Published: Dec 18, 2002 in Knowledge@Emory
When IBM CEO Sam Palmisano announced on October 30 that the company planned to sink a hefty $10 billion investment into “on demand” computing, the business community took note. Public pronouncements of such large-scale investments by a technology company are nonexistent these days, as the industry continues to reel from the dotcom bubble bust and a lag in the sector in general. Nonetheless, IBM moved ahead with their initiative, acquiring PwC Consulting, the global management consulting and technology services unit of PricewaterhouseCoopers, for $3.5 billion, in October, to facilitate the process.
On demand computing–defined as software, capacity and maintenance of computer systems provided to the client on an “as needed” basis via the Internet– is a concept, working in very limited degree at many of the top tech companies, including IBM. (For example, IBM makes WebSphere, an Internet infrastructure software for creating, running and integrating e-business applications across a variety of disparate computing platforms. One thing the technology can do is to provide a framework for business-to-business transactions, allowing such things as procurement systems to work across different programming protocols.) The “pay as you go approach” refers to billing only for the computing services used, much like utilities bill their customers.
Certainly, open standards and industry-wide protocols are needed to allow for on demand computing to succeed at IBM and for similar services to work at the likes of their competitors. The mind shift from tech companies offering their own proprietary software and hardware, to products and services based on open standards, is still an evolving process. However, IBM appears to be championing the cause, as it, along with others in the industry, such as Microsoft, Oracle, and SAP, propel this notion through a variety of standards organizations including OASIS and the World Wide Web Consortium (W3C).
Beyond the standards issue, the hard work of combining the various functions, setting up the large-scale infrastructure necessary to maintain this sort of storage, maintenance and software behemoth, and defining how to meter and bill potential customers for the use of services remains a considerable challenge. Says Benn Konsynski, professor of decision and information analysis at Emory University’s Goizueta Business School, “The metering and billing of the services accurately and effectively may be the most difficult and important piece of the puzzle.”
While it may be difficult to note exactly when the idea behind on demand computing came into being, the concept of using the Internet to provide the service came to life probably along with the popularization of the World Wide Web, says Konsynski. Still, as with any emerging idea, business leaders are particularly reticent about using it until the technology is tried and true. Konsynski admits that the business community was similarly skeptical about using the Internet in the mid-1990s, though they were able to bridge that gap. “We will see the same story and skepticism here. But IBM is a trusted broker, and they have a better chance than many to do it,” he says.
While many agree that utility-based computing services will be eventually accepted industry-wide, few can concur on when the technology will be fully available and when business leaders will buy into it. Now, Wall Street is left to wonder if IBM will be able to get the technology fully operable, and get customers to buy into it, before the costs eat into Big Blue. According to Jagdish Sheth, professor of marketing at Goizueta, the on demand computing effort is a large, but necessary gamble for IBM. “This is the only way that they can protect their existing customer base. There are niche players from all over the world attacking IBM’s systems integration business. They have no choice but to do this to head off the competition. If it succeeds, they will have the advantage of keeping customers for life.”
Big Blue faces competition in the field from the likes of Sun, Hewlett-Packard, Gateway, and Microsoft. However, Sheth remains optimistic that IBM has a better chance of succeeding in their on demand computing initiative. “They have done this sort of large-scale effort before,” he says. “In the past, they built network infrastructure. Simply put, IBM is also capable of building and maintaining the infrastructure necessary to be a utility company.”
While Sheth is confident about IBM’s ability to bring the vision to reality, he does worry about how the actual customer setup will play out. “The main consequence of on demand computing is that IBM may have to outsource and partner with others for some of the implementation of it,” he says. How effective they are in bringing partners in to handle this function effectively, and for the right price, will factor heavily into their success, Sheth adds.
But while outsiders describe the many risks involved in IBM’s effort, company executives paint a different picture. Robert Sutor, director of IBM web services strategy, notes that on demand computing “is not a turnaround for IBM, but an evolutionary technology that has been in the works for some time.” Indicating the company’s commitment to on demand computing, he notes that the effort will act as “the underpinning to all of our services.” Part of the on demand initiative will involve eventually selling hardware and software made by other companies, as well as new programming work, in addition to sales of IBM hardware and software. Sutor notes that IBM will spend the next year to two working on the grid (to provide storage and capacity), and developing a variety of integrated “e-business on demand” services that will continually be rolled out.
Sutor believes that “capacity and storage on demand” will appeal to business clients, as customers need not pay for computing services that would only be used temporarily, such as when E-commerce companies need more capacity for the Christmas buying season or a pharmaceutical company needs the capacity and storage for a modeling exercise. He also notes that “autonomic computing” (or computer systems maintenance provided by a hosting company, such as IBM) will be a key selling point of on demand computing, relieving the client of the maintenance burden.
Despite the many selling points, business executives do worry about the security of e-based on demand computing services, as important business functions get transmitted over the Internet, says Anandhi Bharadwaj, a professor of decision and information analysis at Goizueta. However, she adds that this is “more of a perception problem than anything that needs to be addressed by the IT vendors. I think ultimately, data can be more secure with large IT vendors who can provide state-of-the-art security mechanisms, than trying to protect it within the firm. This is especially true for smaller firms that may not want to incur the high cost of ensuring data security. They may find it a lot more economical to go with larger providers who offer storage and security as an integrated package.”
Even the best-laid plans can go awry, however, if the customers refuse to quickly adapt to, and use in large scale, the new technology available. First, Big Blue will need to convince larger businesses to ditch most of their expensive and pre-existing computer networks and systems, to move to on demand computing. Bharadwaj says, “Currently many see it as a risk they are unwilling to take until the concept is proven. Few large companies will be willing to give up the control of the core IT resources and homegrown systems and processes. “Pay as you go” solutions for data storage are more likely to be adopted first, before we see adoption for more complex services such as business processes.”
As competitors offer similar services, Big Blue will also need to differentiate itself as the “utility” of choice for customers. To its credit, the well–known IBM name, deep pockets and history will help the company to attract and keep some customers. One newcomer, Santa Clara, California-based Exodus Communications, for example, failed at capturing a large market for its “pay as you go” computing services last year, and eventually filed for Chapter 11 bankruptcy under the debt load of the build up of their data centers.
Still, the industry stalwart will not only need to come up with the goods in the way of research and services, but it will also need to work harder at marketing a more forward-thinking vision to the business community-at-large. To IBM’s benefit, says Bharadwaj, “Some of the other companies, such as HP and Sun, are more focused on the technical infrastructure than on software processes and services that will run on top of an adaptive infrastructure.” She adds, “Ultimately, the driver for this is “agility” or the need to quickly reconfigure your IT resources and capabilities to meet market demand. If the firm’s internal IT organization is unable to provide this capability in-house, top management will feel compelled to explore on demand computing options.”