Read Who Says Elephants Can't Dance?: Leading a Great Enterprise through Dramatic Change Online
Authors: Jr. Louis V. Gerstner
Tags: #Collins Business, #ISBN-13: 9780060523800
The highest levels of IBM executives were almost obsessed with the effort to unwind the decisions of the 1980s and take back control of the operating system from Microsoft (and, to a lesser extent, gain control of the microprocessor from Intel). From my perspective, it was an extraordinary gamble for a company to be taking at a time when it was in such a weak financial state.
The pro-OS/2 argument was based on technical superiority. I can say without bias that many people outside IBM believed OS/2 was the better product. The anti-Windows argument was that the legendary Microsoft hype machine was using clever marketing and wily PR to foist an inferior product on consumers, take greater control of the industry, and, in the process, destroy IBM.
WHO SAYS ELEPHANTS CAN’T DANCE? / 139
What my colleagues seemed unwilling or unable to accept was that the war was already over and was a resounding defeat—90
percent market share for Windows to OS/2’s 5 percent or 6 percent.
Not only were we banging our heads against a very hard, unre-lenting wall, but I had to wonder if anyone was paying attention to the strategic direction we were talking about. If we truly believed that the reign of the PC was coming to an end, why were we pouring energy, resources, and our image into yesterday’s war? Desktop leadership might have been nice to have, but it was no longer strategically vital. Continuing to chase it was more than an expensive distraction, not to mention a source of considerable tension with customers. It was counter to our view of where the world was headed.
The last gasp was the introduction of a product called OS/2 Warp in 1994, but in my mind the exit strategy was a foregone conclusion.
All that remained was to figure out how to withdraw. I asked for alternatives and was presented with three. The first two would have involved fairly abrupt terminations of the product line. The third involved a five- to six-year winding down that would cost us hundreds of millions of dollars but would provide support to allow customers using OS/2 to migrate to Windows- or UNIX-based systems in a more manageable fashion. I think you know the decision a former customer made, and IBM today is providing support for customers who still depend on OS/2.
The OS/2 decision created immense emotional distress in the company. Thousands of IBMers of all stripes—technical, marketing, and strategy—had been engaged in this struggle. They believed in their product and the cause for which they were fighting. The doomsday scenario of IBM’s losing its role in the industry because it didn’t make PC operating systems proved to be little more than an emotional reaction, but I still get letters from a small number of OS/2 diehards.
140 / LOUIS V. GERSTNER, JR.
A Future After the War
With yesterday’s war behind us, it was easier to start planning for what lay ahead. As I took inventory of what we had available to us inside IBM, it was a mixed picture: a software business that was big but fragmented and unmanaged; a software portfolio that was closed in a world destined to be open; software built for mainframes, rather than for smaller and more widely dispersed systems; and a business that, aside from operating systems tied to the hardware, was losing huge sums of money.
We needed far more focus. Toward the end of 1994, I decided to pull together all of IBM’s software assets under a single executive and ask him to build a distinct, stand-alone software business. John Thompson had attracted my attention very early as one of the most thoughtful and capable managers at IBM. He demonstrated a deep understanding of the technology, had a bright and thoughtful intellect, and, perhaps most important for me, he was able to translate arcane technology into business terms.
At the time, John was running our Server Group—the heart of the company. We were managing a critical technology transition, and he’d been in his position for only about fourteen months, so he was shocked when I asked him to take up the task of creating a new business from scratch. But as he did many other times during my tenure at IBM, he accepted the challenge and brought his many talents to bear quickly and effectively.
It is really difficult to exaggerate the enormity of the problem that John faced as 1995 began. IBM had 4,000 software products, all of which were branded with separate names (most of which were un-memorable and un-“rememberable”). They were made in more than thirty different laboratories around the world. There was no management system, no model for how a software company should run, and no skills in selling software as a separate product.
WHO SAYS ELEPHANTS CAN’T DANCE? / 141
Over the next two years John and his colleagues recruited and trained 5,000 software sales specialists; they became the backbone of a new sales function in IBM that eventually reached 10,000 by the year 2000.
John reduced the thirty labs to eight and consolidated sixty brands to six. He built a management team, developed strategies, and created marketing programs. His team redirected hundreds of millions of dollars of research expenditures and, in particular, shifted substantial sums of money into the new marketing and sales functions.
IBM salespeople loved to sell hardware, and why not? That was how they made their quotas and their money. They had little appetite or skill to go up against Oracle or Computer Associates salespeople who were trained exclusively to sell software.
What remained was to find a sense of direction, a focus, a leadership position that would send a message that IBM was serious about software. John thought he had the answer.
To set up the software bet we were about to make, think about software doing its work on three levels:
• At the base, there are the operating systems that tell the hardware what to do.
• At the top, there is all the application software, like a spreadsheet, a program for calculating your income tax, or a graphic design program. This is what an end user sees on the screen.
• In between, there is a collection of software products that connect the two.
At the base: Microsoft owned the dominant operating system, which, regardless of the fate of OS/2, we believed would become increasingly commoditized in a world of open standards.
At the top: Companies like SAP, PeopleSoft, and JD Edwards dominated the applications software market, while we were an un-important player.
142 / LOUIS V. GERSTNER, JR.
In the space between: products like databases, systems management software, and transaction management programs. It was the complex, largely invisible layer (aptly named “middleware”) about which only the most hardcore techies could get excited.
Yet the more we considered what was going to matter if client/server computing gave way to networked computing, middleware started to look less like a backwater and more like the key strategic battleground. We couldn’t see the entire picture at the time, but we could see enough. More users. More devices. More transactions. And more demand for ways to integrate applications, processes, systems, users, and institutions. No operating system was going to be able to tie it all together. But middleware existed to do exactly that.
To provide this kind of integration, however, middleware was going to have to work on all of the major vendors’ computer systems that would be linked together over vast new networks. In the industry’s jargon, the new middleware would have to work “cross platform,” and this represented a major obstacle that we would have to overcome. Up until this point in 1995, all of IBM’s software was proprietary and worked only with IBM hardware and other IBM
software.
Thus, we launched a massive, multi-year effort to rewrite all of our critical software, not only to be network-enabled, but to run on Sun, HP, Microsoft, and other platforms. It was a hugely expensive and complicated project and it created many of the same internal tensions that our services strategy had evoked. Once again we were collaborating with the enemy!
Acquiring Lotus
In early 1995 John came to me with a bold idea to accomplish two objectives: fill a hole in our middleware portfolio, and plant a flag firmly in the world of collaborative, rather than stand-alone, computing. His idea: to acquire Lotus Development Corporation. He has told
WHO SAYS ELEPHANTS CAN’T DANCE? / 143
me since then that he did it with great trepidation, because it was very much out of character for IBM to “buy” technology rather than build it. Besides that, John and I really didn’t know each other very well, and he was asking me to sign a very big check.
Lotus had made its name with its popular 1-2-3 spreadsheet software. But what we wanted most, the crown jewel, was an elegant product called Notes—pioneering software that supported collaboration between large numbers of computer users.
By May John convinced me that IBM should acquire Lotus. Thus began the largest software acquisition in the history of the industry.
John had approached Lotus CEO Jim Manzi several times about a possible deal, but he had been turned down. We decided to launch an unsolicited bid. I called Manzi on June 5 to inform him that we were initiating a hostile takeover. I can’t be sure, but I do not believe he was surprised. His reply was noncommittal, cold but polite—and very brief.
Anyone will tell you that software acquisitions are risky. The asset you’re acquiring is human. If the critical people decide to walk (and a lot of them would certainly have the financial wherewithal to do that once the deal closed), then you’ve spent a lot of money for some buildings, office equipment, and access to a customer-installed base.
That part of the deal didn’t scare me. At American Express we’d acquired First Data Resources, which had a very distinct culture, was privately owned, and didn’t want any part of being assimilated into a great big company. I knew it could be managed. But in the case of a maverick software company, I also had a feeling that the effort to win over the workforce was a battle that had to be won before the first shot was fired. We understood that our every move was going to be scrutinized by the Lotus employees, whose trust we desperately needed to win.
We knew that the Lotus workforce—much more than IBM’s at the time—was an Internet-centric culture. We mixed the Internet and the IBM home page for fast, unfiltered delivery of our position di 144 / LOUIS V. GERSTNER, JR.
rect to Lotus employees and shareholders. One minute after I’d ended the call to Manzi, our message, including the letter we’d faxed to Lotus, was live on the Net. As expected, the Lotus workforce knew where to look. They came and they read. We had crossed the first hurdle in the merger of two diametrically opposed corporate cultures.
Privately we were still afraid the deal could take months to complete. Hostile takeovers often become weighty with drama, complete with white knights, court battles, and golden parachutes. But our offer made good strategic sense for everyone, and I think we took all the right precautions. On Tuesday Manzi called me and we talked over dinner in Manhattan that night. Our two companies met in small groups over the next two days to talk about culture, legal issues, employees, and pricing. In a law office on Friday night, we shook hands over the final price: $3.2 billion. By Sunday the boards on both sides had approved the deal. In one week we had wrapped up the biggest software deal in history.
This was also the first hostile takeover enacted by IBM—as well as something of a novelty in the business world.
The New York Times
said: “Perhaps the most striking aspect of IBM’s takeover bid, and the one that says the most about these times, is that it defies the accepted wisdom on the difficulties of trying to acquire a company whose primary value isn’t in its machinery or real estate but rather, in that most mercurial of assets, people.” Fortunately, we were able to keep all of the key people, including Ray Ozzie, the development genius behind Notes. (Jim Manzi stayed on for a few months, but he wasn’t the kind of person who was comfortable in a large, complex enterprise.)
There were approximately 2 million Notes “seats” installed in customer locations when the deal closed, in July 1995. That grew to 90 million at the end of 2001. And while the Internet has subsequently obviated much of the need for basic collaborative software, Lotus remains in the sweet spot of the world of knowledge management and collaboration.
In the end we gained more than a software company. Culturally WHO SAYS ELEPHANTS CAN’T DANCE? / 145
we proved that we could keep some organizational distance and allow a fast-moving team to thrive. Perhaps most important, the hostile acquisition sent a clear signal inside and outside IBM that we were out of survival-mode status and serious about reclaiming a position of influence in the industry.
About nine months later, and again at the urging of John Thompson, we made another big software acquisition—Tivoli Systems—that leapfrogged us into the market for distributed systems management products (more gorpy, but very critical, middleware).
Tivoli was a $50 million company when we bought it. Its revenues, augmented by some business we transferred from IBM, are now in excess of $1 billion.
As I write this, IBM’s Software Group is one of the most powerful software companies in the world and is positioned as the leading software company in networked computing. With 2001 revenues of $13 billion (second only to Microsoft) and pre-tax profits of about $3 billion (growing at a double-digit rate), we are number one or number two in every segment of the market in which we participate.
The IBM software story is a wonderful microcosm of the overall turnaround that took place at the company over the past decade.
Incredible technical resources were unleashed to deploy across the entire industry. The catalyst to drive this transformation was an external event—the arrival of the Internet. Spurred on by this emerging opportunity, we rapidly restructured our assets and organization, and poured resources into acquisitions and development strategies.
S
o far the logic underlying our big strategic bets has been fairly straightforward. If you’re going to be a company that designs, builds, and delivers integrated technology solutions, you need a services capability. If you already develop and sell more software than any other company does, and if you believe software is going to be the connective tissue of the networked world, you ought to run your software business as, well, a business.