The symbolism could not have been more obvious to those who had long endured predictions of IBM's extinction: The new chief executive's 2002 coming-out party—the public launch of the company's ambitious on-demand initiative—was at New York's American Museum of Natural History, an institution renowned for its collection of fossils.
Samuel Palmisano, on the job as CEO for less than a year at the time, was intent on proving that IBM was anything but a relic.
"We are on the cusp of a dramatic shift in this industry," he told customers and employees at the event. "Think about this change. See the world from a different dimension. Don't manage it strictly by functional silos, because a lot of this is really about culture—cultural transformation."
More IT news stories
Microsoft file patent faces exam
Flash brightens AMD outlook
HP seals open-source e-mail deal
Microsoft: Linux threat is rising
Since then, IBM has effectively erased its epitaph as corporate dinosaur. In the past two years, the company has gained market share in categories such as server and id="5210844">middleware—two strategic and high-ticket corporate businesses that have contributed to its $89 billion in annual revenues. No longer a lumbering giant living off a faltering mainframe franchise, Big Blue has shifted its focus to services and made its $14 billion , once dismissed as an ineffectual also-ran, a central part of its long-term strategy.
In many ways, however, it can be argued that IBM has succeeded in spite of itself. As a modern technology patriarch that was once synonymous with computing, the company had every reason to dominate the industry through the end of the 20th century. Instead, because of strategic blunders and bureaucratic inertia through much of the last two decades, IBM fell behind in key technologies—from databases to personal computers—that it had pioneered.
"I see a focus and a drive there that is unprecedented, in my mind, in IBM history, and a vision that is pretty broad and far-reaching in its scope and impact," said Tony Scott, the chief technology officer of General Motors, one of IBM's largest customers. But when asked about the on-demand concept, he replied: "They may think they are further along on the path than I do."
What's 'on demand,' and who's demanding it?
IBM's own executives acknowledge that the on-demand initiative has outside the company. But they say any risks in the marketplace are far outweighed by another value: its power as an internal rallying force to direct an unwieldy empire spread over 170 countries, encompassing more than 350,000 employees, 670,000 stockholders and dozens of manufacturing plants in such far-flung locations as Prachinburi, Thailand, and Szekesfehervar, Hungary.
The on-demand concept has become something of a personal crusade for Palmisano. When IBM launched the campaign, the company had fully recovered from near-death experiences of recent decades but still needed a way to break into new markets. An IBM lifer who joined the company right out of college, Palmisano had to show that he could articulate a strategy for long-term growth.
His charge was made more difficult by inevitable comparisons to predecessor Louis Gerstner, who in 1993 was parachuted in from RJR Nabisco to rescue IBM—a rare outsider appointment in the company's senior ranks, which were usually as incestuous as the rest of the high-tech industry's. Gerstner was elevated to celebrity status after being credited with saving the company. By contrast, Palmisano worked his way up and was known as more of a , a former college football player who had steadily plowed his way through the ranks since joining the company 1973.
As it turned out, blocking and tackling may have been exactly what IBM needed when Palmisano assumed the reins. He is a no-nonsense businessman who enjoys talking to clients and getting things done.
"Every six months, he's turning the crank," said Jonathan Eunice, an analyst at research firm Illuminata. "I hear all the time that 'Sam says we're going to do something,' and people are reacting."
To fully understand what on-demand means within IBM's Armonk, N.Y., headquarters, the initiative must be viewed in the context of the company's extraordinary history. With origins that date back more than a century, IBM had for years been known for a buttoned-down culture that resisted change. It had a stultifying bureaucracy that made it a corporate version of the Pentagon, alongside other industrial institutions like General Motors and AT&T.
As a testament to its military-style hierarchy in the early 1990s, one former employee said he was once assigned to elevator duty during a visit to a North Carolina facility by then-CEO John Akers, making sure the boss would not have to wait for one in the lobby. That kind of mentality still existed a few years ago, according to another company source who said work was once halted for two days at a lab in IBM's hard-drive division so that it could be spruced up for a 15-minute visit from a senior vice president.
By the early 1990s, the company had lost touch with its customers and prevailing industry trends, having based its strategy on locking clients into proprietary mainframe systems. Gerstner, a former IBM customer himself, ridiculed the computing industry's modus operandi of forcing customers to upgrade to solve business problems.
Now, Palmisano is pushing to complete IBM's transformation from primarily a hardware company to one led by software and services. The idea is that customers want bundled "solutions" of software, hardware and services tailored to specific industries, rather than a catalog of products.
"Sam Palmisano talks to the business side of the customer and other IBM people talk to the technical end of the customers," said Amy Wohl, an analyst and president of . "Technology vendors have learned their lessons that if you're going to talk to the CEO, you'd better talk about the business value."
IBM's services-led approach has its risks. The company's high-price consulting talent needs to bring in new business and compete with cheaper alternatives from an consulting and services industry. And some customers may be leery of employing a cadre of consultants to get their technology projects done.
"They have teams of consultants that don't always know the products," said Tim Kelly, technology director of , a credit card transaction processor in Columbus, Ga., that reported $1 billion in revenue last year. "You know how it is when you're trying to hit certain needs, and someone pulls out a book and starts reading it right in front of you to figure out how to do something. The $300 an hour all of a sudden tastes pretty sour."
Moreover, the definition of on-demand remains inexact to almost everyone outside IBM. Some customers associate it with other industry trends and jargon, such as "utility computing" or "hosted outsourcing services." Those most skeptical of the concept reduce it to little more than a repackaging of existing goods and services.
"On-demand reminds me of the emperor with no clothes—there's nothing there that's new or different," said Michael Cusumano, a professor at the Massachusetts Institute of Technology's Sloan School of Management. "It's mostly hand waving on the part of a few technology companies that are trying to get people to buy hardware and services."
IBM managers concede that on-demand is an attempt to redefine the market for computing. The opportunities to make money on smaller-scale jobs, such as PC networks or back-office applications, have largely dried up. Tech companies now sell goods to automate processes that touch many points across a corporation.
Haunted by failures
On-demand is another IBM attempt to chart the IT industry's direction. Many initiatives, such as the PS/2 personal computer or the AD/Cycle development tool, were all hailed at one time or another by IBM management as revolutionary—and all ultimately flopped.
Few, if any, rivals have been as influential or prolific as IBM in the computer business. Its engineers developed the first relational database, hard disk drive, reduced instruction set computing (RISC) chip, speech recognition software and dynamic RAM memory. The company has been granted in excess of 22,000 patents in the last decade, more than its top 10 competitors combined.
Turning those inventions into revenue has been another matter. While IBM engineer Edgar Codd defined the relational database in the 1970s, the technology didn't mesh with corporate strategy. The result: IBM was late to market, and Oracle became the dominant leader in databases in the 1980s. More recently, despite its prowess in hard-disk design and manufacturing, IBM failed to recognize that storage would become a major market and was forced to play catch-up with EMC and other companies.
The consumer market is the scene of IBM's most infamous failure: the personal computer. The company's legendary mistakes around the PC serve as business school case studies in what not to do.
At the dawn of the PC era, IBM management's hardware-centric mentality underestimated the power of software and allowed Microsoft to establish its fortune with its Disc Operating System. IBM later attempted to build a successor to DOS, called OS/2, with Microsoft's cooperation. That ended in a high-profile divorce. IBM got a dead-end product, and Microsoft walked away with a multibillion-dollar franchise in Windows NT.
IBM fared no better in the years that followed, as one disaster after another highlighted the company's tin ear for software marketing. Well after Microsoft had established its Word and Excel applications as de facto standards, IBM repeatedly tried and failed to launch its own desktop software business.
It wasn't until the mid-1990s that IBM finally began to score some wins, seizing on the Java and XML (Extensible Markup Language) programming languages, then Linux technologies—which together form the foundation of its current software products. IBM also bet correctly that the Internet would have an even more dramatic commercial effect on businesses than it did on consumers.
About that time, the company made some high-profile acquisitions, notably a hostile takeover of Lotus Development in 1995 and the purchase of Tivoli Systems in 1996. Bringing in independent software companies shook up IBM's hardware-centric and fractious culture.
"They were clear to me that they brought Tivoli in to change IBM, not other way around," Tivoli founder said in an interview last year. "One change that Tivoli had a significant impact on was the ability to make decisions like a software company. And I don't doubt that we pissed some people off along the way."
IBM now has a structured process for evaluating emerging technologies, such as and computing. Some investments have yet to pay off, including a billion-dollar bet on , and IBM admits that its multibillion-dollar is not performing up to speed.
As it pursues its own on-demand ideal, IBM values the ability to react quickly to customer demand, which stands in contrast to the days when IBM's decision making was centrally planned. If it works, analysts say, the on-demand initiative could have an even more profound impact on the company than the e-business mantra born under Gerstner during the Internet boom.
"E-business was all about using IBM's services capabilities to pull all the technology together, but there weren't any fundamental changes in the technology," said Tom Bittman, an analyst at Gartner. "Now you can go to any division at IBM and hear an on-demand story that doesn't conflict with other parts of IBM. That's pretty powerful."
Old enemies take notice
Already, rivals that once pitied IBM have responded to its new aggressiveness. A broad agreement between , announced in April, included plans to collaborate on technical interoperability. Closer interoperability between the two companies' gear is a hedge against IBM, which earns a lot of money in services by making disparate systems work together, said John Rymer, an analyst at Forrester Research.
"There's no question who our biggest competitor is," said Microsoft senior vice president Bob Muglia. "It is IBM."
The new IBM has become so confident that it is even returning to the scene of its painful defeat on the desktop to challenge Microsoft on its own turf. In May, IBM , which has relatively simple applications that can be delivered via a Web browser from a network server.
Workplace demonstrates IBM's technology strategy in a nutshell. The entire package draws on several components already in IBM's arsenal, including its Lotus collaboration software and open-source technology. IBM's strategy is to mix and match the components—whether they're processors, software programs or business process templates used in consulting—to tailor products for specific industries.In targeting specific industries, IBM has pledged to spend and independent software providers with industry expertise. The partner outreach program is a critical component of IBM's plan to garner more business from midsize companies, which are expected to fuel technology spending in the years ahead.
But this is uncertain terrain for IBM, and Microsoft is well-prepared to defend its ground, with thousands of independent software providers in its camp and a strong customer base among smaller organizations. To push into midsize companies, IBM must overcome its own legacy of catering to large corporations, which favor performance over cost and ease of use. A Microsoft executive said IBM is more interested in dominating any partner network than in being an equitable partner.
"All IBM is trying to do is sell IBM mainframes and service hours," said Adam Sohn, product manager of the Platform Strategy and Partner Group at Microsoft.
Competitors also criticize IBM's offerings as disjointed. Indeed, have been instrumental to IBM's plan to provide one-stop shopping to corporate technology buyers. But accommodating new technologies, as well as decades of existing IBM systems, makes tightly integrating IBM's many parts a tough challenge.
"To really to catch up—to get in a leading-edge position in the market—IBM has made some purchases. And some of the purchases are really not well-integrated into their product suite," said Jackie Barretta, vice president of information services at transportation services company .
The big in 'Big Blue'
Despite integration issues, the breadth of IBM's products will provide at least some advantage over the competition, and many customers will be drawn to its expertise in business consulting. IBM strengthened its services in 2002 by buying the management consulting unit of PricewaterhouseCoopers. That acquisition of 30,000 consultants serves as the linchpin of IBM's efforts in the fast-growing category of "," or BPO.
In this form of outsourcing, service providers take over entire business processes, such as human resources, finance or supply chain logistics. IBM is aiming for a of the BPO market, which it projects will account for about $500 billion beyond the $1.2 trillion estimated for the traditional information technology industry.
All of these efforts play to what is perhaps IBM's greatest strength: its size. That alone has lent credibility to on-demand computing, regardless of the initiative's debated merits, and has influenced others in the industry to follow suit. Hewlett-Packard, Sun, Oracle, Microsoft, Computer Associates International and Veritas Software are all pursuing the notion of data centers that can automatically meet spikes in computing demand. Ultimately, corporate customers will save money by only paying for the computing resources they use.
"IBM is setting the model for what an IT provider ought to be, as we move away from the era of IT vendors just selling hard goods," Illuminata analyst Eunice said. "Many of IBM's competitors are quite explicitly aping IBM."
CNET News.com's Michael Kanellos contributed to this report.