When knowing isn't doing: Strategic delay and the cost of inertia

The story of IBM and the personal computer is not, as commonly assumed, a tale of a company that failed to enter a market. IBM launched the IBM 5150 in August 1981, just a few years after the Apple II. The machine was a runaway success, exceeding sales expectations by over 800% in its first year.

The real story is more instructive for leaders: IBM knew what it needed to do at critical junctures, but delayed doing so. The company underestimated the PC's strategic importance, failed to control the platform's key components, and responded too slowly when competitors exploited these gaps. By the time IBM attempted to reassert control in 1987, the market had moved on without them.

This case study examines IBM's PC strategy through the lens of decision-making and personal leadership. It demonstrates how even the most successful organizations can know the right course of action and still fail to execute, how institutional culture can override strategic clarity, and how the gap between knowing and doing can transform market leaders into also-rans.

Part I: The Landscape

IBM in 1980: The Unchallenged Giant

By 1980, IBM dominated the computer industry. As James W. Cortada documents in IBM: The Rise and Fall and Reinvention of a Global Icon, "For decades, IBM shaped the way the world did business. IBM products were in every large organization, and IBM corporate culture established a management style that was imitated by companies around the globe. It was 'Big Blue,' an icon."

The company's strength was in mainframes and minicomputers, high-margin machines sold to large enterprises. The corporate culture that had built this dominance was legendary. IBM employees were known for their white shirts, conservative dress, and unwavering commitment to customer service. The saying in corporate America was simple: "No one ever got fired for buying IBM."

But a new market was emerging. Apple had introduced the Apple II in 1977. Commodore and Tandy were selling machines to hobbyists and small businesses. The personal computer market, though tiny compared to IBM's mainframe business, was growing rapidly.

IBM's leadership saw this. They understood the trend. The question was what to do about it.

The Decision to Enter

In 1980, William Lowe, director of IBM's General Systems Division lab in Boca Raton, Florida, pitched IBM CEO Frank Cary on the idea of a computer for small businesses and consumers that would sell for around $1,500. Cary gave Lowe an aggressive timeline: a month to develop a prototype and a year to get a product to market.

This timeline forced a critical decision. IBM could not build the PC the way it built mainframes. There wasn't time to develop proprietary components, and the economics didn't justify the investment. As one industry analysis noted, Lowe's plan "called for buying existing components and software and bolting them together into a package aimed at the consumer market. There would be no homegrown operating system or IBM-made chips."

Don Estridge, who took over the project after Lowe's promotion, assembled a small team in Boca Raton. They operated as a "skunkworks," isolated from IBM's standard procedures and bureaucracy. The team, nicknamed the "Dirty Dozen," made decisions quickly and built a machine from off-the-shelf components: an Intel 8088 processor, commodity parts, and an operating system licensed from a small company called Microsoft.

The IBM PC launched on August 12, 1981. It was an immediate success. IBM sold 200,000 units in the first year, then 200,000 units per month. By 1984, the company had 75% of the personal computer market.

From the outside, this looked like flawless execution. From the inside, the seeds of IBM's decline were already planted.

Part II: Three Strategic Delays

Delay One: Underestimating Strategic Importance

The IBM PC was never treated as central to the company's future. Cortada's analysis reveals that IBM's leadership "still saw mainframes/minis as the real business." The PC was a side project, organizationally and psychologically separated from IBM's core operations.

This wasn't ignorance. IBM's leaders could see the PC market growing. They had the data. They understood the technology trends. But the mainframe business was enormously profitable, and the PC business, even at scale, would never match those margins. As one Amazon reviewer of Cortada's book observed about IBM's leadership: "Poor leadership helped to give away the PC business... over focus on the mainframe for years after it was a dinosaur."

The Boca Raton operation that created the PC was deliberately kept small and autonomous. When the PC succeeded beyond all expectations, this became a problem. On August 1, 1983, Estridge's skunkworks was redesignated the Entry Systems Division, which meant the PC business was absorbed into IBM's corporate bureaucracy. Estridge's 4,000-person group mushroomed to 10,000. He protested that corporate had transferred thousands of programmers who "knew nothing about PCs."

The very qualities that had made the PC possible (speed, autonomy, and willingness to use outside components) were systematically eliminated once the product succeeded. IBM's institutional immune system rejected the transplant.

Delay Two: Failing to Control the Platform

The most consequential decisions IBM made were about what it chose not to own.

Microsoft had been brought in to provide the operating system. Bill Gates and Microsoft didn't actually have an operating system when IBM came calling. Gates quickly licensed a product called QDOS ("Quick and Dirty Operating System") from Seattle Computer Products for $75,000, adapted it for the IBM PC, and delivered it as PC-DOS.

Here is where IBM made its critical error: the company allowed Microsoft to retain the rights to license the same operating system to other manufacturers. IBM paid a flat fee rather than taking an exclusive license. As one analysis put it, "IBM's strategic error in not retaining rights to the operating system went far beyond [the immediate financial impact]; it meant that Microsoft would set the standards for the PC operating system."

IBM also chose an open architecture with commodity parts. Anyone could see exactly how the machine was built and replicate it, except for the BIOS (Basic Input/Output System), which was IBM's proprietary code.

Within a year, competitors were working to clone the BIOS. Columbia Data Products produced the first IBM PC clone in June 1982. Compaq followed in November 1982 with the Compaq Portable, featuring a legally reverse-engineered BIOS created using "clean room" techniques. The company spent $1 million to ensure their BIOS was functionally equivalent to IBM's without copying any code.

The clean room approach worked. IBM's legal challenges failed. And once Phoenix Technologies began selling a clone BIOS to anyone in 1984, the floodgates opened. Any manufacturer could now build a fully IBM-compatible PC without paying IBM a cent.

IBM knew about the clone threat almost immediately. The company understood that its open architecture was being exploited. Yet it delayed taking decisive action to close the platform or develop proprietary advantages. The mainframe mindset, in which IBM controlled everything and customers came to IBM, prevented leaders from recognizing that the PC market operated by different rules.

Delay Three: The PS/2 Comes Too Late

By the mid-1980s, the damage was becoming clear. IBM's market share plummeted from 76% in 1983 to 26% in 1986. Clone manufacturers were eating IBM's lunch with machines that were cheaper, often faster, and fully compatible with IBM's software ecosystem.

In 1987, six years after the PC's launch and five years after clones began flooding the market, IBM finally attempted to reassert control. The company launched the Personal System/2 (PS/2) line with a proprietary bus architecture called Micro Channel Architecture (MCA). The strategy was explicit: create a new standard that would require clone makers to pay licensing fees to maintain compatibility.

The market rejected it.

MCA was technically superior to the Industry Standard Architecture (ISA) bus used in existing PCs. But the clone manufacturers had no intention of surrendering their independence. Led by Compaq, nine major PC makers formed the "Gang of Nine" and announced a competing standard called Extended Industry Standard Architecture (EISA).

Customers who had invested heavily in ISA expansion cards and peripherals saw no compelling reason to abandon their existing hardware for IBM's new, proprietary system. As one analysis noted, "With what was widely seen as a technically competent but cynical attempt to gain undisputed control of the market, IBM unleashed an industry backlash."

The PS/2 sold reasonably well initially, moving about 3 million units in its first two years. But it failed to recapture the market. IBM's share continued to decline. By 1994, Compaq had replaced IBM as the number one PC vendor in the United States.

IBM had waited too long. The trigger that might have worked in 1982 or 1983 misfired in 1987. The market had evolved past the point where IBM could dictate terms.

Part III: The Consequences

The Near-Collapse

The PC debacle was part of a larger institutional failure at IBM. By the early 1990s, the company was in crisis. Its stock price fell from $43 in 1987 to $13 by the early 1990s. The company lost $8 billion in 1993 alone, marking a three-year loss of nearly $16 billion.

IBM, the company that had shaped global business for decades, was on the verge of bankruptcy. Bill Gates predicted the company would fold within seven years.

Cortada argues in his history that IBM had developed a corporate "sclerosis," an over-bureaucratization that impeded its ability to adapt. A generation of senior executives was out of touch and unprepared to respond to new market demands.

The board eventually hired Lou Gerstner, a CEO with no technology background, who had previously run American Express and RJR Nabisco. Gerstner kept IBM intact instead of breaking it apart as many advised, laid off nearly 100,000 employees, and shifted the company's focus from hardware to services and software.

IBM survived. But it eventually sold its PC division to Lenovo in 2005 for $1.75 billion, an inglorious exit from a market it had once dominated.

Where the Value Went

The value that IBM failed to capture didn't disappear. It simply went elsewhere.

Microsoft, with its non-exclusive DOS license, became one of the most valuable companies in the world. Every clone manufacturer needed an operating system, and Microsoft was happy to sell them one. By the early 1990s, Microsoft had leveraged that operating system dominance into Windows and Office, building a software empire worth tens of billions of dollars.

Intel, which supplied the processors, captured another piece. The "Wintel" duopoly (Windows plus Intel) defined the PC industry for decades, with IBM relegated to one manufacturer among many.

The clone manufacturers, Compaq, Dell, Gateway, and eventually Asian manufacturers, captured the hardware business by competing on price, features, and speed to market. These companies had no legacy mainframe business to protect, no institutional culture telling them that PCs were a side project.

IBM had created the standard, validated the market, and then watched others capture most of the value.

Part IV: Lessons for Leaders

The Gap Between Knowing and Doing

IBM's story is not about ignorance. The company had some of the smartest people in technology. They understood the PC market. They saw the clone threat. They knew, eventually, that their open architecture was a strategic vulnerability.

They knew all of this and still delayed action for years.

This is the gap between knowing and doing that defines personal leadership failures. It's not that leaders lack information. It's that something prevents them from acting on what they know.

For IBM, that something was institutional culture. The company had been successful for so long with one model (proprietary systems, high margins, enterprise customers) that it couldn't fully commit to a different model even when the evidence demanded it. The PC division was always treated as peripheral to the "real" business, even as it was transforming the entire industry.

When the Trigger Gets Heavier

IBM had multiple opportunities to pull the trigger on a different strategy. In 1980, they could have developed a proprietary operating system. In 1981, they could have negotiated exclusive rights to DOS. In 1982, they could have moved aggressively against clones. In 1983, they could have committed to the PC as the company's strategic future.

Each year they delayed, the trigger got heavier. The clone ecosystem grew. Microsoft's installed base expanded. Customers invested in ISA hardware. By 1987, when IBM finally attempted to reassert control with the PS/2, it was too late. The market had reached a tipping point where IBM no longer had the power to dictate terms.

This is how procrastination works in strategic decisions. Early action is relatively easy; the forces arrayed against you are still weak. But delay allows those forces to strengthen. What would have been a manageable challenge becomes an impossible one.

The Illusion of Time

IBM's leaders likely believed they had time. The mainframe business was still profitable. The PC was growing but hadn't yet overtaken traditional computing. The clone threat was concerning but not yet existential.

This is the most dangerous form of the knowing-doing gap: the situation where the right action is clear but doesn't feel urgent. Leaders tell themselves they'll address it after the current quarter, after the next product launch, after the board meeting. Month after month. Year after year.

The market doesn't wait for leaders to feel ready. IBM's competitors weren't constrained by IBM's internal debates about strategic priorities. They simply built products, captured customers, and established facts on the ground.

Values Revealed by Decisions

What did IBM's decisions reveal about its actual values, as opposed to its stated ones?

The company said it valued innovation. But it treated its most innovative division, the Boca Raton skunkworks, as an aberration to be normalized once the immediate pressure was off.

The company said it valued customers. But it tried to force customers onto a new proprietary standard (MCA) that served IBM's interests rather than customer needs.

The company said it wanted to win in personal computing. But it never committed the resources, attention, or organizational priority that would have been required.

Actions reveal values. IBM's actions revealed that it valued its existing business model, corporate culture, and short-term mainframe profits more than it valued the strategic transformation the PC market demanded.

The Decision That Wasn't Made

IBM's PC story is ultimately about a decision that wasn't made: the decision to treat the personal computer as the company's strategic future and to organize everything around winning that market.

This wasn't a decision that required a single dramatic moment. It required sustained commitment, organizational change, and the willingness to cannibalize existing businesses. It required IBM's leadership to see beyond the current quarter's mainframe profits to the long-term trajectory of computing.

They knew what that trajectory looked like. The data was clear. The trends were visible. The competitive threats were obvious.

They knew and still didn't act. Not because they were stupid, evil, or incompetent, but because they were human. Because institutional inertia is powerful. Because it's easier to optimize the existing business than to transform it. Because the mainframe profits were real today, and the PC opportunity was uncertain tomorrow.

This is why personal leadership matters. Organizations don't make decisions; leaders do. And leaders are subject to all the predictable irrationalities that Dan Ariely documents: the tendency to overvalue the present, to avoid loss, to anchor on past success, to delay difficult choices.

The gap between knowing and doing isn't closed by better information or smarter analysis. It's closed by leaders who recognize their own patterns, design systems that account for institutional inertia, and make the difficult calls before the trigger becomes too heavy to pull.

IBM eventually found such a leader in Lou Gerstner, who arrived in 1993 after the company had nearly collapsed. Gerstner pulled the trigger on dramatic changes, including laying off 100,000 employees and fundamentally transforming IBM's business model.

But by then, the PC market was already lost. The changes that could have saved IBM's dominance in personal computing would have been much less painful if made in 1982 or 1983. Instead, IBM waited too long, and by the time they acted, only painful options remained.

The lesson for leaders is not complicated. It's just hard.

When you know what needs to be done, do it. The gap between knowing and doing is where companies go to die.

Sources

Cortada, James W. IBM: The Rise and Fall and Reinvention of a Global Icon. MIT Press, 2019.

Gerstner, Louis V. Who Says Elephants Can't Dance? Inside IBM's Historic Turnaround. Harper Business, 2002.

"How the IBM PC Won, Then Lost, the Personal Computer Market." IEEE Spectrum, March 2023.

"The IBM PS/2: 25 Years of PC History." PCWorld, April 2012.

Computer History Museum. "Send in the Clones: Capitalizing on IBM's Success." CHM Revolution.

IBM Corporate Archives. "The IBM PC" and "The PS/2." IBM History.

Notes on Sourcing Methodology

1. Primary reliance on Cortada: As the most comprehensive scholarly treatment of IBM's history, written by a 38-year IBM veteran, Cortada's work serves as the authoritative backbone for historical claims about IBM's culture, decision-making, and strategic failures.
2. Cross-verification: Key facts (dates, market share figures, product launches) were verified across multiple sources, including IEEE Spectrum, PCWorld, Computer History Museum, and Wikipedia.
3. Contemporary sources: Where possible, claims about IBM's strategic thinking are supported by contemporary accounts and insider perspectives rather than retrospective analysis alone.
4. Quantitative claims: All numerical data (market share percentages, sales figures, financial losses) are cited to specific sources and cross-referenced where possible.