The software business is full of dramatic, pivotal moments: moments when the gut-wrenching pressure can be almost unbearable, when egos and reputations are on the line — and when, in the end, you either win or lose. At Tivoli Systems Inc., one such moment is game point in a round of office headball.
Headball is something that software developers at Tivoli dreamed up while cranking code deep into the night, and it’s become a favorite pastime. Basically, it’s handball using heads instead of hands; the ball is green and purple, and the size of a volleyball. The game today is between Mitch Medford and Marty Stich, and the score is tied at four all. Medford, 36, director of work-group systems, is stocky, like a catcher, but his trash talking compensates for his lack of agility: “Here we go! Game point! He can’t handle it!” Medford launches a serve that bounces off his forehead and then off a wall. Stich, 34, a lanky staff engineer who has the quickness of a tennis player, lunges headfirst for the return and misses. Game, set, and match to Medford. He howls with delight. Then he and Stich return to work.
Tivoli could be any startup in Palo Alto, Austin, Seattle, or Boston. There’s a sense of mission in the air. The staff is serious and passionate about its work, yet the environment remains playful and irreverent: A hallway doubles as a bowling alley, with plastic water bottles substituting for pins. One office has been converted into a game room, complete with a foosball table. And everyone is fair game for the top-10 lists posted on a wall (“Top 10 reasons why the strongest hurricane of the year was named after Mitch — No. 4: What else is big, loud and disruptive to people’s lives?”).
But this isn’t a startup at all. Tivoli is actually part of IBM. Headball Central is located in the heart of IBM’s wooded campus in Research Triangle Park, North Carolina, between Raleigh and Durham. Forget crisp blue suits and starched white shirts. This operation represents the future, not the past. It is IBM’s answer to a set of vexing questions that haunt big companies everywhere:
Must the advantages of size — financial resources, clout with customers, deep reservoirs of talent and technology — come at the expense of speed?
Must a commitment to professional behavior — clear lines of authority, sound rules for making decisions — come at the expense of agility and flexibility?
Must a proud tradition of success — a well-honed sense of “how we do things here” — come at the expense of changing with the times and doing things differently?
In short, must being big mean being out of touch?
IBM’s answer to all of those questions is no. IBM bought Tivoli, which is headquartered in Austin, Texas, in early 1996. The acquisition raised eyebrows and expectations throughout the computer world. Some pundits commented that IBM had paid too much — $743 million for a seven-year-old outfit with annual revenues of just $50 million. But there was a clear strategic fit. Tivoli was strong where IBM was weak: Tivoli makes powerful software that manages networked PCs and distributed systems from a single location, regardless of the platform — IBM, HP, Microsoft, Sun — on which they run.
But the deal involved more than importing new technology or inventing new strategies. The real objective was to accelerate change. IBM did not fold Tivoli, the young upstart, into its existing business — as is customary in such lopsided acquisitions. Instead, IBM placed Tivoli in charge of IBM’s entire systems-management business. Almost overnight, Tivoli boosted its workforce by a factor of 5 (from 236 people to 1,300 people) and increased its revenues by a factor of 10 (from $50 million to more than $500 million). It was an acquisition all right, says Frank Moss, 49, Tivoli’s then-CEO — but it was a “reverse acquisition.” At the staff meeting announcing the deal, Moss dispelled fears about Tivoli’s future by projecting an IBM logo onto a screen. Below the logo were the words “A Tivoli Subsidiary.” He was only half joking.
Three years later, that reverse acquisition is a promising new model for how to make big change fast. It blends the power of a big company with the speed of a small one — and produces something that neither company could create on its own. The result is, to be sure, an unlikely mix. Suits meet shorts. Wing tips meet sandals. Big Blue meets little red. (Tivoli’s corporate color is bright red.) But sometimes opposites do more than attract — they blend.
Tivoli veterans admit that they still occasionally study the org chart of the new Tivoli to trace the bloodlines of the people they work with: Blue signifies original IBMers; red, original Tivolians. But when Bill Foster, 41, a product-verification manager based in Raleigh, looks around, he mostly sees a third color. “We’re not blue, and we’re not red,” Foster boasts. “We’re purple. We’re the best of both worlds.”
From Blue to Purple
Tivoli and IBM joined forces because they needed each other. By 1996, the young company had doubled its annual revenues for four years in a row. But Tivoli was still a David — a small, overmatched underdog — compared with Computer Associates International Inc., the aggressive Goliath of the systems-management business. The deal with IBM made this David bigger, stronger, and more resilient. As Todd Praisner, 34, Tivoli’s director of infrastructure development, puts it, “We got a steroid injection.” Suddenly Tivoli had access to IBM’s global salesforce, its abundant technical and R&D resources, and a portfolio of software applications designed to manage mainframe systems.
Since the acquisition, that steroid injection has created a lot of muscle. Tivoli now has 3,600 employees around the world and annual revenues of more than $1 billion. Talk about a blend! The company celebrated its passing of the $1 billion revenue mark with a Tivoli-style party on an IBM-style budget: Lyle Lovett performed private shows for Tivoli’s employees, in Austin and in Raleigh, charging a hefty $75,000 per event.
But the lasting impact of the IBM-Tivoli deal is best measured not in terms of numbers but in terms of behavior. Just as IBM has bolstered the young company’s business performance, so Tivoli has reshaped IBM’s human performance — changing how IBM’s people think, work, and make decisions. Most of the IBMers who joined Tivoli no longer sound like IBMers. Or rather, they no longer sound like what IBMers used to sound like. They say that they work in a new way, with a single-minded focus on their customers and on their competition. They say that they attend fewer meetings. They say that they’re free to take risks, that they see the link between their work and the bottom line, that they can raise questions about anything with anyone. It’s as if the IBMers in the Tivoli unit work for a different company. And, in a sense, they do.
“When someone asks where I work, I tell them that I work for Tivoli,” says Lynn Wilczak, 41, vice president of enterprise R&D at Tivoli, who has been an IBMer for 18 years. “I don’t know whether it happened right away or whether it happened when I got my new business cards: My cards don’t have the name IBM anywhere on them.”
Tivoli’s influence is visible in the red Ferrari that rotated to the unit’s top U.S. salesperson each quarter in 1998. It’s visible in the Texas-style saloon doors at a converted IBM development lab in Rome. It’s visible in the free soda (or, in Rome, espresso) that’s available in company break rooms. And, reflecting a long-standing tradition, it’s visible in Tivoli’s Friday-afternoon beer bashes.
It’s hard to overemphasize the significance of these Friday gatherings. Since Tivoli’s early days, its people have gotten together over beer. (In Rome, employees drink wine.) A typical Friday afternoon would involve a few hundred mild-mannered employees mingling under the trees in a courtyard outside one of Tivoli’s offices in Austin, passing around Tivoli bottle openers, emptying six coolers of beer, talking about software. “You wouldn’t believe how much work gets done at one of these,” one Tivolian remarks. The weekly event creates a unity among the Tivoli staff that’s rare in most companies.
Which is why, at the official announcement of the acquisition, back in 1996, one of the first questions that people from Tivoli asked John Thompson, 56, the powerful head of IBM’s software division, concerned the beer bashes: Would they continue? The answer was not long in coming: At the next bash, held a few days later, a large banner announced that Thompson had provided the beer.
But Tivoli’s reverse acquisition hasn’t worked just because an IBM big shot bought some beer. It has worked because rank-and-file IBMers were quick to buy into what Tivoli represented. Despite their loyalty to IBM, people in the company’s systems-management unit had grown frustrated with its conflicting objectives: Support IBM products — but also dominate the market for other products. For these people, working with Tivoli didn’t require an uncomfortable adjustment. Rather, it made for a better fit. “These were Tivoli people who just happened to work at IBM,” says Martin Neath, 35, executive vice president of Tivoli.
Lynn Wilczak was one of those people. She was the IBMer who made the final presentation to Thompson about buying Tivoli. At the time, she didn’t know where she would wind up in the new company. But that wasn’t important. She longed for the systems-management group to move faster — to let it rip, like Mark Martin, the NASCAR driver whose picture adorns her office in Raleigh.
“I will never forget sitting in our first meeting, with Tivoli management on one side and IBM management on the other side,” Wilczak remembers. “I was just so excited about the opportunity that when Frank Moss went around the room and asked, ‘What do you folks think?’ I told him, ‘I absolutely love this fast pace. Let’s make decisions and go.’ “
So Moss told her that he wanted a demo of a new product that would bridge mainframe and desktop environments. Then he gave her five days to complete it. She says that her jaw dropped. If working at IBM had often seemed like running an obstacle course — trudging through bureaucracy, hopping over corporate initiatives — then working at Tivoli would be a flat-out sprint. The demo ended up taking 10 days to complete. But only six months later, the product shipped. “That’s what I mean when I tell people that Tivoli is focused,” says Wilczak.
But life at Tivoli has represented more than a change of pace for most IBMers. Indeed, it has represented a whole new way of working. Leo Cole, 41, director of network management, knew about Tivoli before it joined IBM. He had been watching the company since 1992, and he couldn’t help being impressed by — and a little envious of — its speed and innovation. Now he understands the roots of both. In his 16 years at IBM, Cole had worked with dozens of developers and executives, but he had rarely socialized with them. They knew one another as coworkers only, not as friends, and that unfamiliarity bred a certain formality. Cole just didn’t feel comfortable speaking with honesty and candor. IBM resembled a big family that couldn’t communicate; Tivoli, he discovered, is like one of those families that just let it all hang out.
“At Tivoli, you have a more open relationship with your peers — even with the top executive team — because you know them,” says Cole. “If I feel that a Tivoli executive is wrong, I can say so. I probably had that freedom before, but because I didn’t know that person, I didn’t feel that I could say anything. Here, we’re not only coworkers — we’re friends. It’s very powerful when everybody has the same goal, when you really know your teammates, and when you trust them. We know who the real opponent is: It’s the competition; it’s not each other.”
To IBM veteran (and headball champ) Mitch Medford, working for Tivoli means working for a software company instead of a hardware company. “IBM is one of the biggest software companies in the world,” he says. “But we didn’t know how to develop software in the same way that everyone else does.” After the acquisition, he and his fellow IBMers received a crash course in that skill. Tivoli’s leaders gave him a product, a budget, a deadline — and total control. They also gave him the chance to develop software that involved fewer documents and more code than what he was used to. Typically, IBM developers had to go through a long review process before coding could begin. Medford prefers an approach called “iterative prototyping”: Code a prototype, allow other programmers to “touch” it and critique it, and then revise, revise, revise. “It allows for a more natural evolution of software,” he says.
Last year, Medford’s team worked as many as 90 hours a week to get a new product out the door. Instinctively, Medford adopted what Tivolians call a “rock-star management” style: He bought his programmers dinner and brought in cots during late-night testing sessions. He gave each of their spouses a dozen red roses and a $200 gift certificate redeemable at a local mall. Sensing that his team members needed a break, he rented a theater and treated them to a private screening of Star Wars. And, of course, he not only let them play headball — he learned how to play the game himself.
Another brainstorm of Medford’s was to let members of his development team decide who should be awarded bonuses. He gave each of them money to distribute among his or her peers. In that way, the members were able to recognize the unsung heroes and to show their appreciation for one another. “A lot of my programmers say that they couldn’t go back to traditional IBM practices,” says Medford. He couldn’t either. “Before Tivoli, I spent a lot of time navigating internal processes, fighting political conflicts, trying to get resources, playing games — just to get the job done,” he recalls. “It probably took 20 people’s signatures to launch a product. When we launched Tivoli IT Director, two people signed off on it. That was all it took.”
At the time of the acquisition, a lot of IBMers in Raleigh were a bit uneasy about their new colleagues. Tivoli’s confidence came across as arrogance, and because Tivoli was in charge of the systems-management business, some IBMers feared layoffs. Ken Edwards, 42, then a product manager in Raleigh, was one such skeptic. Unlike Cole and Wilczak, he was surprised by the deal, and like other longtime IBMers, Edwards, a 19-year employee, wasn’t sure where — or whether — he would fit in. “I’m the kind of person who tends to get into a regimented pattern and to stay there,” he admits. “And IBM is the kind of place where you can easily get a little brainwashed about how you should act.”
Nonetheless, he quickly became a convert. Several factors helped to win him over: the number of decisions that were made over the phone or through email, rather than in meetings; the environment at Tivoli; the holiday party in Raleigh, during which IBMers danced and socialized with colleagues, in many cases for the first time; the quarterly bonuses (IBM’s bonuses are annual); the sense of autonomy. That sense of autonomy is especially attractive: Edwards, who is now vice president of application-management R&D at Tivoli, says that he no longer fears being “escalated” (IBM-speak for “going over someone’s head to the boss”). “When people tell me that they’re going to escalate, I tell them, ‘You can, but I don’t know who you’re going to escalate to, or how far you’re going to get. The first thing that person is going to ask you is, Did you talk to Ken? So it’s best for us to work through the issue,’ ” Edwards says.
From Red to Purple
At the time of the acquisition, the folks at Tivoli were experiencing their own sense of apprehension. Many of them were happy to accept the resources that IBM had to offer, but they feared that they might be smothered in the process. People who had come to Tivoli as refugees from big companies worried that they would experience bureaucratic déjà vu. Almost everyone at Tivoli was concerned that their company would lose its culture.
Martin Neath, 35, vowed to himself that none of those fears would be realized. Neath, the seventh employee ever to work for Tivoli, had risen through the ranks at the company, going from programmer to executive vice president, and in the eyes of his coworkers, he as much as anybody else embodied the company’s spirit. His office in Austin is the unofficial Tivoli museum. On his bulletin board are the single-malt scotch labels that inspired the code names for Tivoli products; a copy of former CEO Frank Moss’s statement of company goals from 1994 (Number one: “Make Tivoli a great place to work”); a stock certificate from 1995, the year in which Tivoli went public; and a “Liquidator” water gun that was used in Tivoli’s legendary shoot-outs. Neath also has memorabilia from the company’s ongoing war with Computer Associates, including a roll of toilet paper featuring the face of CA’s CEO. But the most significant artifact is his desk: It happens to be the company’s original conference table. He remembers when all of Tivoli’s employees could fit around that table, just like a family.
Now that Tivoli was part of IBM, Neath was determined to preserve Tivoli’s status as a close-knit company and as a “high-energy, results-oriented kind of place.” That small-company spirit, after all, is what fueled Tivoli’s survival and success during its early days.
“Systems-management software is important stuff, but it ain’t that sexy,” explains Mark McClain, 36, Tivoli’s vice president of marketing. “Frank Moss made it sexy, though. He made us believe that we were a cutting-edge company doing cool stuff. It’s like the famous story of the three bricklayers: The first says, ‘I’m laying bricks,’ the second says, ‘I’m making a wall,’ and the third says, ‘I’m building a cathedral.’ Tivoli is that third bricklayer.”
History also figured into the fears of many Tivolians. After all, IBM was largely responsible for Tivoli’s creation in the first place. Were they now characters in some Dickensian plot twist? In the late 1980s, more and more companies were moving away from their centralized mainframe computers and toward distributed networks. In response to that trend, a handful of IBMers in Austin wanted to build software that would help manage the far-flung systems on those networks.
IBM chose to pass up that opportunity. But three IBMers — Robert Fabbio, 41; Steve Marcie, 35; and Todd Smith, 47 — grabbed it. They left IBM to start Tivoli; the company was incorporated in August 1989.
“We envisioned a single product to manage computers on a network, and as far as we knew, nobody had done that before,” recalls Marcie, now a technology ambassador for Tivoli. That lack of precedent, he adds, “is what created most of our problems inside IBM.” As it turned out, they solved those problems so well that their startup managed to attract financing from top venture-capital firms, including Kleiner Perkins Caufield & Byers — the VC outfit that helped launch other well-known software startups, such as Lotus and Netscape.
Seven years later, when they were considering the deal with IBM, Moss, Neath, and the rest of the Tivoli executive team understood that IBM would give them a competitive edge against Computer Associates. But how would the two companies work together? “The first pillar of the Tivoli culture was a willingness to take risks,” says Moss. “There was a real sense of urgency. We used to say, ‘No mañana’ — there is no tomorrow.”
The first year after the deal took place, Neath says, “Tivoli was basically in denial.” Indeed, the company was so worried about preserving its independence that it largely ignored the rest of IBM. John Thompson helped ease the transition by creating an office that screened the flood of concerns and questions that came from employees at both companies. Meanwhile, Tivoli made decisions not according to IBM policy but according to the demands of the market. Thanks to Thompson, people at the smaller company had permission to say no without fear of being “escalated.” This decision-making ability made sense early on. “We wanted to preserve and grow Tivoli and to use the things at IBM that made sense,” Neath says.
Eventually, though, Tivoli began to let down its guard. Even the most die-hard Tivolians began to recognize that IBM could be useful to an operation that had 1,000 new employees, along with huge new customers from around the world. “We were like [NBA sensation] Kobe Bryant,” says McClain. “We went straight from high school to the pros. We had a lot of raw talent, but we had missed four years of college, so we had to learn a lot on the fly.”
Like most small software companies, for example, Tivoli had lacked a sophisticated process for testing its software. Its coders far outnumbered its testers; bugs got worked out in the field. Today, using thousands of computers at test labs in Austin, Raleigh, and Rome, Tivoli can simulate the dynamics of customer networks and put new software through realistic test runs.
“As an organization, Tivoli got fast-forwarded about five years,” says Todd Praisner. “And some of us weren’t prepared for that change. I heard IBM people talk about being a ‘mature software company’ so much that I wanted to throw them out of my window. But now I realize that they understood what it means to be a billion-dollar software operation.”
Praisner didn’t expect to stick around after the merger. He didn’t see himself joining some “nameless, faceless, corporate tide.” After all, back in 1992, he had left a big company, General Dynamics, to work for Tivoli. So why has he stayed? Because, he says, “none of the things that attracted me to Tivoli have changed. We have cool machines and work on cool software, we get all the caffeine that we could possibly consume, we get to walk around in shorts, and we get to work with really smart people.”
As with any merger, of course, some people did leave: IBMers who couldn’t fit in with the Tivoli style; Tivolians, including Moss and most of his management team, who missed the raw thrill of working at a startup. But, according to Tivoli, the turnover rate has been remarkably low, hovering at 5%. And of Tivoli’s first 236 employees, 203 remain with the company. No wonder that when Bill Foster looks around, he sees purple.
Tivoli’s chairman and CEO, Jan Lindelow, is a purple guy. Lindelow, 53, a former engineer in Sweden and a former executive at Asea Brown Boveri, a European engineering giant, appreciates both sides of Tivoli: its culture and speed, and its access to IBM’s resources. “I see myself as having two jobs,” he says. “One is to lead this company — to be visionary, to make sure that we are efficient in our execution. The other is to be a bridge to the parent company. I’m proud to be part of the senior IBM management team. I don’t make any attempt to hide that.”
Today’s Tivoli is the fastest-growing segment within IBM’s $14 billion software division. Despite its relatively modest size, Tivoli is often mentioned in IBM’s quarterly earnings reports. Since its acquisition by IBM, Tivoli itself has acquired three companies of its own. It has also established a global presence: Half of its business is now international. The development process of the combined operation is twice as fast as the same process was at the old IBM, and the software is more polished than it was at the old Tivoli.
Last September, for example, the company released Tivoli Enterprise, its most powerful software package and its “highest-quality” release yet, according to Neath. Tivoli beta-tested the product for eight months, working with 50 of its biggest customers until the software met with their approval. Tivoli’s software has to be better, Neath says, because the stakes are now higher. Only a few years ago, when the company passed the 1,000-desktop mark, Tivoli celebrated by tossing Frank Moss into a fountain. Now its software manages hundreds of thousands of desktops — and Tivoli is counting on Tivoli Enterprise to generate about $1 billion in revenue from customers like Ford, USAA Real Estate Co., Reuters, Wachovia Corp., and AT&T.
Tivoli is playing for higher stakes within IBM as well. Its culture is rubbing off on other parts of the parent company, creating ripples of change throughout the organization. IBM’s other software groups are copying Tivoli’s market-driven focus, its solution-oriented sales approach, and, occasionally, even its beer bashes. And IBM’s hardware groups are working with Tivoli to make their products Tivoli-ready. These days, in fact, Tivoli is often the one that approaches IBM with suggestions — rather than the other way around. “IBM gave Tivoli a lot, and we’re just trying to give a little back,” says Neath.
Indeed, things are looking more purple every day. “For those customers who say, ‘I want to know that you’re part of the IBM team and the IBM solution,’ we can be that,” says Steve Basile, 39, director of technical evangelism. “And for those customers who say, ‘I want Tivoli, the independent software company,’ we can be that as well. We can use whatever persona helps the customer — without betraying one persona or the other. It’s the best possible relationship.”
IBM and Tivoli Systems are setting the agenda for fast change — blending the power of a corporate giant with the agility of a startup. Here are some of the principles that guide the work of this billion-dollar “small” company.
1. The front line drives the bottom line.
One reason that Tivoli moves so fast is that Tivolians are comfortable with taking stands, speaking their minds, and challenging authority. And that’s because the company’s senior executives believe that the most valuable ideas often come from the people with the least seniority — the people who actually create Tivoli’s software. “The smartest employees are the ones on the front lines,” says Mark McClain, 36, vice president of marketing.
2. Business is not a spectator sport.
One of the big problems with big companies, argues Tom Bishop, 45, chief technology officer and vice president of the embedded-solutions business unit, is that their size creates too many opportunities for people to sit and watch. “You have too many fans watching too few people play the game,” he says. “You have to look at your people and ask, ‘Are they just going to meetings and writing reports — or are they actually writing code, doing work?’ “
3. Just the (minimal) facts, ma’am.
The software business, like most businesses, is a moving target. Wait too long to introduce a new product, and it will be obsolete before it arrives. That’s why Tivoli doesn’t overstudy or overanalyze a problem. Instead, it defines the “minimum set of facts” that it needs to make a product decision, and then it hustles the software to market — after which it adjusts the product as necessary.
4. Focus on the real enemy.
The reality inside most big companies is that lots of people spend time jousting with internal rivals rather than battling competitors in the marketplace. Why haven’t such internal divisions emerged at IBM and Tivoli — two companies with vastly different cultures? Because people from both companies rallied around a shared passion: beating their archrival, Computer Associates (CA).
After the 1996 acquisition announcement, for example, CA ran an ad in Austin inviting Tivoli employees to a recruiting event. “Tivoli just lost its independence,” the ad declared. “That doesn’t mean you have to.” Tivoli fought back. Some of its newest employees flew in from Raleigh to make signs that read, “CA? No way!” The signs lined the road to the hotel where CA was holding the recruiting party — and where Tivoli threw a party of its own at the same time.
5. Compete with yourself — before someone else does.
It’s hard for big companies to get beyond their first big success. That’s why Tivoli created “the courtyard gang,” a small team of marketers and developers who are charged with scrutinizing the company from the perspective of its competition.
The team strives to find vulnerabilities in Tivoli’s products or in its business strategy before another company finds — and exploits — them. “The courtyard gang tries to create a scenario that would put Tivoli out of business,” says Mike Turner, 33, vice president of Tivoli’s enterprise business. “After all, if we don’t come up with it, someone else will.”
Chuck Salter (firstname.lastname@example.org), a Fast Company contributing editor, is based in Baltimore. To learn more about Tivoli Systems inc., visit the Web (www.tivoli.com).