Enterprises will deploy significantly more applications to pervasive devices during the next three to four years. By 2004/05, we expect Pocket PC (Windows CE-based) to have 50% share of the enterprise personal digital assistant (PDA) market, with Palm at 35%-40% and Linux-based devices at 10%-15% (primarily specialized devices for Java applications, single-purpose applications, etc.). Within four to five years, we expect improved smart phones (e.g., MSFT Stinger) and improved network rollouts (e.g., 2.5G/3G) to drive enterprise deployment of mission-critical applications to wireless devices (both smart phone-based and wireless modems attached to PDAs), with 15%-20% of enterprise mobile workers using these devices.
We believe enterprises must focus on the following 10 key issues to make pervasive communication a reality:
- 1. Devices will continue to evolve: We expect the pervasive device diversity to continue. Indeed, within three to four years, new devices - such as telematics (i.e., wireless services, often location-based, delivered to vehicles to provide improved comfort, safety, productivity, or mobility), kiosks, information appliances, smart phones, and PDAs - will outnumber traditional PCs. Most enterprise users will have access to multiple devices, and will employ different devices based on work or lifestyle requirements. During the next 12-18 months, companies will need to establish interface and connection standards to enable support for the widest array of current and future device types, rather than standardizing only on specific platforms. Although enterprise deployments can be controlled, companies dealing with consumers will have to support multiple device types (unless they are willing to furnish each consumer with a device, which is an unlikely scenario).
- 2. Connections will (slowly) improve: During the next three to five years, wireless speeds will increase substantially (to 50 kbps-100 kbps, on average). However, the real challenge will be reliability and availability of a connection when needed, not just speed - indeed, for many applications that send small interaction data, 9,600 bps is adequate. Ubiquitous connections are related to network build-out, and will take five to seven years to complete (though parts of Asia and Europe may achieve this by 2005-07).
- 3. Application mobilization is the hard part: Current implementations of pervasive applications require specialized mobile application servers (i.e., Broadbeam, Mobileum, Brience, HiddenMind, Everypath, AlterEgo, Air2Web, Aether, Synchrologic, etc.) as an adjunct to existing corporate applications. By 2004/05, most application vendors (e.g., SAP, Oracle, Sybase, MSFT, IBM, PeopleSoft, Siebel) will offer embedded capability to extend their applications to pervasive devices (much as they currently enable Web services), thus virtually eliminating third-party add-on solutions. Companies that currently require pervasive extensions should expect solutions to be tactical - acceptable if return on investment is less than 18 months - potentially requiring an upgrade in two to three years. Application mobilization will entail considerable effort ---a 6- to 12-month time frame-- and cost ($250K+), with integration effort having the greatest project cost (65%-80%).
- 4. Don't forget security: Although much effort has been made to secure transactions over wireless networks (yet without complete success to date), few enterprises have concentrated on a potentially more risky situation: the protection of potentially sensitive or mission-critical enterprise data stored on pervasive devices with expanding storage capability (64MB-1GB). We believe companies will establish security standards for pervasive devices within the next 12-18 months (2003/04), mirrored on current corporate standards. In addition, we expect device manufacturers to enhance the currently limited security available on these devices with security chips, biometrics, smart cards, data encryption, etc. during 2002-04.
- 5. ROI from wireless: Few enterprises understand the return-on-investment potential of wireless/pervasive deployments. Indeed, many organizations that have deployed such solutions have done so without a firm understanding of ROI. We believe enterprises should require returns on investment that improve end-user efficiency by at least 10%-15%, or offer a payback of 12-18 months, before projects are approved. Furthermore, companies should concentrate on small, well-defined projects that can be bounded and handled by the current relatively inexperienced staff. During 2002-04, companies should concentrate on B2E (business-to-employee) projects, and then migrate to B2B (business-to-business) and ultimately to B2C (business-to-consumer) projects during the next three to four years, as expertise improves and stability and scalability of solutions becomes available.
- 6. Deployment: During 2003/04, we expect many companies to deploy pilot projects (involving a few dozen to a few hundred users) as they learn and gain experience. By 2004/05, with increasing budgets paralleled by decreasing costs (for software, integration, and services), we believe larger-scale deployments will become the norm.
- 7. The market remains immature: During the next 6-12 months, with the downturn in the economy and hesitant corporate spending on pervasive projects, we expect significant consolidation in the mobile application server space. Indeed, several smaller players have either been purchased or are running short of cash. We view this consolidation phase as a good thing, because it will cull the weaker players and leave the strongest survivors. Although we expect the market to be immature for the next one or two years, companies that can show valid payback should not use market immaturity as an excuse for inaction.
- 8. Acting tactically, but thinking strategically: Enterprises deploying pervasive applications should be prepared to replace tactical applications (by 2003-05) with fully integrated strategic applications from existing vendors (e.g., SAP, PeopleSoft, Oracle, Siebel). This will require a strategic plan for pervasive technology that includes a future road map and a migration path for any tactical applications. During the next 6-12 months, companies should establish at least a three-year pervasive plan, including an architecture and deployment strategy.
- 9. Keeping efforts targeted: Companies should focus efforts for pervasive/wireless deployment on well-defined, limited application solutions that perform a few functions well, rather than attempt to port complete and complicated solutions to input-output-challenged devices. Specialized workflow and data input flows need to be checked with users to achieve the maximum ergonomically correct application, or companies will run the risk of deploying applications that users will reject (we expect failure rates of 35%-50% during 2002-05 until enterprises become more familiar with the success factors involved in pervasive deployments). During the next three to four years, we believe most pervasive deployments will be field force-oriented (e.g., service technicians, salespeople, insurance agents, utility meter readers), in industries with many mobile workers, such as finance, insurance, pharmaceuticals, energy, and retail.
- 10. "Remember, it's about business, not technology": Currently, twice as many pervasive projects are driven by the line of business versus being initiated by IT staffs. We expect this to continue, as lines of business focus on improving business practices, while IT staffs focus on improving infrastructure (a necessary step, but not a solution driver). During 2002/03, companies should establish guidelines that set out, in broad terms, the types of applications that should be considered, and how ROI can be measured. Ultimately, this is a business decision and should not driven by "sexy" new technology.
Business Impact: Wireless/pervasive applications will become a business focus between today and 2006. If selected carefully and implemented effectively in targeted business areas, these applications can improve field-force effectiveness and increase return on investment. However, we expect many implementations to fail until organizations learn the success factors associated with this new technology. Investments should be made carefully.
Bottom Line: Through 2005/06, most enterprises will implement pervasive technology solutions. These organizations must focus on effective strategies that include a number of planning and implementation steps. ROI will be king, and implementations will be driven by business need, not by availability of sexy new technologies.