eaching critical mass for wireless in the enterprise has been a long time coming. Interest in and reliance on wireless technology has crept steadily ahead, but for several years now, wireless has failed to meet past expectations. Certainly, an economic slump has played a major factor, but there are several key technological reasons why enterprise wireless adoption has lagged.
Gartner research director Bill Clark cites the following as the top 10 missed expectations of wireless:
#1. Wide Area Woes: Missed 2.5G and 3G expectations
Although Europe invested heavily in 2G and 3G networks, the reality is that there are very few handsets that utilize 2.5 and 3G systems. Further, Clark says, there are issues with battery life, with the overall performance of the network due to higher data transmission, and with the footprint. "It's a clear case of overhype and oversetting expectations," he says.
North America has been more conservative about 3G, primarily investing in an alternate protocol, CDMA, whose main proponents are Sprint PCS and Verizon. CDMA has a better chance of survival in the short term than 3G and can currently deliver throughput of between 50Kbps and 80Kbps. However, CDMA has problems, too, the main one being that the vendors have not "done a good job of creating a mobile ecosystem of applications" for CDMA. Location-based services, for example, were promised by some providers but the implementations have been poor.
#2. Failure to Prepare Mobile Workforce
Perhaps because they've been so perplexed themselves, IT departments have left users in the dark. They have failed to educate users about what to expect from wireless, its business efficiencies, how to use it, and how it will be supported. Wireless technology opens up all kinds of new problems for enterprises and users alike, including major security, privacy, and spam issues. In most cases, users have heard nothing from the IT department about these problems nor have IT departments prepared recommended procedures for handling them.
As a result, Gartner predicts that 7 percent of mobile deployments through 2006 will failnot because of technological issues, but because of social ones. (See "Wireless Applications: A Radical Shift in Software Development.")
Begin working now to set users' expectations for social factors and training requirements. And talk to Human Resources staff to prepare them for new legal risks and cultural issues.
#3. Failed User Interfaces
Organizations have seemingly failed to learn that a wireless device, at least right now, cannot be a shrunken desktop. User interface is a huge issue with wireless applications (see "Become a Cross-platform Wireless UI Expert .") and corporate IT has been neither sufficiently innovative nor sufficiently aggressive in designing new ways to port their applications to small-screen form factors. Developers, Clark says, keep repeating the same failures, the ones they should have learned from the many failures of WAP interfaces.
#4. Chronic Overestimation of Wireless Consumer Demand
ISVs have been unrealistic about both consumer demand for high-throughput data applications and the possible revenue opportunities from providing them. By focusing initially on building applications that they thought would be used (and happily paid for) by millions of consumers and ultimately being disappointed, the industry largely ignored the opportunities that existed in corporate wireless computing.
Clark says that in 2003, there will be a maximum market penetration of 5 percent for GPRS in Europe, but GPRS penetration will fall far short of that in North Americaprobably less than 1 percent. Right now, about 99 percent of revenue for wireless comes from SMS, which is a big opportunity for mass marketing. "That's the real driver," Clark says, and "we don't foresee that changing significantly in 2003."
#5. Enterprise Architectures that Don't Stretch
The average enterprise either doesn't have business applications that can be extended to mobile devices, or they have severe network or application bottlenecks. Clark says 70 to 80 percent of mobile application efforts are held up by issues in the infrastructure of the wired side of the network.
#6. One Mobile Architecture Doesn't Fit All Workers
Enterprises that have tackled wireless have largely done so in a one-size-fits-all fashion. Such narrow applications may work for specific vertical market tasks, but more generally, applications should be optimized for the profile of the intended user group. For example, are the users of the application true "road warriors" or are they what Clark calls "micro-mobile workers," who are mobile within the confines of a local area or campus? The answers to such questions can help you build a strategy for mobile deployments that is more likely to succeed. (See "Choose Wireless Application Candidates Wisely.")
#7. Mobile Email: The Creation of "Messaging" Silos
Clark says there's a myth that mobile email has taken off, when, in fact, it hasn't yet. Gartner's research found that though there are 60 million mobile workers in the USdefined as those who work away from their desks at least 20 percent of the timebut only around 1.8 million of those workers are currently using mobile email. However, that's going to change soon. Enterprises should be prepared for a sharp increase in demand for mobile email over the next few years. Your demand for mobile email today is probably just the tip of the iceberg.
#8. Mobile Software Vendors and Integrators Failed to Deliver "Always-On" Experience
A seamless "always-on" experience is a big part of what's missing in the wireless user experience. When software vendors can deliver applications and middleware that create an atmosphere of persistent connectivity, productivity and satisfaction will skyrocket.
#9. "Asleep at the Wheel": Failure to Benchmark within Industry
Enterprises have not given sufficient thought as to how they will sell applications into the business or how they will benchmark their success once deployed. Applications without a strong advance level of managerial support often fail, even in non-wireless development. Benchmarking is important, because there's no other bottom-line way to know whether applications justify the investment.
#10. Lack of Metrics: Not Linking Wireless Initiatives to Business Value of IT
Finally, organizations need to be able to articulate the value of mobile applications. Clark says, "In many cases what we see is that a particular project will have success as a point solution, but will fail after that in creating a beachhead." What you end up with, is an application that has value but no roadmap for extending that value and no buy-in or support from the business decision makers. To mitigate this problem, document requirements, do pilots of all your projects to understand recurring costs, and conduct assessments for feedback.
Guidelines for Success
This list of guidelines is comprehensive, but is still a vast oversimplification. The myriad factors that play into the underadoption of wireless vary by industry, by technology, and by organization. But in a general sense, probably one of the key problems is the complexity of the relationships between service providers. The average company will need to work with four or five different vendors in each of several solutions categories including mobile device managers, voice providers, wireless wide area carriers, security companies, etc. Too many companies rely too heavily on one vendor in each category, which Clark characterizes as a quick route to failure. And, he says, systems integrators have biases that may not end up working in your favor.
One way to steer your organization toward successful projects, obviously, is to know what types of projects are patently inappropriate for wireless deployment and avoid them like the plague. What are some applications that shouldn't be considered for mobile right now?:
- Safety critical systems.
- Systems where you can't meet user expectations.
- Systems that don't have a fallback. Any key business process that you deploy on wireless technology needs to have offline multimodal technology that allows the system to work when the wireless link is down.
- Systems that have mandatory performance or quality of service requirements are high risk in the near term.
- Systems with regulatory or service requirements. For example, if your business is beholden to certain industry privacy restrictions you have to consider that the devices are the same as a PC or server and you have protect the data in the same way.
That's a lot of negatives. Clark also offers some "killer characteristics" of wireless applications. Use this list as a guide to select the applications that will have a better chance of succeeding as a wireless deployment.
- Actionablenot contextually dependent on another device, e.g. does not require synching with a PC
- Localizedimportant for providing context across applications
- Easy to usekeeping keystrokes down to one or two
- Appropriately richdon't use bandwidth wantonly