Who Says Elephants Can't Dance?: Leading a Great Enterprise through Dramatic Change

by Louis V. Gerstner, Jr.

  This doesn’t mean it wasn’t a good transaction for AT&T. It allowed AT&T to leapfrog its competitors. But for IBM it was a strategic coup.

  We got out of a business whose value was going to deteriorate very quickly, as massive capacity was added around the globe. We avoided the huge capital investment to maintain the network. And we exited from another part of the stack that was not strategically vital.

  To say there was heavy resistance inside parts of IBM understates the point. People argued, passionately, that we were shortchanging our future. They simply couldn’t see the logic in jettisoning a global

  160 / LOUIS V. GERSTNER, JR.

  data network when we all believed we were on the brink of a networked world. Once again there was the “Do it all to be the best”

  argument. And once again we opted for focus over breadth.

  The PC Dilemma

  Perhaps the most difficult part of the business that needed to be overhauled during my tenure at IBM was the PC segment of our portfolio. Over the course of nearly fifteen years, IBM had made little or no money from PCs. We sold tens of billions of dollars’ worth of PCs during that time. We’d won awards for technical achievement and ergonomic design (especially in our ThinkPad line of mobile computers). But at the end of the day it had been a relatively unprofitable activity. There were times when we lost money on every PC

  we sold, and so we were conflicted—if sales were down, was that bad news or good news?

  The single most important factor in our overall performance was that Intel and Microsoft controlled the key hardware and software architectures and were able to price accordingly. However, we weren’t innocent bystanders as they had achieved those dominant positions. We had entered the business in the 1980s with a lack of enthusiasm for the product, as I’ve already noted. We had consistently underestimated the size and importance of the PC market. We had never developed a sustained leadership position in distribution, vacillating between company-owned stores at one time, to dealers, to distributors, to telephone sales systems. Finally, we couldn’t manufacture PCs in a world-class manner in respect to cost and speed to market.

  Despite this unacceptable performance, we were never prepared to get out. There were many reasons for this, some more applicable in the early 1990s than they are today. But suffice it to say that, at that time, the PC represented a lot of revenue and critical customer mind-share. In very real ways, a company’s PC drove the company’s image in

  WHO SAYS ELEPHANTS CAN’T DANCE? / 161

  the industry. There were raging internal debates about this, but ultimately we felt we couldn’t abandon the PC completely and still be the integrator we needed to be for our customers.

  So we adopted a strategy of playing to our strengths, primarily in mobile computing and in the market for systems that connected other PCs and helped them function in an integrated way. We waited too long to do it, but we finally abandoned the more commodity-like segments, ceasing to sell to consumers through retail stores and shifting more of our consumer business to direct channels such as ibm.com and telesales. Later on we turned over the development and manufacturing of most of our PCs to third parties, lowering our exposure to this segment even further. Still, it’s a spotty record at best, and I am not terribly proud of it.

  There were many other steps taken to withdraw from parts of the stack and focus our portfolio. We exited network hardware. Even though we had invented this business, we simply failed to exploit it over the subsequent fifteen or twenty years. We exited the DRAM

  business. As mentioned earlier, it is a commodity-based, notoriously cyclical market. Midway through 2002 we agreed to divest our hard-disk-drive business through an agreement with Hitachi. As I write these words, other candidates are under active review. This process of selecting markets and competing on the basis of a distinctive, sustainable competency is essential to the new IBM, and I know it will be an ongoing challenge.

  Fallacies and Myths and Lessons

  As that work proceeds, it is my hope that the company’s new leaders keep sight of some of the higher-level lessons that resulted from these decisions.

  162 / LOUIS V. GERSTNER, JR.

  With os/2-the fallacy that the best technology always wins.

  I can understand why this one, in particular, was hard for IBM to accept. In the early days of the computer industry, systems failed frequently and the winner was usually the one with the best technology. So we came to the OS/2 v. Windows conformation with a product that was technically superior and a cultural inability to understand why we were getting flogged in the marketplace.

  First, the buyers were individual consumers, not senior technology officers. Consumers didn’t care much about advanced, but arcane, technical capability. They wanted a PC that was easy to use, with a lot of handy applications. And, as with any consumer product—from automobiles to bubble gum to credit cards or cookies—marketing and merchandising mattered.

  Second, Microsoft had all the software developers locked up, so all the best applications ran on Windows. Microsoft’s terms and conditions with the PC manufacturers made it impossible for them to do anything but deliver Windows—ready to go, preloaded on every PC they sold. (Even IBM’s own PCs came preloaded with OS/2

  and Windows!) And in the mid-1980s, the Windows marketing and PR machine alone had more people than IBM had working with software partners or distributors. Our wonderful technology was whipped by a product that was merely okay, but supported by a company that truly understood what the customer wanted. For a

  “solutions” company like IBM, it was a bitter but vital lesson.

  In the case of application software-the myth of “account control.”

  This was a term used by IBM and others to talk about how a company maintained its hold on customers and their wallets. It suggests that once customers buy something from a company, then train their people on that product and get familiar with how to support it, it’s very hard for them to move to a competitor.

  WHO SAYS ELEPHANTS CAN’T DANCE? / 163

  As a former customer, I was always offended and indignant that information technology companies talked about controlling customers. I had this quaint view that it was the job of a supplier to serve customers, not control them!

  What IBM has learned from this part of the “unstacking” is that we can be the premier integrator and, at the same time, partner with many other companies in delivering an integrated solution. In fact, wearing my customer hat, I could argue that the role of IT partner, or integrator, cannot be fulfilled by companies that support only one technology or one stack. Beware, customers, of suppliers who provide only UNIX or Wintel answers to your problems. Beware of totally proprietary vendors who fight new developments like Linux.

  These vendors still view the world through the window of their proprietary stack.

  At IBM we now focus on a different stack: the customer’s business processes and how we can bring world-class technology—both our own and that provided by other leading companies—to improve those processes.

  In the case of PCS, there are still unanswered questions.

  Why did we make the decision to exit the application software, network hardware, DRAMs, or the data transmission businesses, but not PCs? Why did we decide to stay in the hardware end of this business, even as we folded our hand on OS/2? In hindsight, was this the right decision? I think it was at the time, but the decision has been painful and costly for IBM.

  If there is one lesson to be extracted from this saga, I think it’s about staying true to one’s strategic vision. I said in 1993 that the marketplace would drive every important decision we at IBM made.

  But when it came to the PC business, we weren’t paying attention to either our customers or our competitors.

  One competitor in the PC industry was proving that customers were perfectly willing to buy direct—over the phone, or later via a

  164 / LOUIS V. GERSTNER, JR.

  Web site. But we were painfully slow
to move away from our existing distribution channels. Why? The incomplete and unsatisfying answer at the time was that we’d always done it that way.

  I’m not saying the shift to lower-cost, direct channels wouldn’t have involved some pain, and I’m not saying that actually sticking to a strategic vision is as easy as articulating it. The tendency, especially in a hyper-competitive marketplace, is to establish a position, hunker down, and defend it. But if we had focused on the marketplace and done our homework, there’s no reason the IBM PC business today would be looking up the leader board at Dell.

  Opening up our stack (and our minds) to others had many positive effects on IBM. It cut our losses and improved our integrated offerings to customers. And it freed up resources to invest in the future. Huge sums of money and huge quantities of brainpower have been redeployed from wall-banging futility to exciting new work in areas such as storage systems, self-directing computers, bioinformatics, and nanotechnology.

  It is all about focus—a subject I will return to later.

  18

  The Emergence

  of e-business

  Y ou’ll recall that earlier I said I’ve gotten lucky twice. I described the first piece of luck as my initial encounter with Dennie Welsh, the executive who shared my vision for transforming IBM into a services-led company. Believe it or not, Dennie also played an integral role in my second bit of luck (more on that in a moment).

  Long before my arrival at IBM, one of the most widely discussed technology trends in business centered on what was called “convergence”—the melding of telecommunications, computing, and consumer electronics; or, stated differently, the merger of traditional analog technologies with their emerging digital kin. Depending on one’s point of view, this either promised or threatened to transform multiple industries.

  The subject was not foreign to me. In February 1983, in a speech at the University of Virginia, I made the following observation:

  “Computer and telecommunications technologies give us a reach and flexibility that were beyond imagination just a few years ago.…Technology has virtually eliminated distance as an obstacle to doing

  166 / LOUIS V. GERSTNER, JR.

  business. Today, an American Express Cardmember from Dallas can buy a plane ticket in Kuala Lumpur and have the purchase authorized in less than six seconds by our computer system in Phoenix, Arizona.”

  During my brief tenure on the Board of Directors of AT&T, I had learned a great deal about the allure of convergence. That company had bought a computer company, NCR. Some years before that, IBM

  had bought a telecommunications equipment company, ROLM. Both had placed bets on convergence. And they weren’t alone.

  If you were a telephone company back then, you were salivating at the prospect of getting beyond dial tone and voice services, to provide all kinds of higher-value services—data, entertainment, and commerce—to homes and businesses.

  If you were in the entertainment or media business, convergence represented the ultimate distribution channel. Not only would you be able to digitize all your content, you’d be able to deliver it to devices from personal computers and smart TVs to cell phones and network-enabled wristwatches!

  Consumer electronics companies were dreaming up a panoply of devices that would allow billions of people to access this world of digital information and entertainment.

  And the information technology industry was gearing up for an explosion of demand for the hardware and software that would manage, process, and store the world’s digital content.

  So, as I started to probe the strategic thinking inside IBM in 1993, I wasn’t surprised to find a number of people who were very excited about this issue. Which brings me back to straight-shooting Dennie Welsh.

  Discovering the Cloud

  In August 1992 Dennie had landed IBM’s largest single contract ever, the outsourcing of all of Sears’ data center operations. As part of

  WHO SAYS ELEPHANTS CAN’T DANCE? / 167

  that contract IBM and Sears merged their private data networks into a joint-venture company called Advantis. It was classic Dennie. He’d been battling IBM to make our global network part of his services organization, but he was getting nowhere with colleagues who couldn’t quite see his vision for making the network a big-time profit center. In one stroke he not only closed an $8 billion contract with Sears, but he gained control of more network capacity.

  Eventually the joint venture with Sears was dissolved, we took total ownership, and Advantis became part of what we called the IBM Global Network. In its day, IGN was one of the world’s most sophisticated networks. It was also the largest Internet service provider in the world—truly an asset without equal, a moneymaker and our stake-in-the-ground in the networking business.

  Years later, and as I described in the previous chapter, we divested ourselves of the Global Network in a $5 billion transaction with AT&T. What I didn’t mention was that as early as 1993 we knew we’d eventually sell this business. In one of the first conversations I had with Dennie, we agreed that in the long term we’d never be able to justify the massive capital investments required to compete with the telcos. With their assets and base of capital equipment, they could easily undercut our prices. What we couldn’t see back then was the Internet, which would entirely obviate the need for us to own a network.

  It had to be in one of these early discussions with Dennie that I was introduced to “the cloud”—a graphic much loved and used on IBM charts showing how networks were going to change computing, communications, and all manner of business and human interaction.

  The cloud would be shown in the middle. To one side there would be little icons representing people using PCs, cell phones, and other kinds of network-connected devices. On the other side of the cloud were businesses, governments, universities, and institutions also connected to the network. The idea was that the cloud—the network—would enable and support incredible amounts of communications and transactions among people and businesses and institutions.

  168 / LOUIS V. GERSTNER, JR.

  If the strategists were right, and the cloud really did become the locus of all this interaction, it would cause two revolutions—one in computing and one in business.

  It would change computing because it would shift the workloads from PCs and other so-called client devices to larger enterprise systems inside companies and to the cloud—the network—itself. This would reverse the trend that had made the PC the center of innovation and investment—with all the obvious implications for IT companies that had made their fortunes on PC technologies.

  Far more important, the massive, global connectivity that the cloud depicted would create a revolution in the interactions among millions of businesses, schools, governments, and consumers. It would change commerce, education, health care, government services, and on and on. It would cause the biggest wave of business transformation since the introduction of digital data processing in the 1960s.

  So it was natural that when I decided to put someone in charge of a team that would investigate whether we truly believed that convergence was the future—and if so, what to do about it—Dennie got the call. He was passionate on the subject, by then he “owned”

  our network, and yet I knew Dennie would make sure the team was objective. The team delivered its answer and a set of detailed recommendations in three months.

  A Network Computing Blueprint

  Dennie’s group believed zealously—from the standpoints of technical feasibility and the business opportunity—that this was where the industry was headed. But again, what they presented was not primarily an Internet strategy. That’s not surprising, because this was back in the days when few people outside of universities and government labs had heard of the Internet. Fewer still believed that the Internet could be a mass-market communications medium, much less

  WHO SAYS ELEPHANTS CAN’T DANCE? / 169

  a platform for mainstream business transactions. (One notable exception inside IBM was a m
arketing executive named John Patrick, who had a unique ability to bring the networked world out of the jargon of technology and into the minds of everyday people. John became our spokesperson, demonstrating to our customers and to IBM employees how real-life activities would benefit from the Net.) Whether we were thinking of networks in general or the Internet specifically, the single most important outcome of Dennie’s task force was a recommendation to commit the resources of the entire company to lead this next wave of computing—to mobilize on every front.

  On its face, and given the success we’d enjoyed based on proprietary architectures like the System/360, this could have presented a staggering barrier to success. But in truth this one wasn’t as tough for us as it would be for some of our mainstream competitors, because we’d long since made our commitment to open, standards-based computing, including embracing all the important Internet standards and protocols. Still, there was plenty to do, much of it described previously.

  In software we were fortunate to recognize that middleware would be the integrating glue of networked applications. We had to step up the Internet enablement of these products and develop some new ones.

  We’d have to build a significant new services business around what came to be known as Web hosting. And because the networked world was about helping customers transform their businesses, we had to build capability in consulting and implementation services associated with e-business.

  In component technology, what began in the early 1990s as a search for a new revenue stream turned into the foundry of specialized chips that would be in high demand.