The New Blue

Lou Gerstner saved Big Blue. Now it's up to new CEO Sam Palmisano to restore it to greatness

The directors were just sitting down for the first IBM (IBM ) board meeting of the year on Jan. 28 when CEO Samuel J. Palmisano dropped a bombshell. For years, the board had lavished wealth upon Louis V. Gerstner Jr., keeping his pay in line with other pinstriped superstars across Corporate America. But in a surprise break from the past, Palmisano asked the board to cut his 2003 bonus and set it aside as a pool of money to be shared by about 20 top executives based on their performance as a team. Palmisano doesn't want to say how much he's pitching in, but insiders say it's $3 million to $5 million -- nearly half his bonus.

A crowd-pleasing gesture? It was just his latest salvo. Five days earlier, he took aim at a bastion of power and privilege at Big Blue, the 92-year-old executive management committee. For generations, this 12-person body presiding over IBM's strategy and initiatives represented the inner sanctum for every aspiring Big Blue executive. Palmisano himself was anointed back in 1997, a promotion that signaled the shimmering possibilities ahead. But on Jan. 23, the CEO hit the send button on an e-mail to 300 senior managers announcing that this venerable committee was finito, kaput. Palmisano instead would work directly with three teams he had put in place the year before -- they comprised people from all over the company who could bring the best ideas to the table. The old committee, with its monthly meetings, just slowed things down.

All the while, Palmisano was piecing together an audacious program to catapult IBM back to the zenith of technology. It started at an Aug. 5 strategy meeting, when he asked his team to draw up a project as epochal as the mainframe computer -- IBM's big bet from 40 years ago. Through the day, the team cobbled together a vision of systems that would alter the very nature of how technology is delivered. IBM would supply computing power as if it were water or electricity. But how to tackle a project this vast? No one knew where to begin. A frustrated Palmisano abruptly cut short the meeting and gave the team 90 days to assemble and launch the megaproject. Three months later, the CEO unveiled "e-business on demand." Standing in New York's American Museum of Natural History, not far from the hulking dinosaurs whose fate IBM narrowly skirted, Palmisano vowed to lead a new world of computing. "We have an opportunity to set the agenda in our industry," he says.

After one year on the job, Palmisano is putting his imprint on the company -- and with a vengeance. Sure, IBM roared back to strength in the late '90s. But Palmisano is out to remake the company and hoist it back to greatness. Through much of the 20th century, under the leadership of Thomas J. Watson and his son, Thomas Jr., IBM not only ruled computing and defined the American multinational, it was the gold standard for corporations. From the days of tabulating machines all the way to the Space Age, when its mainframes helped chart the path to the moon, IBM was a paragon of power, prestige, and farsightedness. It was tops in technology, but also a leader in bringing women and minorities into a well-paid workforce and in creating a corporate culture that inspired lifelong loyalty. "We stood for something back then," Palmisano says.

To return IBM to greatness, the 51-year-old Palmisano is turning the company inside out. He's the first true-blue IBMer to take the reins since the company's fall from grace more than a decade ago. And while the new CEO never criticizes his predecessor, who rescued IBM and pushed many key technologies, Palmisano is quietly emerging as the antithesis of Gerstner. Where Gerstner raked in money, Palmisano makes a point of splitting the booty with his team. While Gerstner ruled IBM regally, Palmisano is egalitarian. The revolution he is leading spells the end of the imperial CEO at IBM. "Creativity in any large organization does not come from one individual, the celebrity CEO," Palmisano says. "That stuff's B.S. Creativity in an organization starts where the action is -- either in the laboratory, or in R&D sites, at a customer place, in manufacturing."

If that sounds like the IBM of old, that's exactly what Palmisano is hoping for. The flattening of the organization, the lowering of CEO pay, the emphasis on teams -- it's all part of his broad campaign to return to IBM's roots. Palmisano believes that core values remain in what he calls the company's DNA, waiting to be awakened. And he thinks that this message, which might have elicited chortles during the tech boom, resonates in the wake of the market crash and corporate scandals. More important, he believes that only by returning to what made IBM great can the company rise again to assume its place of leadership in America and the world.

At the heart of Palmisano's plan is e-business on demand. The project, which is already gobbling up a third of IBM's $5 billion research and development budget, puts Big Blue in the vanguard of a massive computing shift. The company starts by helping customers standardize all of their computing needs. Then, in the course of the next 10 years, it will handle growing amounts of this work on its own massive computer grids. And this won't be just techie grunt work. The eventual goal is to imbue these systems with deep industrial expertise so that IBM is not only crunching numbers and dispatching e-mails but also delivering technology that helps companies solve thorny technical problems -- from testing drugs to simulating car crashes. It's a soaring vision. But Palmisano has believers. "Sam is aiming to go where the market's going, not to where it's been," says Cisco Systems Inc. CEO John Chambers.

The obstacles he faces are immense. Start with the technology. The vision of on-demand computing is downright audacious. It proposes joining all of the thousands of computers and applications in enormous enterprises, and putting them to work seamlessly and in unison -- not only in-house, but with partners and customers. Assembling the pieces will require every bit of IBM's vaunted smarts, and a scrap of luck as well. IBM officials say only 10% of the technology needed for this system is ready. And many of the necessary pieces, including futuristic software programs that will heal themselves, are at the basic test stage in IBM's labs. "There are huge, huge technical challenges," says A. Richard Newton, dean of the College of Engineering at the University of California at Berkeley.

Palmisano faces an equally imposing job at home. To make good on his vision, he must turn IBM itself into a user of on-demand computing and become a prototype for its customers. This entails recharting the path of every bit of information flowing inside the company. It means not just shifting the computer systems, but redefining nearly everyone's job. And if IBM meets resistance to these changes, it could stumble in producing the new technology. This could undermine IBM's $800 million marketing campaign for e-business on demand -- and scare away customers in droves. Such a failure could punish IBM financially, forcing a retreat toward fiercely competitive markets such as servers and chips. "The two most important parts of their business -- services and software -- are tied to the [on-demand] strategy," says Gartner analyst Tom Bittman. "They need to succeed."

Is history on Palmisano's side? Try to think of a great technology company that took a life-threatening fall and then scratched and clawed all the way back to the very top. Westinghouse? Digital Equipment (DEC )? Xerox (XRX )? Some have survived. But if Palmisano leads IBM back to the summit, Big Blue will be the first full-fledged round-tripper.

To get there, he must win a brutal battle raging among the titans of tech. From Hewlett-Packard Co. (HPQ ) to Microsoft Corp. (MSFT ), the industry's bruisers are all pushing research into next-generation computing systems that will rival IBM's. Big Blue appears to be better positioned than its foes, thanks to a wider range of offerings. But, warns Irving Wladawsky-Berger, IBM's general manager for e-business on demand: "In 1996, we had the benefit of being considered irrelevant. [Microsoft's William H.] Gates and [Steven A.] Ballmer felt pity on us. Now they are all watching us. If we don't move fast, they will pass us."

The new initiative provides Palmisano with a prodigious tool to remake the company. Gerstner's reforms began the process, directing IBM toward software and services. But Palmisano's e-business on demand goes much further. It extends into nearly every nook of Big Blue, from its sales force and its army of systems consultants to the big brains cooking up the software code in the research and development labs. Management expert Jim Collins, author of Good to Great, says Palmisano's willingness to think and act boldly bodes well, and recalls earlier outsize bets in IBM's history, such as the development of the tabulating machine. "It reminds me of what Tom Watson Sr. did during the Depression," he says.

Palmisano already is banking on winning his share of the new business. Last year IBM saw revenue slip 2%, to $81.2 billion, with earnings tumbling 54% to $3.6 billion. But this year Palmisano is counting on e-business on demand to fuel the hottest sales growth at Big Blue since 1995. Analysts predict 9% revenue growth this year. And Palmisano expects 40% -- nearly $3 billion -- to come from new offerings in e-business on demand. These include servers running the free Linux operating system and grid software that pools the power of scores of networked computers into a virtual supercomputer.

By pursuing this plan, Palmisano is fleeing the brutish world of hardware and seeking refuge in profitable software and services businesses. He bulked up for this drive last year by spending $3.5 billion for PricewaterhouseCoopers Consulting and another $2.1 billion for Rational Software Corp., a maker of software tools to write programs. And why not? According to IBM's internal research, 60% of the profits in the $1 trillion high-tech industry will come from software and services by 2005. That's up from 45% in 2002. "We're just going where the profit is," Palmisano says.

And he's leading Big Blue in a way it has never been led before. One year before Palmisano disbanded the Executive Management Committee, he had put in place his management teams for the future. He created three of them: strategy, operations, and technology. Instead of picking only high-level executives for each team, Palmisano selected managers and engineers most familiar with the issues. "Heads are spinning," says J. Bruce Harreld, senior vice-president for strategy. "He's reaching six levels down and asking questions."

Talk to Palmisano for an hour and he'll mention teamwork 20 times. His entire on-demand strategy hinges upon it. Why? For IBM to come up with a broad array of on-demand technologies in a hurry, the whole company has to work smoothly from one far-flung cubicle to another. That means bringing researchers in touch not only with product developers, but with consultants and even customers. Only by reaching across these old boundaries will IBM find out what customers are clamoring for -- and produce it fast.

To head up this process, Palmisano has chosen Wladawsky-Berger, the renowned Cuban-born computer scientist who was IBM's e-biz guru in the 1990s. Today, Wladawsky-Berger's mission is to drive the strategy across the company. In the last two months, he has assembled 28 people working in every division of IBM into what he calls a "virtual team." These are Wladawsky-Berger's on-demand agents. They nose around their areas of expertise, looking for on-demand possibilities. New servers coming out later this year, for example, will be equipped to dispatch excess work to other machines on the network.

Still, it's no easy job coaxing separate divisions to dance in unison. Clashes are common, for example, when IBM's 160,000 Global Service workers descend into the research labs. Last year, researchers were hard at work on a program for supply chains in the electronics industry. Consultants ordered up a quick version of the same program for a carmaker. The two sides battled briefly until the researchers adapted a program for cars, and then went back to work on electronics. The consultants' timeframe, says William Grey, manager of IBM's Finance Research, "is milliseconds. Ours is five years. There's a cultural gap that needs to be bridged."

The key is getting IBM itself to function as an e-business-on-demand enterprise. To drive this message through the company, Palmisano in January grabbed a star manager, Linda Sanford, and put her in charge of internal e-business on demand. Sanford, a senior vice-president, had revived IBM's storage business and was viewed as a bona fide up-and-comer at Big Blue. "I take a senior vice-president who has a great job, and say, 'O.K., you're going to make IBM on demand,"' Palmisano says. "Then, 320,000 people say, 'Holy ..., this guy's serious."

Sanford faces an imposing job. First, she has to supervise the overhaul of IBM's massive supply chain. That means piling $44 billion of purchases into a single system. It's a slog. It means pushing IBM's engineers to switch to company-approved suppliers. Then a procurement rep is assigned to each development team, to make sure that they all use industry standard parts. It's intrusive. But like the rest of the on-demand program, it focuses the company onto a single effort. And it should pay dividends. Palmisano expects the entire initiative to yield 5% productivity gains, worth $2 billion to $3 billion a year, for the next five to 10 years.

Sanford also is working to create an online inventory of IBM's knowledge. She's turning the company's intranet into a giant collaboration portal. One feature is an "expertise locator" that helps an employee find, say, a software engineer with expertise in building databases in Linux. But at a meeting of the operations team at Armonk, N.Y. on a cold mid-February morning, a frustrated Sanford told key executives, including Palmisano, that the concept was a hard sell.

Palmisano, his face cupped in his hands, looked concerned. "There's a huge level of expectation on this portal," he said. "I just hope we can deliver." Sanford responded with a blunt message: If Palmisano wants the portal to succeed, he and his teams must lead by example, offering their own areas of expertise within a 30-day deadline. "We have to lead the way," she said.

For Palmisano, this means rallying the biggest brains and deepest thinkers in the company to the cause. In January he flew to Harvard University in Cambridge for a meeting with IBM's top computer scientists. His message was simple and straightforward: The dream of on-demand computing hinged upon their ability to produce technology breakthroughs.

While scientists are wrestling with future iterations of on-demand computing, IBM's sales team is rolling out the first products. New IBM servers include a feature called "hypervisors." These allow technicians to monitor as many as 100 servers at a time, shifting work from one machine to another. A new program from IBM's Tivoli group performs similar work, patrolling the network, constantly on the lookout for servers running short of memory. When it finds one, it automatically shifts the work to other computers. This is a key aspect of on-demand computing, and a potential money saver. Once systems can distribute work, companies will be able to run their servers at a high level, much closer to capacity. This reduces costs. And if work piles up, customers will ship excess tasks to IBM.

Many of them, IBM hopes, will eventually exit the computing business altogether and ship all their digital work to IBM. American Express likes the idea. A year ago, before Palmisano even came up with the new vision, AmEx signed a seven-year, $4 billion services contract with Big Blue. At first blush it looks like a standard outsourcing deal. The company has shifted its computers and 2,000 tech employees to IBM. But what makes it different is the economics. AmEx pays only for technology it uses every month. The advantages? AmEx is looking to save hundreds of millions of dollars over the course of the contract. And with IBM running the system, says Glen Salow, chief information officer at American Express, "they can upgrade technology five times faster."

Palmisano's vision for e-business on demand stretches beyond the technical challenges to the realm of human knowledge. In the services division, IBM has experts on industries ranging from banking to metals to autos. He wants to gather their knowhow -- "deep process insights," he calls it -- into the systems. Eventually, he sees IBM's on-demand offerings reinventing the company's corporate customers and shaking up entire industries.

IBM is developing 17 different industrial road maps for on demand. Pharmaceuticals is one. There, a computer grid will handle simulation and modeling to reduce the number of clinical trials needed. That, IBM says, could lead to improving the success rates of drugs, now from 5% to 10%, to 50% or better. IBM also believes it can help cut the time it takes to identify and launch a new drug to three to five years, down from 10 to 12, slicing the pre-launch cost of drug development to less than $200 million, from $800 million.

It's a splendid vision -- and far too rich with opportunity for IBM alone. Microsoft has more cash than any tech company -- $43 billion -- and its .net Windows initiative is an effort to rule the next generation of computing every bit as ambitious as Palmisano's. But Microsoft trails Big Blue in the upper end of the corporate computing world. Sun Microsystems (SUNW ), an early advocate of on demand computing, is pushing its own effort to develop software, called N1, that will more efficiently manage Sun gear. Sun claims N1 will offer superior performance -- at one-tenth the cost -- because the software is designed only for Sun products. "Diversity is great in your workforce," says Sun Executive Vice President Jonathan Schwartz. "It sucks in your data center." IBM software head Steven A. Mills shoots back: "Nothing they have in N1 is unique."

A stronger contender is Hewlett-Packard, thanks to its array of hardware, software, and services. Analysts say HP leads IBM in a few important niches. HP, for example, has software called Utility Data Center, that shifts work across all of a company's computers, networks, and storage devices.

For now, IBM's wide-angle vision and broader range of technology gives it the overall lead. But to keep ahead, Palmisano maintains a routine of near-constant work. Even while on a Vermont ski vacation in early March, Palmisano spent a snowy Sunday afternoon reading briefing papers while his family hit the slopes. Rest assured, Palmisano won't be getting his weekends back anytime soon. He is remaking IBM, and that's a job that could last a full decade. If he pulls it off, though, a giant of technology will be reborn.

By Spencer E. Ante in Armonk, N.Y.

    Before it's here, it's on the Bloomberg Terminal.