E-sourcing: Information Technology on demand: Utility computing marks a new stage in the evolution of outsourcing, and may be the first to fully deliver on the strategy's true potential. (CEO Perspectives).

The business objective behind the concept of outsourcing has always been a fundamentally sound one: handing off responsibility for non-core functions to a third-party provider, thus freeing up capital and human resources that can be redirected to more strategic uses. In theory, the strategy is supposed to result in improved quality and efficiency in both the outsourced support functions and the in-house strategic ones, along with a net cost savings. In practice, results may not meet expectations. Customer demands increase, as well as the need for greater speed and cost efficiency to remain competitive. In the area of information technology, however, that may be about to change, thanks to a new concept dubbed utility computing, or "e-sourcing."

Utility computing is a pay-as-you-go model frequently likened to the electric industry. Just as end-users tap into their local electric grid and pay only for the power they use, utility computing customers will be able to tap into a mammoth computer grid and be charged on the same basis for their IT services. Transforming fixed costs into variable costs has long been one of the key objectives of IT outsourcing, and utility computing promises to provide a more granular and flexible way to achieve that end, notes Bill Martorelli, vice president, enterprise services strategies, at the Hurwitz Group, an information technology industry research and consulting firm.

"The emerging excitement over utility computing stands in stark contrast to the dismal current picture for co-location and managed services providers," Martorelli says. "By offering a variable-cost alternative, the advent of utility computing opens up exciting new possibilities for enterprise customers increasingly concerned with the high costs of current computing models."

Simply stated, e-sourcing entails the delivery of standardized processes, applications and infrastructure as a service over the network, with both business and IT functionality. It differs from traditional outsourcing and current hosting services in three key ways:

* E-sourcing is shared, simultaneously serving multiple customers in a flexible, automated fashion.

* It is standardized requiring little customization or integration.

* It is scalable, providing capacity on demand in a pay-as-you-go model.

The concept also holds out the promise of additional benefits to end-users. It can help companies simplify the adoption of new technologies, minimize IT hiring and training obligations, and compress time-to-market for new, value-adding projects and initiatives. As IBM Chairman and Chief Executive Officer Lou Gerstner Jr. pointed out in a security analyst briefing in May 2001, that last point may be the most significant to chief executives.

"Speed drives every CEO and every business manager today," Gerstner told the analysts in supporting his contention that IT will increasingly be a services-led industry in the years ahead. "It is the top item on every manager's agenda -- specifically, speed and response time to competitors. When you couple this with the [IT] skills shortage, you see the explosive demand for services."

Utility computing's promise is prompting many enterprises to view the concept as a continuum that extends beyond commonplace IT resources on demand. Indeed, utility computing aims to strengthen the delivery of business process and management functions integral to the way the organization works. It is in that expanded view that its ultimate potential will be realized, most analysts believe.

THE FUTURE OF CORPORATE COMPUTING

Virtually all the IT industry's heavyweights, along with a host of smaller companies, are positioning themselves to jump on the utility computing bandwagon -- or have already done so. Some are focusing on providing the enabling hardware and software infrastructure, some on content and processes. One of the most comprehensive and wide-reaching efforts to date is IBM's "e-business on demand" project, and senior executives at the company have clearly signaled their belief that this is the future of corporate computing.

Gerstner set the tone in December 2000, when he declared that the future of business computing would be the sale and delivery of information technology as a utility-like service over the Internet. "We are rapidly approaching a day when customers aren't going to have to own, manage or even house any of the traditional assets of our industry," he said. "The processing, the storage, the applications, the systems management, security, load-balancing -- all of it -- can be provided over the Net as a service." Gerstner also promised that IBM would focus on this new utility model, delivering software, server time and data storage to customers on a pay-as-you-go basis, like power from a socket.

To be sure, some observers have been skeptical. They point to the hype surrounding e-commerce in the past and previous outsourcing trends that did not live up to initial expectations. "They want to know if utility computing is for real or merely the latest hope for vendors anxious for a way out of the current industry slowdown," Martorelli says.

A growing chorus of independent voices and a developing base of satisfied customers suggest that e-sourcing really is something new, different and potentially valuable. One such voice is 12 Entrepreneuring, a San Francisco-based operating company that builds technology-enabled businesses, including Grand Central and Oxygen Software. John Hagel Ill, the company's chief strategy officer, and John Seely Brown, its chief innovation officer, voiced their confidence in this new IT strategy in an article in the Harvard Business Review [October 2001]:

"No doubt, many executives are skeptical [of the concept]," they write. "They've heard outsized promises and indecipherable buzzwords before, and they've wasted a lot of time and money on Internet initiatives that went nowhere. This time, though, there's an important difference. The technology providers are not making empty promises: They're backing up their words with massive investments to help create the infrastructure needed to make the new IT approach work. As these efforts continue, over the next year or two, a steady stream of new, Internet-based services will come on-line, providing significant cost savings over traditional, internal systems and offering new opportunities for collaboration among companies. Slowly but surely, all your old assumptions about IT management will be overturned."

David Tapper, program manager of networked infrastructure management services at IDC, the technology industry analysis firm, offers a similar assessment. "Essentially, there appears to be a move under way to deliver IT on a 'buy-by-the-drink' service model that IDC refers to as computing utility. In the world of computing utility, customers can purchase storage, server and software from a central location on a pay-as-you-go basis, just as they can buy telephone services or electric power," Tapper explains. This trend appears to involve "a dramatic evolution, perhaps a revolution," in how IT services will be bought, delivered and paid for in the future, he adds.

GRAVITATING TOWARD THE GRID

At the heart of the e-sourcing revolution is Grid computing, a technology that enables scalable virtual organizations. Grid computing got started in the scientific and academic sector and is just now migrating to commercial applications. It is distinguished from conventional distributed computing by its focus on large-scale resource sharing, innovative applications and high-performance orientation. Grids are clusters of servers joined together over the Internet, using protocols provided by the Globus open source community and other open technologies, such as Linux. (The Globus Project is a research and development initiative originally focused on enabling the application of grid concepts in scientific and engineering computing.] Computing grids allow geographically distributed organizations to share applications, data and computing resources.

Grid protocols are designed to allow companies to work more closely and more efficiently with colleagues, partners and suppliers, Tapper says. They do that through:

* resource aggregation, which lets corporate users treat a company's entire IT infrastructure as one computer through more efficient management:

* database sharing, through access to remote databases -- particularly useful to life sciences, engineering and financial firms: and

* collaboration, allowing widely dispersed organizations to work together on a project with the ability to share everything from engineering blueprints to software applications.

Making Grid computing a reality on a commercial basis will require a massive, if disparate, infrastructure, and, as Hagel and Brown note in their article, the equally massive investment needed to create that infrastructure has already begun. IBM Global Services, for example, has announced plans to invest in a worldwide network of computer farms to support Grid computing. The company has already garnered important experience in this area through its work on scientific and academic Grid computing projects, including a $50 million initiative in the United Kingdom.

IBM is providing a storage controller and open-source Globus middleware to Oxford University, one of nine universities participating in the U.K.'s National Grid project, expected to be completed by fall 2003. IBM is also supplying 1 6 Intel-based xSeries 330 Linux servers to link Oxford to the other institutions in the network, and it has sold 200 xSeries servers to The Netherlands for use in a five-university grid being built in that country.

"We believe that Grid computing and the emerging Grid protocols will grow beyond their current home in the academic world and become a foundation for the delivery of computing to business customers as a utility-like service over the Internet," says Dev Mukherjee, IBM's vice president of e-sourcing strategy. "It is not surprising that Grid computing first emerged in the scientific and technical community, the same community that invented and developed the Internet, the World Wide Web, Mosaic and the concept of the Web server. As these technologies propelled the sharing and usage of information over the network, IBM believes Grids will propel the sharing and usage of computing resources over the network."

While the development of commercial computing Grids is key to the emergence of utility computing, this new concept is also dependent on the convergence of other technologies, most notably Web services and enhanced, "self-healing" distributed systems management, Martorelli says.

THE NEED FOR NEW MODELS

Just about all software companies are adopting Web services as the basis for their middleware offerings to facilitate e-business integration requirements, notes Jeff Gore, vice president of e-sourcing technology at IBM. "Grid protocols have been designed primarily to support the needs of the scientific community in the past. They need to emulate Web services and be extended to support the requirements of commercial e-business applications," he says. Chief among them is the ability to dynamically integrate applications and business processes, and the ability to execute high volumes of transactions against shared databases. We are working closely with the Grid community to help evolve Grid protocols to support commercial as well as scientific applications, encompassing Web services when appropriate."

Hagel and Brown point out that construction of the Web services architecture is still in its early stages and will require years of investment and refinement. Yet companies shouldn't wait to begin the transition to a utility computing IT strategy. "Even today, benefits can be gained by moving to a Web services model for certain activities and processes." they say, suggesting that companies take a pragmatic, measured approach to deploying that strategy, to which the Web services architecture is ideally suited.

For the Internet to be an open and robust computing platform capable of supporting the e-sourcing model, its management capabilities must be significantly enhanced. The increasing complexities and sophistication of the Internet call for an autonomic computing infrastructure. "The network needs the intelligence to manage itself, in the sense of detecting all its capabilities and self-configuring itself, self-optimizing its allocation of resources, self-healing upon failures of individual components and self-protecting against security intrusions of all sorts," Gore explains.

Users of the utility computing infrastructure must be able to specify their requirements as policies to be followed for performance, scalability, availability, security and other parameters. The various systems in the infrastructure must then execute those policies and adjust their computing capabilities as needed. "That is the basis for IBM's Project eLiza, with its goal of delivering systems that manage themselves," Gore says. Project eLiza focuses a number of programs all across IBM on the common goal of developing automated management functions, essentially asking the systems to diagnose and heal themselves. "The objective is to give customers the power to manage environments that are hundreds of times more complex and broadly distributed than those that exist today." (See "Matching Computer Use with Demand," page 7.)

To be sure, there are obstacles that go beyond technology that must be overcome before adoption of the utility computing model becomes widespread. "Utility computing models are tantalizing in their ability to provide for variable-cost computing architectures, but the technology itself is probably the least of the problem," Martorelli contends. "Both existing buyer behavior and organizational structures typically built around technology silos will have to evolve to make it a reality. New business models for products and services enabled by Web services will also have to emerge.

Hagel and Brown concur, noting that the shift to a Web services architecture for corporate computing will require broad organizational and managerial changes, as well as the development of new technologies. The impact will be particularly pronounced in corporate IT departments, which will have to outsource many traditional activities while simultaneously leveraging internal capabilities to design distinctive Web services that can be sold to other companies. Chief information officers will have to become strategists, entrepreneurs, knowledge brokers, relationship managers and negotiators.

"AS RELIABLE AS FLIPPING A SWITCH"

Still, there is strong evidence that adoption of the utility computing model may turn out to be one of the fastest and most far-reaching developments in the history of the IT industry. Trends of this sort are generally led by smaller, cutting-edge enterprises, and there has certainly been some of that activity in the e-sourcing arena. However, IBM has already inked a deal that brings utility computing to the top echelon of corporate America. In what the company describes as "a first-of-its-kind collaboration between Fortune 100 companies," IBM will provide hosting services to manage Dow Chemical Company's rapidly growing Internet-based business, which has experienced a nearly 4,000 percent increase in customers since 1999. IBM researchers will work directly with Dow tech engineers to build utility-like computing pipelines on the Web that will allow Dow to access IT -- software, server time and data storage -- from IBM on a pay-as-you-go basis.

"Information delivery is becoming a utility, like electricity, gas and water," says Dow CIO David E. Kepler, who also serves as the company's corporate vice president of electronic business. "Quality customer care in today's dynamic e-business environment requires the predictability of a utility-based computing model. Our agreement with IBM enables us to maximize the availability, performance and quality of our online information and services for customers. Information delivery should be as simple and reliable as flipping a switch and turning on a light."

As things now stand, the most relevant question about utility computing's future appears to be simply when, not if, Martorelli believes it will play a significant role in the futures of IBM and other leading IT industry companies, although it will take some time.

"What they are offering is a new form of outsourcing that is more effectively aligned with business impact than previous attempts, such as IT outsourcing and business process outsourcing, have been," he says. "Conceptually, this is targeted at the chief executive level, with the prospect of variable costs and real payback in CEO-relevant terms. That's the kind of message that really resonates in those precincts."

[Graph omitted]

[Graph omitted]

[Graph omitted]

RELATED ARTICLE: RX FOR E-SOURCING

About two years ago, ChartOne, an 18-year-old company that is an industry leader in the administration and management of medical records for health care providers nationwide, sensed an impending business opportunity in the recently enacted Health Insurance Portability and Accountability Act. Provisions of the Act created new security rules to ensure the safety and privacy of individually identifiable health care information and records.

"We saw an opportunity to create digital warehouses and remove the paper from the back end of the process," says Peter Henderson, ChartOne's senior vice president of marketing and strategic planning. Paper-based medical records would be electronically scanned into a data warehouse, where the company's customers could access them via the Internet. However, the San Jose, Calif.-based company wanted to avoid the massive up-front investment and protracted time-to-market that an in-house solution would entail. Henderson and Sharad Patel, ChartOne's executive vice president and chief technical officer, agreed that an outsourced solution was the answer.

"We wanted to create a best-of-breed solution from an application standpoint that would be complemented with a best-of-breed infrastructure on the back end," Henderson says. ChartOne wanted to be able to focus its resources on its core competency -- the creation of medical information-related applications -- while relying on a third party for the storage solution. It approached a number of vendors with a set of parameters for the new service:

* an affordable plug-and-play solution;

* leading-edge backup, security and recovery services;

* state-of-the-art facilities; and

* the ability to deliver additional storage capacity within a two-week time frame.

The company also wanted management flexibility that would allow it to incorporate its legacy servers into the solution and have a third party oversee facilities hosting operations. "We evaluated several vendors, and IBM, which at that time was just creating the idea of e-sourcing, had the most flexible model and demonstrated the best understanding of what we were trying to do for our customers on the application side," Henderson says.

ChartOne's new View Manager service went live in April 2001, providing the company and its clients with online access to records management on a cost-efficient, pay-as-you-go basis. Implementation was completed in a matter of weeks, shaving six months off the initial launch forecast.

"Essentially, we are able to offer an infrastructure solution to our clients at about a third of what it would cost if they built it themselves," Henderson says. He estimates that e-sourcing has lowered ChartOne's total cost of ownership by 60 percent, saving $4 million over the new service's first year of operation.

The experience has made Henderson a believer in utility computing. "I think being able to purchase things like software, server time and data storage on a pay-as-you-go basis is where the market is going to head in the future," he says, "not just in health care, but across the board."

IN CONTROL -- AND WITHOUT THE HEADACHE

As director of information technology for Memphis City Schools, Linda Mainord oversees more than 2,000 devices connected in a wide area network that links 187 sites in Tennessee's largest school district. Deploying the network was a multiyear project costing millions of dollars, and the WAN is now the focal point for both educational and administrative functions in the district. Making sure it is available to deliver data, video and voice wherever needed 24 hours a day, 365 days a year is one of Mainord's biggest responsibilities. It's one she meets by relying on outside help.

"Managing and monitoring the network and controlling costs are huge issues for my department," she says. "We have a conservative budget and a limited IT staff. We need to manage risk and ensure network reliability in the most cost-effective manner possible. Using a third-party provider for network management and monitoring allows us to focus on what we do best and not add to the load of an already overburdened staff."

Although the school district originally planned to handle network management in-house, Mainord realized early on that there were problems with this approach. "We went in thinking we could do it ourselves, but the complexity of what we were proposing quickly became apparent. Implementation time was one big issue," she says. "We would have had to double up on what we were already doing, and network reliability would have been vulnerable as a result."

The answer was to "out-task," the district's term for selective outsourcing. Memphis turned to an alliance between IBM and NetSolve, a leading provider of remote network management services.

Cost, reliability, predictability and implementation time were all issues Mainord considered in making this choice. Deploying an in-house network management solution would have to have been done in a piecemeal fashion, and she estimates it would have taken two to three years. The network's efficiency and reliability would have suffered in the process, and its ability to meet the needs of the 119,000 students and 16,000 employees who rely on it would have been compromised.

"IBM has been our vendor of choice for a number of years, so I went to them and told them what I wanted, that I wanted it quickly and that I couldn't get the bodies to do it myself," Mainord recalls. "They immediately began looking at how we could make it happen, and what we have today is a direct result of that effort. IBM and NetSolve monitor the network 24/7 and are responsible for all the headaches that go with it -- change management, problem management, software distribution and all of the other elements involved in systems and network management."

Mainord likes the fact that she is getting just what she needs for a flat monthly fee that makes budgeting easy while retaining full control of the network. She estimates that using IBM's e-business-on-demand solution costs the school district about half what an in-house solution would cost. "I still own the network; it belongs to me," she says. "I make all the decisions, and I'm getting the support and services I need to make sure everything is always up to speed, but I don't have to buy extra hardware and software or worry about training and additional skill levels."

Mainord is sold on the concept of utility computing. "It works extremely well in the educational environment," she says. "It allows us to augment our core competencies and leverage some skills, such as architecture engineering, that we would never be able to attract on our own."

MATCHING COMPUTER USE WITH DEMAND

What effect will a-sourcing have on your company? IBM's Dev Mukherjee, vice president of e-sourcing strategy, and Jeff Gore, vice president of e-sourcing technology, explain how utility computing works and how your company can benefit.

What types of businesses and organizations are likely to gain the most from e-sourcing?

DEV MUKHERJEE: Literally all businesses and most organizations can benefit from e-sourcing, almost immediately. As we make more infrastructure and business processes available over the network, businesses and other organizations will have access to more resilient and flexible infrastructure and new applications or business processes that previously required significant investment.

What are the requirements in terms of infrastructure, expertise and so on that organizations need to tap into this new form of IT?

JEFF GORE: That's an interesting question. E-sourcing is about removing the need for firms to be technology experts, or to manage investment or technology risks. I would say that the key requirement is to know how you want to create competitive advantage in your industry using technology.

What critical technological developments led to the development of e-sourcing?

GORE: First, advances in telecommunications have allowed more and more data to travel at ever-faster speeds, along with the build-out that has led to bandwidth prices falling dramatically over the past 12 months. Second, distributed systems and middleware technology today make it easier to design and manage systems. Finally, there have been tremendous strides made in server and storage virtualization, including Grid technology, and that has gained a lot of interest over the last year.

What else needs to be done from a technology standpoint before e-sourcing can reach its full potential?

MUKHERJEE: The real issue is adoption: How quickly will businesses realize the benefit of "e-business on demand?" But there are technologies that will make adoption and migration easier. We're always looking at technology that makes our operations more efficient. Likewise, our software partners need to continue to make it easier for service providers to host and manage their applications for multiple customers.

Are there technologies or programs in the pipeline that will improve or otherwise affect e-sourcing's future?

GORE: Project eLiza, our whole autonomic computing effort spearheaded by IBM Research and server colleagues, will make a huge difference to data center operations for us, our partners and customers. (Project eLiza coordinates all of IBM's efforts to develop automated management functions.) Self-managing and self-healing technology reduces risk and makes the infrastructure more efficient. Grid computing and the work our software colleagues have done in middleware are also helping.

What are the major benefits e-sourcing offers to companies?

MUKHERJEE: The greatest benefits are around time to value -- that is, getting systems that create new value or provide access to new markets up and running quickly, all as efficiently as possible and with the minimum distraction to management and employees. To paraphrase Alice in Wonderland, in this environment, you have to run twice as fast just to keep up. E-sourcing helps you get ahead -- right now.

What is IBM's pricing model for e-business on demand?

MUKHERJEE: The simple answer is that we're moving toward a pay-as-you-go or utility model. However, the better way to think of this is as a tighter coupling of the investment in technology and the business value -- for example, e-procurement or document exchange systems billed by the number of transactions they handle rather than the number of servers, people and data centers required to make them work. Clearly, this is a transition. Computer users want the benefit of being able to scale up and down as required, but they also want some predictability in their costs.

What is IBM's vision for the future of e-sourcing?

MUKHERJEE: We expect to have more and more infrastructure and process offerings going forward, almost like a cable TV model where you can click on what you want to use. In a relatively short period, buying infrastructure and process capability and know-how over the network will become the norm, just as plugging into a wall socket to get electricity is today.

WHY E-SOURCING IS A CEO ISSUE

Why e-sourcing? When your business needs electricity, you don't spend millions to build your own power plant. Yet when your business needs more e-business capabilities or infrastructure, you spend millions of dollars and expend thousands of man hours to build it yourself. Why?

Because e-business has revolutionized the speed, reach, efficiency, leverage and responsiveness of business. You need to be e-business ready. And unfortunately, there have been no short cuts to e-business readiness.

Until now.

The solution is e-sourcing, or e-business on demand.

Imagine being able to flick a switch and get more storage instantly. Imagine being able to scale your servers up and down in seconds rather than days. Imagine hardware, turnkey business services, industry-specific solutions and major e-business applications all as accessible, reliable, dependable, affordable and ubiquitous as water, gas, electricity or the telephone.

That's the value of utility computing. E-business on demand is just there. Turn it up when you need more. Turn it down when you need less. Your company doesn't pay massive up-front costs or take on major infrastructure decisions -- just pay as you go, leaving the infrastructure, integration, implementation and management to the e-business and infrastructure specialists at IBM. While you spend your time, energy and resources focusing on your business, we help you get e-business gains without e-business pains.

This is a technology that has the potential to transform companies, literally, in important ways. It's already working for several major companies, cutting implementation time and giving them the opportunity to experiment, to better manage cash flow and resources, and to access increasingly hard-to-find technology skills and know-how.

Why e-sourcing? It's the wrong question. The question should be "why not?"

Jim Corgel, General Manager

e-business Hosting Services &e-sourcing