Why CEOs Fail It's rarely for lack of smarts or vision. Most unsuccessful CEOs stumble because of one simple, fatal shortcoming.
By Ram Charan and Geoffrey Colvin

(FORTUNE Magazine) – What got Eckhard Pfeiffer fired? What fault did in Bob Allen? Or Gil Amelio, Bob Stempel, John Akers, or any of the dozens of other chief executives who took public pratfalls in this unforgiving decade? Suppose what brought down all these powerful and undeniably talented executives was just one common failing? It's an intriguing question and one of deep importance not just to CEOs and their boards, but also to investors, customers, suppliers, alliance partners, employees, and the many others who suffer when the top man stumbles. The answer even matters to the country; America is the world's most competitive nation, thanks in large part to the overall high quality of its CEOs. If people knew how to spot CEOs headed for failure--even if the company's results still looked fine--they could save themselves much pain. Trouble is, they usually look in the wrong place.

Consider the Pfeiffer episode. The pundits opined, as they usually do in these cases, that his problem was with grand-scale vision and strategy. Compaq's board removed the CEO for lack of "an Internet vision," said USA Today. Yep, agreed the New York Times, Pfeiffer had to go because of "a strategy that appeared to pull the company in opposite directions."

But was flawed strategy really Pfeiffer's sin? Not according to the man who led the coup, Compaq Chairman Benjamin Rosen. "The change [will not be in] our fundamental strategy--we think that strategy is sound--but in execution," Rosen said. "Our plans are to speed up decision-making and make the company more efficient."

You'd never guess it from reading the papers or talking to your broker or studying most business books, but what's true at Compaq is true at most companies where the CEO fails. In the majority of cases--we estimate 70%--the real problem isn't the high-concept boners the boffins love to talk about.

It's bad execution. As simple as that: not getting things done, being indecisive, not delivering on commitments. We base our conclusions on careful study of several dozen CEO failures we've observed over the decades--through our respective work as a consultant to major corporations and a journalist covering them. The results are beyond doubt.

Here's what we aren't saying: That failed CEOs are dumb or evil. In fact they tend to be highly intelligent, articulate, dedicated, and accomplished. They worked hard, made sacrifices, and may have performed terrifically for years; Pfeiffer, for example, transformed the company more than once and multiplied Compaq's revenues, profits, and market values, a remarkable achievement. And failure as a CEO is never final. These are strong people who can come back successfully in other roles.

Nor are we saying execution is the only reason CEOs falter. Sometimes they adopt a strategy so flawed that it's doomed, or they refuse to confront reality in their markets, or they antagonize their board. And when a CEO really goes down in flames, there's almost always more than one reason. But business people learn to focus on the main thing, the explanation that accounts for most of what they're worried about, and in the realm of CEO failures that explanation is clear.

It's clear, as well, that getting execution right will only become more crucial. The worldwide revolution of free markets, open economies, and lowered trade barriers and the advent of e-commerce has made virtually every business far more brutally competitive. The frantic spread of information through technology is making customers everywhere more powerful and pushing toward the commoditization of everything. Institutional investors now own more than half the equities in U.S. corporations and relentlessly demand results. Indeed, two of the nation's preeminent headhunters, Tom Neff and Dayton Ogden of Spencer Stuart, calculated recently that while average CEO tenure in the biggest companies has remained fairly steady at seven to eight years, those who don't deliver are getting pushed out quicker. (See the graph later in the article.) A new academic study reaches the same conclusion--poorly performing CEOs are three times more likely to get booted than they were a generation ago. Even if their boards spare them, their companies often get taken over, like Digital Equipment under Robert Palmer and Rubbermaid under Wolfgang Schmitt. Bottom line: whatever cover CEOs used to hide behind has been blasted away. Either they deliver, soon, or they're gone.

So how do CEOs blow it? More than any other way, by failure to put the right people in the right jobs--and the related failure to fix people problems in time. Specifically, failed CEOs are often unable to deal with a few key subordinates whose sustained poor performance deeply harms the company. What is striking, as many CEOs told us, is that they usually know there's a problem; their inner voice is telling them, but they suppress it. Those around the CEO often recognize the problem first, but he isn't seeking information from multiple sources. As one CEO says, "It was staring me in the face, but I refused to see it." The failure is one of emotional strength.

The excuses and rationalizations that CEOs concoct are largely unconscious, a mechanism for avoidance. They make an impressive list; six cover most cases:

"He has to succeed." The CEO may become a victim of "intellectual seduction," installing a subordinate so talented that the CEO persuades himself failure is impossible. If the protege then fails to deliver, the CEO can't come to terms with it, especially if the protege is a succession candidate. Often these subordinates have been promoted into line jobs from staff positions or consulting firms, with their high-level executional abilities untested.

"He's my guy!" The problem of blind loyalty shows up more often than you may suspect. The boss and the subordinate may have worked together a long time; in some cases their families vacationed together. Judgment becomes blurred. Mention this to people who were around General Motors in the early '90s and they tend to nod vigorously and say, "Lloyd Reuss!" He became president when Robert Stempel became CEO, and many GM managers considered him a smooth talker who belonged nowhere near the company's pinnacle. Stempel emphatically disagreed, often putting his arm around Reuss' shoulders and exclaiming, "Lloyd's my guy!" Not anymore, said the board, as GM's losses sank to historic depths. When the directors took the chairman's title away from Stempel, they also demoted Reuss, and when they fired Stempel six months later, they booted Reuss too.

"I can coach him." The CEO of a FORTUNE 500 manufacturer brought in an outsider a few years ago to run North American operations and eventually become the next CEO. The executive missed his commitment the first year, then missed it again the second, causing the whole company to fall short of its publicly stated promises to Wall Street. The CEO decided he wasn't giving the subordinate enough coaching and resolved to help more. He was human. But was this response humane? It wasn't. Results continued to decline, the stock collapsed, and the company was taken over. Both executives are gone, later joined by several thousand employees deemed unneeded by the new owner. It isn't uncommon for a strong CEO, otherwise decisive, to be blind to this fatal flaw.

"Wall Street and the press like him--I'd better keep him around." When a failing subordinate forms strong links with these important constituencies--sometimes through his own public relations efforts--the CEO faces a dilemma. Poor performance hurts the company's results, but taking out the subordinate may hurt its image. Typically the CEO doesn't act until the problem is acute, and by then it's sometimes too late.

"I've fired a lot of people lately. The board won't like it if I sack another." Specifically, the board may begin to worry that the CEO isn't developing the company's leadership. But if the subordinate is failing, delaying action just makes the problem worse.

"He's in the job, and I'll take the devil I know over the devil I don't." The CEO may be insecure about his ability to hire an outsider, especially someone from outside the industry. If the company has a strong, insular culture, he may rationalize that the culture wouldn't accept an outsider.

We've heard all these statements, and they're virtually always a sign of trouble ahead. Quick action on problems in the top team is simply imperative. Bob Allen of AT&T deserves credit for trying to break company (and Bell System) tradition by concluding that his successor had to come from outside. He recruited four candidates--most notably President John Walter--but none worked out. When Walter got fired, the board seized control of the process, and the company took considerable heat from Wall Street and the press. "If you have three or four people in the mill and some run short along the way, you can't wait," says Larry Bossidy of AlliedSignal, one of America's most successful CEOs. "You've got to make a change right then."

Yet you needn't be ruthless to get things done. Ron Allen's willingness to swing the ax so antagonized Delta's work force that the board asked him to leave. When Lou Gerstner parachuted in to fix the shambles John Akers had left of IBM, famously declaring that "the last thing IBM needs right now is a vision," he focused on execution, decisiveness, simplifying the organization for speed, and breaking the gridlock. Many expected heads to roll, yet initially Gerstner changed only the CFO, the HR chief, and three key line executives--and he has multiplied the stock's value tenfold. The best CEOs never hesitate to fire when they must, but the larger point is that they're deeply interested in people--far more so than failed CEOs are.

GE's Jack Welch loves to spot people early, follow them, grow them, and stretch them in jobs of increasing complexity. "We spend all our time on people," he says. "The day we screw up the people thing, this company is over." He receives volumes of information--good and bad, from multiple sources--and he and his senior team track executives' progress in detail through a system of regular reviews. His written feedback to subordinates is legendary: specific, constructive, to the point. Of course some come up short. When Welch committed the company to achieving six-sigma quality a few years ago, he evaluated how the beliefs of high-level executives aligned with six-sigma values. He confronted those who weren't on board and told them GE was not the place for them.

This continual pruning and nurturing gives GE a powerful competitive advantage few companies understand and even fewer achieve--extraordinary longevity in top executives. Consider: Robert Wright is in his 13th year running NBC; vice chairman Dennis Dammerman was CFO for 14 years; Gary Wendt ran GE Capital for 12 years; John Trani ran GE Medical for 11 years; vice chairman Eugene Murphy has been in top positions for 13 years, plastics chief Gary Rogers for 13 years, vice chairman John Opie for 16 years. Because Welch has the right people in the right jobs, he can leave them there and things tend to get better, not worse.

The motto of the successful CEO, worthy of inscription on his or her office wall, is "People first, strategy second."

Regular review of subordinates is a vital process, but every process carries a mortal danger--that the CEO will forget its purpose and begin to think that the process itself is what matters. It happens all the time. A CEO becomes committed to an organizational model. Maybe he insists on 100% consensus. Middle managers resort to informal networks to get things done. Cliques form. Indecisiveness takes over, and a fast-moving competitor grabs the advantage.

Decision gridlock can happen to anyone, but it happens most often to CEOs who've spent a career with one company, especially a successful one. The processes have worked, they're part of the company's day-to-day life--so it takes real courage to blow them up.

Listen to Elmer Johnson, a top GM executive, describe this problem to the executive committee: "The meetings of our many committees and policy groups have become little more than time-consuming formalities. The outcomes are almost never in doubt.... There is a dearth of discussion, and almost never anything amounting to lively consideration.... It is a system that results in lengthy delays and faulty decisions by paralyzing the operating people...." That was in 1988, during Roger Smith's troubled tenure, and the problem persisted through Stempel's brief reign. Neither man could break the process machine, and both must be considered failed CEOs.

Process gridlock is never good, but in the unforgivingly fast Internet age it's the way to catastrophe. It was a major problem during Gil Amelio's short time atop Apple Computer. Roger Siboni, who spent 20 years as a KPMG consultant, now runs a Silicon Valley startup called Epiphany and says the differences in process are stark: "You can't imagine the contrast here with the cordialness of corporate America. That whole world--meetings, facilitators...facilitators? Out here that would be ludicrous." There's just no time.

Effective CEOs use processes to drive decisions, not delay them. They start by focusing on initiatives that are clear, specific, and few, and they don't launch a new one until those in progress are embedded in the company's DNA. We've heard many employees, and so have you, speak witheringly about their CEO's flavor of the month--vision statements, quality, empowerment, leadership, all of which beget process and apparatus. By contrast, Welch has introduced just five major initiatives in 18 years as CEO (the most recent is e-commerce).

With their initiatives firm, effective CEOs implement them through a process that seems simple, even obvious, but has profound effects. Watch the likes of Welch or EDS's Richard Brown or Bossidy or any other proven implementer in a meeting. Near the end he'll grab a pen and start writing: He's noting exactly what is supposed to be done by whom, by when. He'll go over this with everyone before the meeting closes, and he'll probably send each one a reminder afterward.

It's fascinating to watch what happens when a CEO who executes well brings these habits into a company where they didn't exist. The whole tone changes. People prepare for meetings differently. They interact differently. They begin to see a fundamental distinction between failed CEOs and effective ones: For many failures, process is everything; for the great ones, commitments are everything. As Dick Brown says, "Delivering on commitments is the most important thing." Great CEOs hold people accountable, always.

Keeping track of all critical assignments, following up on them, evaluating them--isn't that kind of...boring? We may as well say it: Yes. It's boring. It's a grind. At least, plenty of really intelligent, accomplished, failed CEOs have found it so, and you can't blame them. They just shouldn't have been CEOs.

The big problem for them is not brains or even ability to identify the key problems or objectives of the company. When Kodak ousted Kay Whitmore, conventional wisdom said it was because he hadn't answered the big strategic questions about Kodak's role in a digital world. In fact, Kodak had created, though not publicized, a remarkably aggressive plan to remake itself as a digital imaging company. Whitmore reportedly embraced it. But he couldn't even begin to make it happen. Same story with William Agee at Morrison Knudsen--plausible strategy, no execution.

The problem for these CEOs is in the psyche. They find no reward in continually improving operations. Failing CEOs ask, "Why can't people do it themselves?" They're afraid of being seen as too controlling. The winners have what Bossidy calls "a drive to be competitive all the time--competitive in the operational sense." They get a charge out of pushing, pushing, pushing to make change happen.

That's why they're also constantly hungry for information from the battlefield. Effective CEOs have a strong external focus and get stimulated by details of what's happening in their markets, details that others might find boring. They're haunted by a familiar warning: "The CEO is always the last to know." They pull in loads of data from diverse sources. Then, as Welch says, you don't do what you want to do, you do what must be done--what reality demands.

Failed CEOs, by contrast, avoid facing market realities in all sorts of inventive ways. They remain in denial (see next article). They may become prisoners of one or two executives or of a guru or consulting firm, looking nowhere else for advice. Or they may look outward--but not at their markets.

Some CEOs get distracted by serving on too many boards. Others, like former American Express CEO James Robinson, see themselves as global ambassadors and lose focus. John Sculley became enamored of politics--he was a vocal supporter of Bill Clinton. By the final months of his tenure, the board realized he "was not focused on the day-to-day operations of Apple, other than on its technology," said former inside director Albert Eisenstat in a lawsuit. When profits deteriorated, the board asked him to leave.

But wait. In all this talk about CEOs and execution, aren't we forgetting someone? What about the COO? If operating the company isn't the job of the chief operating officer, whose is it?

Good question, but it doesn't get the CEO off the hook. Certainly some CEO-COO partnerships have been terrifically successful. Look at Tom Murphy and Dan Burke at Capital Cities/ABC or Roberto Goizueta and Don Keough at Coke. Today, Steve Case and Bob Pittman at AOL could be a winning team.

But be careful--these partnerships depend on a rare chemistry that's hard to predict, and the stakes are high. If it doesn't work, the resulting trouble is worse than most. Compounding it, the CEO must then fire the COO fast, which is often a problem.

Note how many of today's best CEOs, the master executors, don't even have a COO: Craig Barrett of Intel, Bossidy, John Chambers of Cisco, Michael Dell of Dell, Gerstner of IBM, Ray Gilmartin of Merck, Herb Kelleher of Southwest Airlines, Jacques Nasser of Ford, and Welch, among others. That's a multi-industry all-star team of CEOs who've put themselves squarely in charge of meeting their commitments and getting things done. Of America's ten most admired companies, as determined in FORTUNE's latest survey, eight don't have COOs (Microsoft and Wal-Mart are the exceptions). Most of the best CEOs seem to agree with Bossidy, who acknowledges that COOs can work but believes that someone needs to "know in total what's going on." His view: "It's best to have that responsibility invested in one as opposed to two people."

Any way you look at it, mastering execution turns out to be the odds-on best way for a CEO to keep his job. So what's the right way to think about that sexier obsession, strategy? It's vitally important--obviously. The problem is that our age's fascination with strategy and vision feeds the mistaken belief that developing exactly the right strategy will enable a company to rocket past competitors. In reality, that's less than half the battle.

This shouldn't be surprising. Strategies quickly become public property. Ask Michael Dell the source of his competitive advantage, and he replies, "Our direct business model." Okay, Michael, but that's not exactly a secret. Everyone has known about it for years. How can it be a competitive advantage? His answer: "We execute it. It's all about knowledge and execution." Toyota offers anyone, including competitors, free, in-depth tours of its main U.S. operations--including product development and distributor relations. Why? The company knows visitors will never figure out its real advantage, the way it executes. Southwest Airlines is the only airline that has made money every year for the past 27 years. Everyone knows its strategy, yet no company has successfully copied its execution.

Yes, strategy matters. A good, clear strategy is necessary for success--but not sufficient for survival. So look again at all those derailed CEOs on the cover. They're smart people who worried deeply about a lot of things. They just weren't worrying enough about the right things: execution, decisiveness, follow-through, delivering on commitments.

Are you?

Dallas-based RAM CHARAN advises many FORTUNE 500 CEOs and is author of an acclaimed book on corporate governance, Boards at Work.