We've been hearing for years that mobile applications haven't taken off because there just isn't demand. But there's a chicken-and-egg debate going on between operators, developers, and would-be users that belies that simple explanation.

Which has to come first--the mobile application, or demand? The answer is neither, and it's time developers and network operators rethought the entire question.

It's not that users have no demand for mobile applications. It's that they're afraid to commit to something that's not quite there, or that could end up obsolete. So they throw the blame back on operators and developers. Likewise, developers are also wary of committing to something that could get thrown by the wayside as the complicated Web of standards, consumer predilections and networks evolve.

Aside from a dizzying array of devices, software platforms, and standards, developers also have a gaggle of compliance and certification issues to overcome. Not to mention spotty networks and endlessly delayed rollouts of 3G. With no one committing to roadmaps, taking ROI risks, or committing to a dominant standard, application development is gridlocked.

Still, the predictions are upbeat: Gartner forecasts that revenue from wireless data applications for specific verticals like insurance and retail will grow from US$278.7 million in 2000 to US$894.6 million in 2005. "Horizontal" applications, such as ERP, email, and file transfer, are expected to grow from US$653.6 million in 2000 to US$7.14 billion in 2005.

And a recent Yankee Group survey says 62 percent of large US corporations plan to pilot or deploy a wireless data solution within the next two years.

But those kind of predictions have been bandied about for years. In a comparable Yankee Group survey from 2001, 26 percent of the respondents said they would have a wireless data solution deployed in 12 months. But less than a third of those companies actually got around to it.

Not only have corporations been slow to adopt, some have even dropped wireless projects in mid-stream, like Wells Fargo and Bank of Montreal.

Lack of demand has long been the excuse for abandoning or avoiding wireless applications. But a blurry picture of demand is only a minor part of the problem. What's really at stake is a fear of commitment on both sides. Citing lack of demand, the app suppliers won't commit to more development and users are loath to invest in new devices without assurances that the features and applications they need will be there.

The situation is complicated by the fact that mobile application development is a mess. The absence of a dominant standard means developers are mired in evaluating risks and rewards for a host of different architectures. For microbrowser based content, there's WML, xHTML, and HTML; for downloadable content, J2ME or BREW. And for messaging-based content, there's SMS/EMS and MMS/IM.

The difficulty doesn't end once standards have been chosen. Java's supposed "write once, run anywhere" ability doesn't apply to wireless devices where hardware differences require code to be tailored. There are also certification and compliance testing issues, and unresolved risks related to security and network availability.

How are users supposed to know what they want if developers can't even figure out what their options are?

Ken Dulaney, a Vice President at Gartner Research, is adamant that the biggest thing holding mobile applications back is a conceptual failure on the development, not the demand side.

Dulaney has an example that says it all:

"If you're a traveller, and you need to find out if you're flight is late, and rebook if it is, you could have two options: you can fire up your phone, go onto the airline's WAP site, find out if you're flight is late, and if it is, go to another site and rebook yourself. Or, you could have an SMS message sent to your phone advising you that your flight is late, containing a link that takes you back to a place to change your flight."

The second option is almost universally more appealing. But it's the former option that's available now.

"Browsing should be forbidden on a phone," said Dulaney. But somehow, the development world has failed to grasp that. For that reason, Dulaney predicts that more than 50 percent of mobile applications deployed at the start of 2002 will be obsolete by the end of 2002.

And the jumbled process is going to filter through to the "demand" problem. Delaney also predicts that; "the lack of sufficiently useful and usable applications will be the biggest barrier to consumer acceptance in 2002."

So what's the industry's response? To hold out and wait for that elusive killer app--the thing that drives users en masse to the new technology.

The problem with the mobile universe is, there just might not be one.

Email is the most obvious option for a killer app. According to the Yankee Group study, 80 percent of enterprises said email was the driving application for wireless data. But a murky path to ROI is going to keep most corporations from forging ahead with email, especially in tight economic times.

"The returns (for email) aren't as clear cut as with vertical applications," said Yankee Group analyst Eugene Signorini. "You can't really measure workers' increased productivity, though there are intuitive benefits," he added. In the Yankee Group survey, corporations ranked "specific corporate data/applications," slightly above email as the top thing driving applications for wireless data.

So enterprises go along in their verticals, building their niche applications. Coca Cola, Federal Express, and Charles Schwab are a few examples of companies forging ahead in that direction.

But what's the broader universe of users, and companies looking for horizontal applications, to do? There are a handful of software vendors working towards easing the transition with middleware, such as IBM, AvantGo, Brience, Broadbeam, and Air2Web, but analysts say they aren't the solution.

There's only one player in the game that can call a truce: The network operator.

"The onus is on the operator to make this happen" said Jackson. "There needs to be a concerted effort to streamline the development process, provide device roadmaps and (network) rollout roadmaps," he added.

"They control the environment, and are responsible for educating the people who need to put content on their network. If they don't get things together soon, they're going to blow it again," said Dulaney.

AT&T just announced they would be putting Java on their phones, touting the promise of new revenue streams that would result from the ability to offer games or give phones access to corporate computer networks.

"So what!?" said Gartner's Dulaney to the announcement. "It just shows how backwards these guys are in their thinking."

Copyright © 2009 CBS Interactive, a CBS Company. All Rights Reserved.
ZDNET is a registered service mark of CBS Interactive. ZDNET Logo is a service mark of CBS Interactive.