to look at things, not only because messaging represents the bulk of data revenues for non-Asian carriers, but also because Java information and messaging apps often directly compete with SMS content.

The overwhelming number of existing Java apps -- three-fourths according to the report -- are games, and certainly most of the successful apps are games. This ratio has to change dramatically for Saadi's prediction to come true, since the number of mobile gamers is limited. He says that messaging and informational applications will need to increase, but the market's appetite for these is questionable.

CNN made some waves last fall when they included a Java app with European versions of the Nokia 6600 that downloaded news updates to the device. It hasn't any noise since its launch, and with the 6600 selling so well in Europe, it's a pretty safe bet that the app hasn't found great success. The Nokia content study we reported last week, which said users were willing to spend more on mobile content, also said that people responded they wanted to view information in a browser rather than an application, presumably because it's more familiar.

Like so many other things on the mobile Net, pricing is a mess. It's easy with a Java game, you pay, you download, that's it. But with an information or messaging app, it's difficult for the user to predict their traffic charges. If a user subscribes to the CNN service for $5 a month, they've got that fee, then, in most cases, they've got to pay for traffic on top of that, but how can they gauge their bill, particularly if there's some sort of push mechanism. If a user sends an SMS, they know how much that costs -- but what about mobile IM? What's that going to cost in traffic charges?

An app that gets me news headlines isn't very cool for a monthly subscription fee when I can fire up my browser and get them without paying that charge. Mobile IM doesn't hold a lot of hope if I'm paying AIM $5 a month and can't figure out what chatting with my friends costs when I can send SMS for 5 cents and receive them for free. Some compelling Java apps that aren't games need to emerge to boost revenues, and the billing situation has to radically change along with them.