Who could have predicted it? In an era when the Internet is everything, Silicon Valley is the center of the universe, and young software programmers push rock stars off the covers of magazines, IBM matters again. Big Blue's turnaround is one of the most unexpected comebacks in corporate history.
In April 1993, when CEO Lou Gerstner arrived, IBM looked like one more giant company bent on self-destruction: out of touch, out of steam. Younger, hungrier rivals were stealing its best markets and attracting the best talent. There was lots of talk, in lots of places, of an AT&T-style breakup.
Less than five years later, IBM is back. The company is growing. It is a genuine presence on the Net. It even wants to be cool — relevant to the programmers flocking to Netscape, Starwave, and other fast-moving software companies.
"We used to call them the ponytail brigade, the black turtleneck brigade," says David Gee, 30, IBM's point man in Silicon Valley. "Now they're PIBs — People in Black. We have to be relevant to PIBs."
Gerstner played a decisive role in IBM's comeback. He understood that the company's future lies not in tearing itself apart but in pulling itself together under the cry of "network-centric computing." He became a force for discipline and accountability in an organization coasting on its reputation.
But no turnaround this far-reaching can be the work of a CEO alone. IBM's comeback has been a grassroots revival, driven by a network of leaders who made it their business to change the company. What follow are profiles of these in-the-trenches leaders. They have different titles and work in different parts of the company — and of the world. But they share a passion for their jobs, for the Net, and for making a difference inside IBM. They have created principles and practices — "rules for radicals" — relevant to change agents in any company.
These IBMers are working on the edge. But they relish the challenge. "You want to know why none of us is insecure about what we're doing?" asks David Gee. "Because we're doing the right thing. We're at the forefront of what's coming."
Where do you find the future?
It was late 1993, months after Lou Gerstner had arrived to "save" IBM. As the world around them was collapsing, IBMers around the world were asking the questions employees in all big companies ask during times of crisis: What went wrong? Who was to blame? People were looking inward and pointing fingers.
Well, most people. John Patrick, one of IBM's senior strategy executives, wasn't casting blame. He was messing with his computer. "I was experimenting with Gopher," he says, referring to the Internet software utility. "I became captivated by the idea of sitting at home and cruising around someone else's computer. Being remotely connected was hardly a new idea at IBM. But being inside someone else's computer and having a standard that meant it didn't matter what kind of computer either of us had — a light went off. I thought, this is going to change everything."
In a way, it did. It would be an overstatement to argue that all of what's exciting about IBM's presence on the Net goes back to Patrick's "Gopher epiphany." Lou Gerstner created intense pressure from above. But it was Patrick who let loose pressure from below. Ask in-the-trenches IBMers who has really pushed the company onto the Net, and the name you hear again and again — second only to Gerstner — is John Patrick.
"It's not as if I was the only person in the company who believed in the Internet," he says. "There were lots of people at the grassroots level who thought the same thing. I suppose I became a spiritual leader for many of them."
Patrick, 52, is an unlikely change agent. He is true Blue. He joined IBM in 1967 and worked his way up from jobs in sales and marketing to CFO positions in several business units. As vice president of marketing for personal systems, he was responsible for creating the ThinkPad brand. He works out of IBM's corporate complex in Somers, New York.
But if Patrick's career looks conventional, his attitudes and activities have always defied IBM convention. Even in the dark days of 1993, when most IBMers were polishing their resumes, Patrick, an insatiable gadget freak, was doing what he had always done — thinking about the future and tinkering with ways to get there.
Soon after his Net conversion, Patrick wrote a manifesto called "Get Connected." It identified six principles that would reshape industries and reinvent companies. Each principle came with action items: give every employee an email address; create internal newsgroups; build a corporate Web site.
These ideas are conventional wisdom today. But when Patrick first published them, they were new, exciting — dangerous. "The paper wasn't about getting physically connected," he explains. "What I meant was, Get with it. Connect with other people. If you become externally focused, you can change the whole company."
"Get Connected" got results. Patrick received memos, phone calls, and emails from IBM offices around the world. How could a white paper generate such red heat? Because, he argues, people who've seen the future want to get there.
"People didn't know where I reported in the company, and they didn't care," he says. "We shared a common vision that the Internet was going to change everything and that IBM should be a leader."
Patrick turned that energy into a movement. He began a Get Connected mailing list (using email). That list quickly became the Get Connected team — a virtual organization whose members cut across functions, divisions, and time zones. "We had no budget, no head count, no authority," Patrick laughs. "Everything we did was informal."
The obvious place for the Get Connected team to start was building IBM's corporate Web site. That site http://www.ibm.com debuted on May 24, 1994 — less than six months after Patrick distributed his manifesto. "It was one of the first significant corporate Web sites in the world," he says.
Meanwhile, Patrick kept pushing himself deeper into the Net. The same month that IBM's Web site debuted, he visited Internet World, an industry gathering in San Jose, California. The show was in its infancy, but to Patrick it was a revelation. With no authority to do so, he signed up IBM as a major participant in the next Internet World, in Washington, DC, seven months later.
Then Patrick issued a call to action. He needed the Get Connected team to design and staff the show. He started visiting IBM divisions, tin-cupping them for funds. "I went to the RS/6000 guys and asked them for $5,000," he remembers. "I went to the IBM Global Network and got another $5,000. The next thing I knew, I had the money."
The night before the show, Patrick convened a meeting in his Washington hotel room. He still wasn't sure what, or who, he had to work with. He found out: 54 people representing 12 IBM units had marched on Washington. "We dominated the show," he says. "And the amazing thing was, you couldn't find us in any IBM budget."
IBM soon issued a comprehensive statement of its Internet strategy. Then it convened a task force to turn strategy into reality. On December 1, 1995 IBM created an Internet division responsible for defining the company's Net initiatives. Patrick became vice president and chief technology officer. Nearly 600 people were assigned to the group. The virtual team had become a corporate force.
"People came to me and asked, 'Are you disappointed to see the Internet division take this over?'" Patrick recalls. "Disappointed? It's a victory! We won. That's why we've been doing all this."
Patrick says he can feel a difference at IBM. "Years ago," he says, "you could hear a pin drop in the halls. Today you can feel the excitement. Kids are building Java applets. People are in a hurry. We just have to keep moving faster."
How Do You Reinvent "Not Invented Here"?
It is a sublime midsummer day outside IBM united kingdom Laboratories Ltd. in Hursley Park, England, 90 miles southwest of London. A lonely cloud drifts above 18th-century Manor House, the lab's central office. A cricket match is in progress.
The setting is so postcard-perfect. So terribly British. Which is why it's hard to believe that Hursley is ground zero in a technical revolution with the potential to transform IBM. But it was here that two thoroughly respectable IBM scientists identified the revolutionary potential of Java, the programming language created by Sun Microsystems, and championed it to the point that Java is now a cornerstone of the Internet strategy that IBM calls "e-business."
Talk about grassroots impact. In the summer of 1995, out of 220,000 IBMers worldwide, Mike Cowlishaw and Ian Brackenbury were the only two working on Java. Two summers later, more than 2,400 IBM scientists and engineers on three continents are working on the language. In fact, the company is the biggest Java developer on the planet — with more Java programmers than Sun itself.
"There is no precedent for what has happened," says Brackenbury, Hursley's chief scientist. "It was a miracle."
Mike Cowlishaw saw it first. A quiet, self-effacing software guru, he is one of only 50 IBM Fellows worldwide. It is a lofty status that he earned as one of the company's top computer-language experts. His biggest claim to fame is a programming language called Rexx that he invented nearly 20 years ago. Last year he created NetRexx, a version designed to run on Java.
Cowlishaw recognized Java's potential in May 1995, the moment Sun released the language to the outside world. Most programmers saw it as a tool for jazzing up Web sites. Cowlishaw saw it as something more: a language that could deliver the holy grail of "platform independence." Any computer equipped with Java's "virtual machine" can run any Java program — no matter what operating system it uses or how the application was written. Cowlishaw sketched a chart showing how Java's virtual machine could lead to platform independence: "I thought to myself, I've seen this picture before."
The picture that came to mind was more than 30 years old. It was the IBM System/360 family of mainframes. The System/360 debuted in 1964 and generated $16 billion of revenue. Why such success?
Because its different hardware models all used the same assembly language. It was an early, IBM-only example of platform independence. Java "captures the same idea we had in 1964," he marvels.
By 1995, of course, the dream of platform independence had given way to the reality of partisan competition among hardware standards and operating systems. Wars over standards were good for companies but terrible for developers and users.
"I wrote Rexx in 1979," Cowlishaw says. "It took us twelve years to port it to eight different operating systems. When I created NetRexx, the Java version, I originally wrote on OS/2. Then I ported it to eight different platforms in three weeks. There's no contest."
Cowlishaw shared his enthusiasm with Brackenbury. But how could they translate their commitment into corporate momentum? After all, championing Java meant reckoning with one of IBM's deep-seated pathologies — an arrogance endemic to life at most big companies with deep pockets and a proud heritage. "It's the Not Invented Here syndrome," Cowlishaw says. "And not just Not Invented at IBM. If one division invented something, the other divisions wouldn't look at it."
Cowlishaw and Brackenbury invented a plan to tackle Not Invented Here. They arranged a meeting with James Gosling, the Sun engineer who created Java and himself a former IBMer. They put Gosling through a technical cross-examination. "We asked him about every abyss we'd peeked into," Brackenbury says. "He answered perfectly. I was convinced that we had something golden."
In July 1995 Hursley's lawyers obtained a Java evaluation license from Sun. In a five-day marathon coding session, Cowlishaw ported Java to OS/2. Then Cowlishaw and Brackenbury produced an internal white paper — their version of John Patrick's "Get Connected" call to arms. They also began an "information wave." Brackenbury called a meeting of 40 top IBM technologists at Lotus headquarters in Boston. He created an email list of key executives and 200 opinion leaders. He also set up electronic forums where IBMers could swap ideas, urge each other on, and argue.
"We wanted the grassroots to have as much information as possible about developments inside IBM," Brackenbury says. "People could get information on the Web about what was happening outside. But they might not know what was going on inside."
Brackenbury needed bodies too. As luck would have it, IBM had just scrapped a groupware tool called Person-to-Person. People who had spent years on the product were adrift. "Ian mobilized these miserable and disillusioned people," says Simon Phipps, then Person-to-Person's chief evangelist. "We were funded for the rest of the year, and we didn't have any work to do."
Now they did. Phipps himself proved to be a crucial addition to the Java team. A self-described "professional rebel," he radiates exuberance and wit. He was the perfect evangelist for what was effectively a renegade operation. "I've always been seen as a little bit 'other,'" Phipps says. "Why? Because I'm a bigoted, opinionated swine. I have so little respect for other people that I actually express myself. I think that's one of the reasons Brack asked me to join."
By the fall of 1995, the Java religion was winning converts. Projects began sprouting up in labs around the world. That December, the company announced it would license Java from Sun. The Hursley team became IBM's Centre for Java Technology Development. Today IBM has the world's biggest army of Java programmers.
From the beginning, says Cowlishaw, "it was clear that Java could touch every part of IBM. And it has. This project has helped bring people together in a way that no project has in a long, long time."
Should you start from scratch?
This is IBM. The Almaden Research Center is an imposing 40-acre complex, with buildings made of burnished wood and tinted glass, surrounded by 690 acres of rolling hills in south San Jose, California. More than 500 researchers are busy inventing high-powered digital technologies. It is a place about as open to the outside world as a top-secret government installation.
This is IBM? In the basement of Almaden, a dozen young IBMers are busy reinventing how the company does business. They wear T-shirts, blue jeans, earrings. A pirate flag flies above the cubicle of the team's Webmaster. "Tekken 2", the Sony PlayStation game, runs on a screen in the conference room. A poster that looks like a ransom note blares, "I cannot be managed by anybody."
David Gee has presided over this unruly scene since January. He is program director for AlphaWorks — an "online laboratory" designed to change how IBM commercializes products and collaborates with customers. Unlike John Patrick, who hired Gee, or Hursley's Cowlishaw and Brackenbury, Gee is not a veteran IBMer. He joined the company two years ago from Dun & Bradstreet.
"Our team is mostly people new to IBM," Gee says. "This is our first job at the company. And part of our job is to shake up the status quo. We want to get in trouble. We bend the rules."
Big, successful companies get that way by designing rigorous processes, teaching people to master those processes, and grooming subsequent generations to live those processes. Over time, though, unity easily mutates into uniformity. And uniformity can become obsolescence. Renewal means breaking with tradition, redesigning processes, and recruiting people who play by different rules.
That's what Gee and his AlphaWorks colleagues represent. The team is a start-from-scratch effort to emulate the fast feedback, quick cycle-time, download-for-free model that allows young companies to move at speeds IBM never dreamed possible. It's an experiment to see if Big Blue can operate on Net time.
Gee's organization (www.alphaWorks.ibm.com) is a Web-based community that debuted in August 1996. It now hosts 28 early-stage technologies that registered users are free to demo, comment on, download, and experiment with. More than 60,000 users are part of the community. The site has already helped commercialize five products, including Bamba, a streaming audio-video player for low-bandwidth Net users; PanoramIX, a 360-degree viewer for video and multimedia; and Mike Cowlishaw's NetRexx.
There's no denying what alphaWorks has achieved. But the team's mission is as much about style as substance. Gee's job is not just to release a stream of new products faster to the outside world. It's to infect IBM with new attitudes from the outside world — to bring the reckless abandon of the Net to a company that has valued caution above all else. That's why alphaWorks is based in northern California rather than Westchester County — to be nearer the center of the Internet universe than the center of the IBM universe.
Gee and his colleagues revel in their outsider status. The team has assembled a worldwide network of supporters called FOAs — Friends of alphaWorks. FOAs get email updates on the group's progress, reports on success stories, requests for ideas. FOAs also receive a steady flow of what Gee calls "trash and trinkets" — lab coats, coasters, magnets, and other silly artifacts that exude a sense of play and adventure. "They think it's cool," he says. "That's what we want them to think."
"Cool" is not a word often associated with IBM, but it's a critical piece of the alphaWorks vocabulary. That's because changing IBM's image in Silicon Valley — making the company a cool place to work — is central to the team's mission. "We're probably the most unpopular company in the Valley after Microsoft," Gee concedes. "Over the last five years, if you told your friends you were going to work for IBM, they'd say, 'What? Why? Go work for Netscape, go work for Sun.' Part of our mission is to make IBM a dynamic and exciting place to work."
Gee claims that perceptions are shifting already: "alphaWorks represents a company with network-centric computing ingrained in its DNA," he says. "A company that doesn't touch, create, or develop anything that is not Net-centric, Net-ready, open-standard. I'm not going to tell you that we've changed the world. But we are making a difference. And in a company this size, even a small win is a huge victory."
Can we make money on this?
In June 1996 Hursley Park's Simon Phipps led a guerrilla assault on the JavaOne conference at San Francisco's Moscone Center. It was similar to John Patrick's rebel operatation Internet World two years earlier. Many at the conference were surprised to see such an aggressive IBM presence — none more than Patricia Sueltz, IBM's new vice president for Internet software. "Pat was amazed that we were there and horrified at the potential for damage if there weren't some marketing experience put into place," says Phipps. "I also think Pat realized that Java was a serious part of her portfolio."
Today Pat Sueltz has an even broader portfolio of responsibilities at IBM. As vice president for e-business software, she manages more than 200 marketing specialists and software developers in locations including Westchester County, Silicon Valley, and Raleigh, North Carolina. Her marketing teams create ways to present the company's Web-enabled software and services to IBM's customers. Her developers work on a range of Net-based products. Her overall role is to shape what she calls IBM's "Java fever" into a coherent set of market offerings and messages to customers. Lots of people dream about the Web. Sueltz's job is to make it real.
Every transformation needs its seers — blue-sky visionaries who plant the seeds of change. Every transformation needs its rabble-rousers — young change agents who don't care whom they offend. But there is another breed of activists who are just as important. They are the consolidators: people who can translate the styles of the new guard to company veterans in corner offices. People who embrace new ideas and drive them to the bottom line.
"I'm a change agent," she says. "I'm a cheerleader at times. But I always go back to substance: What does it do for my customers? Does it improve their operations? Does it add value? Is it compelling?"
Sueltz, 45, stands between the grassroots and the gray hairs inside IBM. Gee's alphaWorks crew falls under her umbrella; she is their sponsor and protector. But Sueltz is equally comfortable at the highest levels of the corporate hierarchy. In 1993, soon after Lou Gerstner arrived, she was plucked from a management role in Hursley Park for a high-impact assignment at headquarters — technical assistant to the CEO. Her job was to help Gerstner answer basic questions about IBM's present and speculate on its future.
Sueltz was one of the first people to show Gerstner the Web: "His reaction was, 'This is great, this is a new channel for business. How do we make it real for customers? How do we make money on it?'"
Sueltz's role was to give Gerstner the material he needed to reach his own conclusions about those issues — not to editorialize. But she didn't keep her ideas and opinions on the shelf. One day, she says, Gerstner described her as "uninhibited." She got very nervous very quickly: "But then he said, 'Keep being uninhibited, keep being curious, keep driving this.' It was his exhortation to be wild. I took it as a compliment."
Sueltz left after 17 months as Gerstner's technical assistant, to return to line responsibilities. When she had started her job, IBM's stock was at $44 a share. When she moved on, it was at $100.
"I'm one of Gerstner's guerrillas," she declares. "Lou encouraged me to push the edges, to push every boundary if it meant finding new ways to help our customers. It really was — and I know this is an overused word — empowering. I knew that I could make a difference, that we could change the way we do business."
Sueltz had acquired a taste for the excitement that comes from operating on the edge — and wasn't shy about sharing her convictions and techniques for getting things done. For example, she became frustrated that too many IBM meetings were "collective monologues" rather than genuine dialogues. Her response? Buy a basketball and bring it into meetings: "I held it up and said, 'You may not speak unless you're holding the ball.' People thought I was nuts. But it worked. We've used that in a lot of meetings."
These days, Sueltz's meetings revolve around a clear agenda: how to turn IBM's new-found energy and enthusiasm for the Net into products that can make money. In the summer of 1996, she worked with John Patrick on a range on Internet technologies for the Atlanta Olympics. For example, Patrick and his band of Webheads built a site that handled an average of 11 million hits per day. Sueltz then evaluated their cutting-edge efforts for innovations that IBM could take to market.
One of her biggest wins from that experience was the Interactive Network Dispatcher, load-balancing software that helped the IBM Olympic site handle massive traffic. The code was posted on alphaWorks for other companies to try and is now available as fully supported commercial software. "It's fabulous," Sueltz says. "That product came out of running the Olympics site with beta-code."
It's also just a small example of her long-term plan for creating new products, delivering new services, helping to make IBM a dominant company on the Net. "I get called by headhunters all the time," she says. "They say, 'We can make you very rich.' But they don't get it. I am going to change the world with this. That's why I stay here. I don't know of any other company that can do what we can do. I'm making a difference."
Eric Ransdell (firstname.lastname@example.org) is a Fast Company contributing editor based in San Francisco.
A version of this article appeared in the October/November 1997 issue of Fast Company magazine.