Ending the CEO Succession Crisis

by Ram Charan

WE TALK ABOUT LEADERSHIP as though leaders—like Tolstoy’s happy families—are all alike. But CEO leadership should be a subject apart because it is unique in scope and substance and of incomparable importance. CEOs’ performance determines the fates of corporations, which collectively influence whole economies. Our standard of living depends upon excellence at the very top.

Who, then, would dispute that CEO selection deserves perpetual front-burner attention from the custodians of a company’s welfare? Surely, when time or trauma ushers in change, organizations should be ready with a clear view of current and future needs and with carefully tended pools of candidates.

But they’re not. The CEO succession process is broken in North America and is no better in many other parts of the world. Almost half of companies with revenue greater than $500 million have no meaningful CEO succession plan, according to the National Association of Corporate Directors. Even those that have plans aren’t happy with them. The Corporate Leadership Council (CLC), a human-resource research organization, surveyed 276 large companies last year and found that only 20% of responding HR executives were satisfied with their top-management succession processes.

That deficiency is simply inexcusable. A CEO or board that has been in place for six or seven years and has not yet provided a pool of qualified candidates, and a robust process for selecting the next leader, is a failure. Everyone talks about emulating such best practitioners as General Electric, but few work very hard at it.

The result of poor succession planning is often poor performance, which translates into higher turnover and corporate instability. As increased transparency, more vocal institutional investors, and more active boards make greater demands, CEO tenures continue to shrink. Booz Allen Hamilton reports that the global average is now just 7.6 years, down from 9.5 years in 1995. And two out of every five new CEOs fail in the first 18 months, as Dan Ciampa cites in his article “Almost Ready” (HBR, January 2005).

The problem isn’t just that more CEOs are being replaced. The problem is that, in many cases, CEOs are being replaced badly. Too often, new leaders are plucked from the well-worn Rolodexes of a small recruiting oligarchy and appointed by directors who have little experience hiring anyone for a position higher than COO, vice chairman, CFO, or president of a large business unit. Hiring a CEO is simply different.

Coaxing former leaders out of retirement is another popular way to fill the void. Celebrated examples include Harry Stonecipher at Boeing, Bill Stavropoulos of Dow Chemical, and Jamie Houghton at Corning. But most “boomerang CEOs” return for just a couple of years, long enough to restore credibility and put a real succession candidate in place. They are not the long-term solution.

To increase their chances of finding a leader who will serve long and well, companies must do three things. First, they should have available a deep pool of internal candidates kept well stocked by a leadership development process that reaches from the bottom to the top. Second, boards should create, then continually update and refine, a succession plan and have in place a thoughtful process for making decisions about candidates. Finally, directors considering outside candidates should be exacting, informed drivers of the executive search process, leading recruiters rather than being led by them.

In my 35 years advising corporations, I have participated in dozens of CEO selections and have closely monitored numerous executive pipelines. Drawing on that experience, I will in these pages first explain why companies make poor appointments and then suggest what they can instead do to make good ones. Using these guidelines, organizations can ensure that all participants—directors, executive recruiters, and sitting CEOs—perform wisely and appropriately when it comes time to choose their next leader.

The Trouble with Outsiders

When companies lack the culture or the processes to grow their own heirs apparent, they have no choice but to look outside. More than a third (37%) of the Fortune 1,000 companies are run by external recruits, according to the public affairs firm Burson-Marsteller. Although global data are harder to come by, the worldwide trend appears to be similar. But external candidates are in most cases a greater risk because directors and top management cannot know them as well as they know their own people.

Outsiders are generally chosen because they can do a job—turn around the company or restructure the portfolio. But the job is to lead a hugely complex organization over many years through an unpredictable progression of shifting markets and competitive terrains. Unfortunately, the requirements for that larger job are often not well defined by the board, which may be focused on finding a savior.

The results are not surprising. In North America, 55% of outside CEOs who departed in 2003 were forced to resign by their boards, compared with 34% of insiders, Booz Allen reports. In Europe, 70% of departing outsiders got the boot, compared with 55% of insiders. Some outside CEOs are barely around long enough to see their photographs hung in the headquarters lobby. Gil Amelio left Apple 17 months after he arrived from National Semiconductor. Ex-IBMer Richard Thoman was out of the top spot at Xerox after 13 months. David Siegel gave up the wheel at Avis Rent A Car for US Airways but departed two years later.

Even under the best circumstances, CEO selection is something of a batting average: Companies will not hit successfully every time. But two or more consecutive outsider outs can have a devastating effect on employees, partners, and strategic position. New leaders import new teams and management styles. Continuity and momentum collapse, the energy to execute dwindles, and morale plummets as employees obsess about who will get the next pink slip. Rather than focus on the competition, companies starts to look inward. Bad external appointments are also expensive, since even poor performance is rewarded with rich severance packages.

The Trouble with Insiders

On the other hand, sometimes an external candidate exists who is, very simply, the best available choice. A skillful, diligent board may discover an outstanding fit between an outsider and the job at hand. Lou Gerstner and IBM spring to mind. And boards must remember that just as outsiders are not uniformly bad choices, insiders are not uniformly good ones. In certain situations, internal candidates actually present the greater risk.

Some concerns about insiders, ironically, emerge from their very closeness to the company. For example, as “known quantities,” they may sail through a lax due-diligence process. Or their social networks and psychological ties may complicate efforts to change the culture. Some will not have had the right experience or been tested in the right ways. Individuals from functional areas may not be up to the task of leading the entire business. Or a shift in the industry or market landscape may render carefully nurtured skills irrelevant. In some cases, the credibility of the outgoing CEO or management team may be so sullied that only a new broom can sweep the company clean.

What’s more, companies that have no ongoing senior management development program (currently more the rule than the exception) will in all probability need to look outside, maybe for as long as the next ten to 20 years. Outside candidates, in other words, should always be an option. But so long as they remain the only option, and boards lack rigor in identifying and assessing them, succession is imperiled.

The Trouble with CEO Development

Many organizations do a decent job nurturing middle managers, but meaningful leadership development stops well below the apex. The problem manifests itself as a dearth of senior managers, for which companies must increasingly shop in other neighborhoods. Almost half of respondents to the CLC survey had hired a third or more of their senior executive teams from outside, but only 22% of those did so because they considered external candidates irresistibly appealing. Rather, 45% of all respondents judged that it would take too long or be too expensive to develop successors internally.

It’s easy to understand why they feel that way. Even where strong development programs exist, very few leaders will ever be qualified to run the company. Very few. A $25 billion corporation with 70,000 employees, for instance, may have 3,000 leaders, perhaps 50 to 100 of whom would qualify for one of the ten jobs just below the top. That same company would be fortunate to field five strong internal candidates for CEO—and two or three is a more realistic number. General Electric had around 225,000 workers in 1993 when Jack Welch identified 20 potential successors; over seven years, he winnowed the number to three. In CEO succession, it takes a ton of ore to produce an ounce of gold.

Furthermore, the window in which to spot CEO talent is narrow. Companies require sufficiently seasoned candidates who can be counted on to hold the top job for ten years or more. That puts the age of accession at between 46 and 52. In my experience, for a candidate to be ready by 46, serious development should start by age 30. Recognizing which five saplings in a 3,000-tree forest are the ones to nurture requires a degree of discernment that most line managers and HR departments lack and few are developing.

Some companies do identify candidates early but then fail to evaluate them properly. Such organizations often turn evaluation over to HR, which may rely excessively on packaged databases of leadership traits developed by researchers in the human behavior field. Those programs compare internal high potentials with generic benchmarks along many dimensions, a process that creates fragmented profiles of some cookie-cutter ideal rather than nuanced, individualized portraits. What’s more, most of those dimensions reflect only the personality traits and not the skills required of a CEO.

Nor do many companies properly nurture the candidates they identify. Some misjudge the business’s needs and consequently emphasize the wrong talents. Only 24% of organizations the CLC surveyed believe their leadership development efforts are aligned with their strategic goals. And those goals can be a moving target, changing in response to sometimes tectonic shifts in the external environment. The marketplace changes. Technology changes. Employees’ skills become obsolete even as they develop. What’s more, very few in-house executive education programs are designed to impart the skills and know-how that a CEO needs.

But the larger issue is that true development happens on the job, not in a classroom. Few companies know how to get their best people the experiences that would prepare them for the CEO role or to rigorously evaluate them in the jobs they do perform. Many companies, for example, still equate leadership development with circulating candidates through multiple functions. In the 1970s, that was the rule at AT&T, IBM, and Xerox, companies that produced leaders who went on to become CEOs elsewhere—and in some cases failed.

The problem with that approach is that potential candidates don’t stay long enough in one position to live with the consequences of their decisions. In addition, functional leaders learn to lead functions, not whole companies. Faced with external competition, they fall back on their functional expertise. You can mine all possible lessons from a turn as VP of marketing and still be blindsided by a P&L.

The Trouble with Boards

Bob Stemple’s short stint as the head of General Motors ended ingloriously in 1992—and so did the accepted wisdom that boards should automatically bless the departing CEO’s handpicked successor.

Yet while directors describe CEO succession as one of their most consuming issues, they don’t appear consumed by it. In a survey by Mercer Delta and the University of Southern California, 40% of corporate directors called their involvement in CEO succession planning less than optimal. (I would hazard to add that far fewer are satisfied with the outcome of their involvement.) Only 21% responded that they were satisfied with their level of participation in developing internal candidates for senior management.

A packed agenda is the chief culprit: Governance and fiduciary duties, in particular, command an outsize share of boards’ attention. Mercer Delta asked directors to compare the amount of time they spend now with the amount they spent a year earlier on nine key activities. Large majorities reported devoting more or many more hours to monitoring accounting, Sarbanes-Oxley, risk, and financial performance. They also reported spending less time interacting with and preparing potential CEO successors than on any other activity. Yet boards’ work on succession represents probably 80% of the value they deliver. If the choice of CEO successor is superb, all subsequent decisions become easier.

Another huge problem is that the vast majority of search committee members have had no experience working together on a CEO succession. As a result, they seldom coalesce into deep-delving bodies that get to the pith of their companies’ fundamental needs. So they end up approaching their search with only the demands of the moment or—worse—the broadest of requirements.

As they audition candidates, directors may be seduced by reputation, particularly if they’re considering a Wall Street or media darling. A few aspiring CEOs employ publicists who flog rosy stories to journalists; when those leaders are up for other jobs, their press-bestowed halos follow them. Board members can also be blinded by charisma, by the sheer leaderishness of a candidate. There is nothing intrinsically wrong with charisma, though some criticize it as the sheep’s clothing in which hubris lurks. But too often directors become so focused on what candidates are like that they don’t press hard enough to discover what candidates can and cannot do.

For example, one board looking for a new CEO after firing the old one asked for someone who could build a great team and get things done. The recruiter presented such a person—an energetic, focused candidate whose personal qualities quickly won over directors. What the organization really needed was someone who could create a stream of new products and win shelf space from powerful retailers in a volatile marketplace. Unfortunately, the directors never specified those requirements or raised them either during interviews or the background check.

The candidate’s upstream-marketing skills were poor to nonexistent. The company’s market share declined precipitously, and three years later the CEO was out on his ear. On its second try, the board concentrated so hard on marketing that it ignored execution. The next CEO was a visionary and a marketing genius but was unable to get things done. The company, once first in its market, will probably be sold or stumble into Chapter 11.

Finally, directors too often shunt due diligence onto recruiters. As a result, that process can be quite superficial. One company that left vetting to its recruiter and its investment banker found itself saddled with a leader who botched critical people issues. At a postmortem three years later, directors discovered that at his former company the CEO had routinely punted people problems to the chairman, who had been CEO before him and occupied the office next to his. That would have been nice to know before the pen touched the contract.

The Trouble with Recruiters

Executive recruiters are honest and highly professional. Still, they can wield disproportionate influence in CEO succession decisions. One reason is concentration. Just three recruiters control some 80% of the Fortune 100 CEO search market (a single firm claims fully 60% of it), and one or two people within those companies direct the most important searches. These firms’ social networks are vast and powerful. Anyone with a smidgen of ambition in the corporate world knows whom they have to know to get ahead.

At the same time, board members’ inexperience and consequent inability to precisely define their needs makes recruiters’ task difficult. Recruiters must satisfy their clients yet also manage them, helping the search committee to gel so they can extract the criteria they need while keeping requirements broad enough to cast the widest talent net possible.

When committees don’t gel, recruiters may step into the vacuum with their own criteria, and directors too often let them. Unfortunately, no executive recruiter can grasp the subtleties of a client’s business as well as the client can. In the absence of effective direction, recruiters generally approach each search with a boilerplate of the 20 or so attributes they consider most desirable for any CEO. That formula tends to overemphasize generic qualities like character and vision, as well as team-building, change-management, and relationship skills. Psychology and chemistry are also very important to executive recruiters: Like directors, they may let a personality surplus overshadow a skills deficit.

In one—granted, extreme—case, the longtime CEO of a company with four highly successful businesses and a huge debt level was retiring. The recruiter produced a list of six candidates, pressing one—the head of a very large division at a multinational company—hard on the board. Yet all the recruiter gave the directors was a page-and-a-half description of this candidate’s leadership skills; a list of his extensive connections with unions, customers, and government bodies; and an outline of his swift rise through the organization.

A financial performance history for the candidate’s division was not included and not publicly available, so a member of the search committee began to dig. He discovered that return on assets under the candidate’s supervision was miniscule over the previous five years, even though his division was four times larger than the entire company considering him for CEO. Furthermore, this man had never earned cost of capital in his life. Even so, the recruiter wanted to put him in charge of a business that had certainly done so—and that hoped to rise to the next level.

Fortunately, after much debate, the committee vetoed the recommendation, opting instead for number three on the recruiter’s list—the president of another company, who had consistently improved performance and delivered a 20% return on equity. In his first three years, this new CEO took the stock from 24 to 108 in a slow-moving industry. The board was happy. Management was happy. The recruiter’s preferred candidate was happy when he was placed at another, larger company—but then he was fired in six months.

Executive recruiters also succumb to the usual-suspects bias, primarily looking for new heads above other companies’ necks. It is just plain easier to compile a list of sitting CEOs than to make a case for—or take a risk on—a COO or an executive VP. Some recruiters go so far as to approach sitting CEOs, even with no specific jobs to dangle, and urge them to consider looking elsewhere. The recruiters’ goal is to loosen a prized gem from its setting and thereby beat a fellow recruiter to the punch.

Sometimes, the board’s selection of recruiter is flawed from the start. A director may jump the gun, recommending a recruiter he has worked well with even before the search committee is formed. Nor do most boards examine search firms’ track records—that is, how many of the CEOs the firm has placed have succeeded and how many have failed. Even if directors did ask that question, they’re not likely to get the answer because it appears no one is monitoring recruiters’ performance. The stock-buying public, by contrast, knows exactly how well directors score on their CEO choices.

How to Succeed at Succession

Charlie Bell’s ascension to the top spot at McDonald’s within hours of Jim Cantalupo’s death reflected well on a company that had its house in order, particularly when compared—as it inevitably was—with Coca-Cola’s simultaneous travails. Similarly, NBC’s early, orderly announcement that Brian Williams would replace network news anchor Tom Brokaw stands in stark relief to CBS’s public uncertainty over Dan Rather’s successor. (Anchors are not CEOs, of course, but they are even more visible and arguably as consequential to their organizations’ fortunes.)

By now it should be clear that the most important thing companies can do to improve succession is to bolster their leadership development and focus on those very rare people in their ranks who might one day be CEO. Organizations must identify high-potential candidates early in their careers, and global companies must look in all the countries where they operate. As candidates enter the development pipeline, managers must constantly align their charges’ education and on-the-job experience with the emerging landscape. And they must rigorously assess the candidates’ performance at each developmental stage.

The very best preparation for CEOs is progression through positions with responsibility for steadily larger and more complex P&L centers. A candidate might start by managing a single product, then a customer segment, then a country, then several product lines, then a business unit, and then a division. Whatever the progression, P&L responsibility at every level is key. The Thomson Corporation, a global provider of information solutions, comprises more than 100 P&Ls, so its top people have abundant opportunity to run a $50 million to $100 million business. “That’s the best crucible for formulating leaders that I know of,” says Jim Smith, executive vice president of human resources and administration.

Companies not set up to provide such opportunities should create jobs—large projects or small internal organizations—that exercise the P&L muscle. Otherwise, they risk elevating an internal candidate who is not prepared. For example, one $10 billion company in a highly capital-intensive and unionized industry has targeted as CEO successor the head of its smallest division. The candidate is a brilliant, articulate young man but has no experience running a big business in general or this type of business in particular (his own division is knowledge intensive, and unionized labor has no presence). The board is considering creating a deputy position within its largest division for this person and making the 59-year-old current division head (who will retire in three years) his coach, granting that coach a bonus if he ensures his successor’s success.

Companies with inflexible functional structures will probably be forced to import P&L-tested leaders from outside and place them in very high positions. To reduce the risk, they should bring in such executives three or four years before the expected succession. That can be challenging, however, because many will demand appointment to the top spot in less than a year.

But leadership development is just part of the solution. Boards, too, can greatly improve the chances of finding a strong successor by acting vigilantly before and during the search. Senior executive development should be overseen by the board’s compensation and organization committee, which needs to receive periodic reports on the entire pool of potential CEOs and regular updates on those bobbing near the top of it. The committee should spend a third of its time examining lists of the top 20 candidates in the leadership pipeline. In addition, at least 15% of the 60 or so hours that members meet as a full board should be devoted to succession. At minimum, the board ought to dedicate two sessions a year to hashing over at least five CEO candidates, both internal and external.

And directors should personally get to know the company’s rising stars. Promising leaders should be invited to board meetings and to the dinners that precede board meetings, and members should talk with them informally whenever possible. Directors should also meet with and observe candidates within the natural habitats of their business operations. In this way, when it comes time to single out CEO candidates, directors will be considering a set of very well-known quantities.

The “Fit” Imperative

The goal of all these interactions and deliberations is for board members to reach a highly refined but dynamic understanding of the CEO position and their options for it long before appointing a successor. Company leaders should be as well defined as puzzle pieces; their strengths and experiences must match the shape of their organizations’ needs. That is, they simply must fit. Boards achieve fit by specifying, in terms as precise as possible, three or four aspects of talent, know-how, and experience that are nonnegotiable.

Ideally, these attributes pertain to the organization’s dominant needs for the next several years, but they should also relate to future growth. In one recent CEO succession, the company, in conjunction with a boutique recruiting firm, began with impossibly broad criteria that included everything from industry leader to change agent. The process floundered until the search committee narrowed its focus to three qualities: experience in segmenting markets according to customer needs; the talent to grow the business organically; and a track record of building strong executive teams. Those three skills, in addition to general leadership traits, delineated the pond in which this company fished.

The job of defining such qualities belongs to the search committee, which should form well before succession is scheduled to take place. As they wrestle with requirements, committee members must constantly keep in mind the company’s changing circumstances, so that an understanding of what currently works doesn’t congeal into what works, period.

For example, Bank of America flourished under deal maker par excellence Hugh McColl, Jr., for years. But by the time he stepped down in 2001, integration, rather than acquisition, had become the dominant challenge. Having recognized the altered environment several years before, BOA’s board chose not a leader in McColl’s image but instead Ken Lewis, a company veteran proficient at integration of acquisitions and organic growth. (For an example of how a company integrates its leadership development with its strategy, see the sidebar “The Living Succession Tree.”)

Specific, nonnegotiable criteria also let directors keep control when they work with executive recruiters. With good direction, search firms can be a valuable source of objectivity—benchmarking internal candidates against outsiders and making sure that board members consider all possibilities, even if they prefer an insider. Some companies even bring in recruiters to do independent assessments of insider candidates. Their concurrence with a board’s judgment carries weight with shareholders and potential critics.

Search firms ask boards to recommend candidates, and they take those recommendations seriously. But, ultimately, it is the recruiter who compiles the list, and the compiler of the list wields considerable influence. Directors must require from recruiters detailed explanations of how the candidates fulfill their criteria. A ten-page report on each is reasonable.

When the time comes to select the new CEO, directors—ordinarily a polite breed, unaccustomed to challenging one another or asking discomforting questions—must engage in a vigorous discussion of the candidates’ comparative merits. One search committee that did an outstanding job making the final decision invited five candidates (two internal and three external) to a hotel for a couple of days. The two internal candidates were favorites of two different directors. On the first day, the committee interviewed three candidates, two external and one internal. The directors split into two groups of three, and each group spoke with one candidate for 90 minutes. After these interviews, the directors broke for 45 minutes to share impressions, then switched candidates. Then the two groups of directors took turns interviewing the third candidate, similarly sharing impressions informally. At the end of the first day, the committee members debated over dinner, and the director who had originally advocated for the internal candidate volunteered that he was indeed not the strongest choice. The next day, they repeated the process with the two other candidates, and the results were remarkably the same, with the director who had originally advocated for the internal candidate changing his mind. In the course of these discussions, all hidden agendas fell away, requirements were honed, and directors were able to reach consensus.

Finally, board members must do due diligence on outside candidates—and do it well. Directors must seek reliable external sources and demand candor from them. Board members should ask first about the candidate’s natural talents. If those gifts—admirable as they may be—do not match the position’s specific profile, that candidate is not worth pursuing. Needless to say, due diligence is also the time to root out any fatal character flaws.

At this point, the role of the outgoing CEO is chiefly consultative. He or she must be active in spotting and grooming talent, help define the job’s requirements, provide accurate information about both internal and external candidates, and facilitate discussions between candidates and directors. But when the choice of successor is imminent, make no mistake: That decision belongs to the board.

Inside a Development Engine

Despite the current crisis, we know it is possible to build organizations that reliably produce great CEOs. Starting after World War II, a few corporations emerged as veritable leadership factories. Companies like General Electric, Emerson Electric, Sherwin-Williams, Procter & Gamble, and Johnson & Johnson managed to stock not only their own corner offices but also many others. (Of course even great companies sometimes stumble: Procter & Gamble had a failure from within when it promoted Durk Jager to the top spot. But it is going great guns under the stewardship of company veteran A.G. Lafley.)

Reuben Mark has sat atop Colgate-Palmolive for 20 years, so the company’s CEO succession chops have not been recently proven. But I believe the consumer products giant has a first-rate process for identifying and developing CEO talent. At the very least it produced three internal candidates who are excellent prospects for the job.

Colgate-Palmolive does business in more than 200 countries, and its emerging leaders are correspondingly international and diverse. Leadership evaluation begins during the first year of employment. “It may seem strange to talk about someone who’s been here just a year when discussing the pipeline to the CEO,” says Bob Joy, senior vice president of global human resources. “But the earlier you start to identify talent, the earlier you can provide the job assignments and develop the broad business experience needed by a CEO candidate.”

Each subsidiary identifies its own high potentials and submits that list to local general managers, who add and subtract names and then hand the list off to the division heads. These lists wend their way up the chain until they reach the Colgate-Palmolive Human Resource (CPHR) committee, composed of Colgate’s CEO, president, COO, the senior VP of HR, and the senior candidates up for the top job. CPHR modifies and consolidates the lists into a single master list, dispatching it back down the ranks where managers can contest decisions made by those above them. The process takes place once a year.

Those who make the cut are deployed in one of three tracks. The first track, local talent, is for relatively junior staff who might become the direct reports of a general manager. Someone more advanced would be designated regional talent, and given, for example, a significant position in Asia. The most elevated track—global talent—is the reservoir from which the most senior jobs are filled.

Throughout their careers, all these high potentials receive assignments that stretch their abilities and expand their knowledge, exposing them to a variety of markets, cultures, consumers, and business circumstances. CPHR itself designs career paths for general managers and higher positions because the committee is at the same time dynamically developing the profile of Colgate’s future leadership team. (Also, says Joy, “you can imagine the kind of resistance you’d get from a division president who would like to keep his high-potential people in his own area.”) The thousand or so highest high potentials (out of a total pool of about 2,000) receive outside executive coaching, which includes 360-degree feedback on current and past assignments.

Having identified its high potentials, Colgate strives to bolster their connection to the company. One tactic is recognition: “If you’re talking about the future leaders of your company, you want them to feel special,” says Joy. “You want them to have Colgate in their veins.” Toward that end, the company sponsors a series of “visibility programs.” One, for example, gathers high potentials from all over the world at Colgate’s New York headquarters for week-long sessions during which they meet with every senior leader in the company. In addition, each high potential receives a special stock grant, which arrives with a personal letter from the CEO.

Colgate’s global growth program mandates that all senior managers retain 90% of their high potentials or lose some compensation. If a high potential at any level, anywhere in the world, does resign, the CEO, the COO, the president, and Joy are alerted within 24 hours and move immediately to retain that person.

Perhaps most important, Joy collaborates with the office of the chairman to connect directors early and often with high potentials in all areas. At the most senior level, functional leaders introduce the board to the top two or three most-promising heirs for their own positions, adding detailed analyses of those candidates’ strengths and weaknesses. Emerging leaders routinely take part in presentations to the board and meet informally with directors over lunch. Board members closely track the progress not of one or two people but of the top 200, frequently discussing how each piece fits into the puzzle and what experiences or skills might improve that fit.

As a result, when CEO succession looms, the board and top management will be able to select from candidates they have spent many, many years observing and evaluating. “If you start five years or even ten years before the CEO is going to retire,” says Joy, “it may be too late.”

Of course Colgate-Palmolive—like General Electric—tackles succession from a position of strength. Its CEO has been two decades in the saddle, and he is passionate on the subject of an heir. Companies with less-veteran chiefs—and whose boards have been negligent in this area—will probably need to line up candidates quickly, while laying a deeper pipeline. They will in all likelihood have to bring in outsiders and position them to gain the requisite business and industry experience. That may mean shaking up the leadership team and reporting structures to free up slots in which outsiders can be tested. This restructuring will probably be resented, but it is necessary pain.

A quick infusion of talent may be a company’s only course, but it is no way to run a railroad. Organizations without meaningful pipelines must start now to put them in place. Young companies should create the processes that will come to fruition in five or ten years’ time. Choosing the CEO’s successor is not one decision but the amalgam of thousands of decisions made by many people every day over years and years. Such meticulous, steady attention to defining needs and evaluating candidates produces strong leaders and inspires succession planners at lower levels to exercise the same discipline.

The trend of CEO failures must be reversed. CEO succession is all boards’ paramount responsibility; nothing else so profoundly affects their companies’ futures. Directors must start investing their time and energy today. The call for a new leader could come tomorrow.

Originally published in February 2005. Reprint R0502C