HOW to SUCCEED at SUCCESSION.

To get a closer look at CEO succession planning, CE gathered CEOs - several of whom have recently come to terms with what is said to be the most important responsibility of their job.

Upheaval in the corner office seems a side effect of these shareholder-value-or-else times. Abrupt transitions so often emphasize the separation of old and new, with the "old" becoming an inevitable negative reference point and the "new" embraced largely for its newness. Le roi est mort; vive le roi. Nor are messy transitions confined to U.S. companies coping with the demands of impatient institutions or trigger-happy boards. When Kurt Hellstrom took over control of Sweden's Ericsson AB after the dismissal of Sven-Christer Nilsson last July, the promise of a rapid turnaround helped boost the telecom giant's shares 9 percent despite performance falling below consensus estimates in the first half of 1999. Even staid Swedes will throw out a CEO or two in the name of performance.

In Built to Last, Jim Collins and Jerry Porras observe that companies that succeed over time have developed a culture of renewal. Hewlett-Packard earned high praise from the authors for its promote-from-within tradition. Recently H-P reached not only outside its ranks, but a level down in choosing Lucent Technologies executive Carleton Fiorina.

If good internal development is basic management hygiene, what does it say about a company when succession becomes internally disruptive? The departure of Eckhard Pfeiffer from Compaq and the circus that ensued didn't do wonders for team-building, particularly after several potential candidates turned down Compaq's board. "By promoting COO Mike Capellas to CEO, Compaq's board has decided against major change - a decision that severely limits the value of the Digital and Tandem acquisitions and will doom Compaq to focusing on bloody commodity PC wars," asserts Forrester Research's Matthew Nordan.

A similarly disruptive transition period turned Merrill Lynch upside down when it became clear to Herbert Allison that he wouldn't get the top job when current CEO David Komansky steps down. Merrill's senior people reportedly preferred someone who had come up the broker sales side of the business, the firm's traditional source of top talent. The curious aspect of this case is that succession at Merrill is historically well choreographed. If a technology guy with investment banking experience didn't have the right sales stuff, why wasn't that apparent?

Contrast this with George Fisher's waning rule at Kodak. Say what you will about Fisher's performance at Big Yellow, one of his earliest moves was to reshuffle the Kodak management team, reaching deep into its ranks to create a new team - including president Dan Carp, who will shortly succeed him.

Then there's the case of AlliedSignal, which shows that even the best-laid succession plans can go awry. Dan Burnham and Fred Poses were Larry Bossidy's chosen pair of successors until Raytheon pinched Burnham to be its next CEO. Had the AMP takeover succeeded, Poses would have likely assumed the AlliedSignal top job in October. As it happened, the bid was rebuffed and AlliedSignal merged with Honeywell, with Michael Bonsignore named CEO. A director at AMP who asked not to be identified says that AMP's board actually did Bossidy a favor by forcing him to find a more suitable merger mate in Honeywell, yet it ruined his succession plan and forced Poses to depart.

Still, creating a "succession" culture where the board and the CEO are deeply involved in developing talent at all levels of the firm can take away much of the human drama that surrounds the process. In the following discussion held in partnership with Deloitte Touche Tohmatsu, CEOs who have recently faced what is arguably their most important responsibility raise a number of key issues, such as:

* How closely a company should link management development at all levels with CEO succession.

* Focusing the board so that governance issues don't become divisive when holding formal succession discussions without the CEO present or a party to the discussion.

* Avoiding "horse race" succession.

* Preventing top management talent from heading toward the exits when bringing in an outsider.

* Board capability in closely held or family dominated firms.

* An executive search firm's role. Several participants also raised the issue of the board's responsibility to assert itself, and, if necessary, to force the CEO to step aside and let the board manage a more transparent process. Board committees can facilitate the development of potential candidates and determine whether CEO pay should be linked to sound succession plans. Most agreed boards that fail to develop a successful plan are asking for trouble and will be forced to pursue other, riskier options not necessarily in the shareholders' interest.

- J. P. Donlon

TELESCOPIC VISION

Ed Kangas (Deloitte Touche Tohmatsu): Succession is a little bit like shooting clay pigeons; you've got to understand the business is moving and changing. The CEO that you're looking to have in three or four years might need a very different set of skills than the CEO has today. Ten years ago, in the merger that created Deloitte & Touche, I became managing partner of the U.S. practice, and was chairman and CEO of the global firm. But most of my job was managing the part of the U.S. practice that dealt with integrating the two firms after the merger. Mike Cook was CEO of the U.S. practice only, and his job was primarily to focus on the marketplace and clients. Five years ago, I became chairman and CEO, or continued as chairman and CEO, but with more reality to it because our next step was to become a managerially and financially integrated firm at a global level.

So we went about the process of designing a global firm that would have a real management, a real CEO, and a real structure. We finished our planning around the globalization and the integration of our global firm. We asked a group of senior partners to identify candidates who could not only meet the needs of today, but be the CEO of the global firm when it truly was integrated. Their eight candidates were exposed to the CEO and the Board. We identified special assignments they might need to grow to the job and created a nominating committee who talked to 300 or so partners about what they thought of the leading candidates. In the process we developed, the board hears the results and selects one candidate to be the CEO, and then the partners vote yes or no. The votes are always 95 percent or more in the affirmative. Then we go through a transition.

It's very important to identify candidates years in advance, identify what specific experiences they will need, and make sure those experiences are based on the needs of tomorrow. In the case of Jim, we gave him a wealth of international experience. I created a variety of different places where he could play major roles with the current and future leaders of the global firm - along with similar kinds of things for other candidates. While the partners get to vote, our board is trusted to select the next CEO, with the advice of the current CEO. And our culture is such that we don't campaign. The transitions have worked reasonably well in the last several years.

James Copeland (Deloitte Touche Tohmatsu): A partnership is just different. We do have a culture of our partners supporting our board's decision, but that lasts just so long as our board acts on the direction they're given by our partners through a nominating committee process. And our board feels that responsibility strongly.

They have been elected by that same process, and they know that there are pros and cons. It's like every rose; you get thorns with it. You can't designate someone years in advance. That's the down side. The positive side is that you come into office knowing that you are the choice of your partners. It's very empowering to have partners say, "That was my candidate. I voted for that guy, I support him."

Kangas: Not only can the board not do it in advance, but if the CEO named a chosen successor more than nine months in advance, it would be the kiss of death.

Copeland: Yes, it would likely work to his disadvantage. You can position someone you think is good, but all you're doing is putting that horse in the race.

J.P. Donlon (CE): I've seen some top talent leave after this kind of process.

Copeland: In recent history, they never leave, because if you had several candidates nominated and the partners had to pick one, then you'd have losers. In our process you don't have losers.

Josh Weston (Automatic Data Processing): Some of my gray hair came from being on some public boards. Succession planning is the most important role of the board, occasionally effectively complemented by the CEO, but more often than not, knowingly or unknowingly, weakened by the incumbent. Everybody here can think of cases where the incumbent CEO not only didn't do the job, he was a detriment. I criticize the board when that happens more than the CEO. It's hard to get a CEO to judge when's the right time, who's as good as he or she is.

Steve Kaufman (Arrow Electronics): I'd add a third reason why it doesn't work out - the CEO picks someone who's been his loyal scout Tonto. I told our board that 10 years, plus or minus a year, was the right time for a CEO. That's not always the truth; Jack Welch is clearly the exception that proves the rule. If you take 100 CEOs who served 20 years and compare their first decade to the second decade, rarely is the second decade as good as the first. As we got to the seven- or eight-year mark, we said that the nominating committee should begin meeting twice a year, using me as staff officer, but driving the process.

Dayton Ogden (Spencer Stuart): Recently, it was announced that Matt Barrett had taken on the CEO job at Barclay's. At the beginning of his tenure at Bank of Montreal, Barrett had told the board that he would step down after a decade. He designated his successor early and prepared the successor, with the help of the board, and they had an orderly transition, despite a very difficult circumstance around a merger with the Royal Bank of Canada that never took place. So some companies do this very well. And it happens more effectively when the CEO and the board are working in partnership with the board giving the CEO incentive. The extent to which a CEO does a good job developing succession candidates should be a major performance measure. But, in the end, orchestrating the transition is the board's responsibility.

Donlon: Why the 50-50 success rate?

Kangas: The more dominant the personality and the ego of the current CEO, the more the board needs to take charge. Unfortunately, the reverse happens.

Tom Neff (Spencer Stuart): Many board members were handpicked and close friends with the CEO. If the CEO had been successful up to that point, it's difficult to challenge. And there is a tendency of CEOs to pick someone in their own likeness, or their loyal lieutenant. Succession is the time for reevaluation of the direction of the business and projecting from there what kind of profile is best needed. CEOs aren't necessarily going to do that.

Weston: One of the avenues, not the only one, for developing choices, as well as proving and training grounds for the CEO, is obviously the COO. Some boards are more permissive and let the CEO have the sole say on whether there's even going to be a COO and, if so, who to pick. And there are CEOs, who, whether wittingly or unwittingly, opt for less than the most upwardly mobile COO.

Craig Winn (Value America): There's the partnership model, the large public firm model, and the world I come from, which is entrepreneurs who created a company that now needs to be managed. Almost all entrepreneurs fail when their business goes from small to medium. And those who don't fail going from small to medium fail going from medium to large. I was leading a company that in its first year did tens of millions, and its second year is now doing hundreds of millions, and its third year will do billions. So I recognized the likelihood of me failing was enormous and decided to call the smartest businessman I've ever known, Federal Express's Fred Smith, who happens to be on our board.

Fred Smith said, "What you want is someone who has the opposite characteristics that you have. Someone who can still focus on ethics and morality and who you can trust with your company and with your life, but, in terms of skill set is the opposite of what you are." We consulted and then he called Tom Morgan, then the CEO of a Fortune 500 firm, and said, "We need you." It's hard to tell Fred Smith no, so Morgan became our CEO and I took the role of chairman. It's his company to manage. I spend all of my time on strategic issues. It's a wonderful relationship.

Donlon: Isn't that the structure Ben Rosen had with Eckhard Pfeiffer? Winn: Eckhard Pfeiffer didn't strike me as capable of endearing himself to his team. There are different types of strategies in leadership. The person I picked has a servant leadership style, one of the most popular leadership styles today, and almost no ego. It works under that circumstance.

Donlon: So your model only works if the personalities - yours, the chairman's, and the CEO's - are of a certain type?

Winn: An entrepreneurial company is governed by its founder. So that person has to decide they genuinely want this. And the process works if you have different styles and if you decide who is responsible for what.

Arnie Pollard (CE): Your category of dot-com company is, more than any other category in the history of the public marketplace, selling dreams of the hereafter at a hell of a price. And the job of selling dot-com visions is a unique full-time CVO or chief visionary officer job, which is something you can't do if you also have to think about actually running the company. A firm like that needs a CVO freed up from being CEO.

Winn: Because you are my good friend, I can disrespectfully disagree. [Laughter] Because the world we occupy is driven by pioneers and everything that we do everyday has never been done before, having a CVO officer, if you will, is a valuable thing. But we must execute. We have 2,000 brand partners we have to ship from every day. Products have to be ordered, credit has to be processed, goods have to be shipped, and customers have to be taken care of. There is nothing virtual about that; there's a lot of tactical.

Martin Sperber (Schein Pharmaceutical): It's not only the dot-corn companies that are in this boat. My company was founded in 1985, but until we became a public company, we did not really have totally independent outside directors. After a 10-year tenure as chairman, CEO, and now looking at a plan to replace myself - at least as CEO, maybe not as chairman initially - over the last two years, the company doesn't quite fit the mold of turning it over to outside directors who can do that.

Part of my vision, which I hope will continue, is that the company will always have a balance between our original concept of providing affordable drugs - the generic drugs - and the branded sector, which generates the profit margins we need as a public company. But we're moving rapidly toward a 50-50 mix that will probably take place over the next five years, because that's where the profits are that the quarterly reports demand.

Larry Gloyd (Clarcor): We have a similar situation. I was scheduled to retire three years ago, but, even though we had the man chosen, we were in the process of making an acquisition that would double the size of our company. The board asked me to stay on for two more years, which has given us a chance to prepare my successor, who I think will do a wonderful job.

Donlon: Who did the choosing, the board or you?

Gloyd: Both. I hired him nine years ago, not to become CEO, but to come in and run one of our companies. The board asked me to identify people I felt had the wherewithal and he was one of three candidates I identified. We worked with an industrial psychologist and I gave all three exposure at board meetings. We evaluated them and one man came out head and shoulders above the other two. We identified him as the president and COO, which gave us a chance to really zero in on him.

That's why I'm questioning if I should stay on as the chairman when he becomes CEO. I don't want to be in his way. Most of my net worth is tied up in this firm, and I expect it to grow in the future, but I've got confidence in walking away.

Kangas: At a lot of private companies that are in transition, what really happens is the founder or the dominant shareholder becomes chairman, but I maintain they are still the CEOs. What they hire is not a CEO. They hire more than a COO, it's a chief management officer. The person you hired, Craig, is responsible for everything tactical, running the business, all management, but you're responsible for what's strategic. In most businesses, everything strategic is the CEO.

Ogden: The lack of process, which leads to the lack of interjected view, is what runs most closely held firms off the track. The successor will either come from inside the company from the board, which is often the case, or from outside, which is infrequent. And the process surrounding how you make that decision between internal candidate, executive candidate, trusted member of the board, or outside candidate determines whether or not it goes off the track. Having the owner-founder draft someone - from inside or outside - without a process is a recipe for disaster.

Neff: Another recipe for disaster is when no board members are involved. The CEO will then generally have the outcome that he or she wants, and probably have a continuing role - which might be the right thing. If there isn't at least one independent voice involved in addition to the CEO, then the outcome is predictable - and the pool of candidates that might be attracted could be limited as a result.

Then in some cases the board is not doing its job, or they waited too long, and have no choice but to fire the CEO, and then we get involved. The code words are "We need a new CEO, but deal with the CEO, and let's pretend we're finding a COO." Sometimes you have to go through that charade to get to the right conclusion.

THE DOOR OR THE CHAIR?

Kaufman: Should a former CEO stay on as chairman and, if so, how? I'm grappling with whether I have institutional history to bring, or will I act as a dead weight?

Bob Lear (CE): I've come to the painful conclusion that when the CEO retires he should retire from the board as well. He's served his time and he will cramp the style of the successor if he picked the right successor. You don't need the old CEO around. It's so hard to fire his old friend, get rid of his old product, wipe out his own acquisition, change the technique that he put in yesterday, and tighten the screws on his favorite customer, who he still plays golf with. There are so many things that change when you bring a new CEO in, particularly when you want the direction of the company changed. And usually you need to change the direction of the company.

Donlon: What about continuity?

Lear: If you plan properly, you've had three to eight years in which the CEO has been planning for his leaving. And you can get continuity from the CFO and the operating heads.

Neff: You have to take personalities of not just the CEO, but also the successor into account. Can you imagine Jack Welch staying on his board and what his replacement would be like? I'm sure he will get off. Then there's the issue of whether the successor is ready to take on the whole thing. Some need mentoring by someone, and the outgoing CEO is an obvious choice - although if they can't be mentored and brought up to full speed in a year, then you've probably got the wrong person.

Barry Naft (Environmental International): It's a function of the personality and the roles of the individuals. Will the chairman be the big brother? Will he look over his shoulder at operating functions? That will never work. The idea of continuity is overplayed. There are plenty of examples of guys coming in from the outside and they're extremely successful. A good guy should be able to pick that up quick enough.

Weston: If the new chairman can't stay completely out of operations, he ought to get off the board.

Kangas: It ought to be up to the new CEO.

Neff: When we're doing a search, that's often one of the conditions of acceptance.

Kaufman: When I became CEO, I was grateful that my predecessor stayed as chairman. When he stepped down after five years, I actually asked him not to because he took off my shoulders things that he could do as well or better, which freed me up.

Pollard: You have a unique situation where somebody who's like an extension of yourself can take part of it off your shoulders; that's the exception. If you've got that and can use it, you use it.

Neff: If the firm needs change, the person should definitely get off right away.

Copeland: I'm a little more positive about it, maybe just because of my own experience. But people will absolutely go around you to the former CEO, under the best of circumstances. And everything depends on what that former CEO does then. If he chases them right back to the CEO, you've probably got a relationship.

Kangas: That happens. But Jim and I are very good friends and very good partners. For a while it works very well, because I do not have the ego need or the power need that chief executives have; therefore I am quite comfortable with my new role. I am unusual in that regard, but that comes from the way you manage partnerships. As a result of that, I counsel people to go around Jim, and then I send them back to Jim. I make no decisions. This is terrific! [Laughter]

Keith Morgan (AAMCO): There are two other important decisions relevant for closely held firms. One is estate planning in terms of the ownership of stock, which has interest in terms of tax considerations, but also in terms of control. A lot of founders don't want to address these issues, and some of their kids are in the business, some are not in the business, and it becomes complicated. If the founder doesn't discuss this plan with the adult children and get them to buy in, you get conflicts between personal financial interests and what's in the best interest of all shareholders.

Donlon: Do you have siblings?

Morgan: I do. I'm fortunate that one of my brothers is an oncologist, and another is a college coach. Although we grew up thinking we were going to take over the business, they went on different paths. And my father did do wise planning, and I believe they recognize that I am the most qualified professional manager on their behalf as shareholders of the company. My father also recognized that I should have voting control, and that is in their best interests, which may have been difficult initially for them to recognize. [Laughter]

Weston: There isn't any one model that will survive this lunch. Among privately owned companies, you've got two subsets: those where the founder already has family members involved who he or she wants to keep involved, and those that are privately run that don't have family members involved. I've been on the boards of both kinds. A privately run company would be well advised to get some outside board members and listen to them.

With public companies, there are some entrepreneurial ones that just became public and still don't know how to make that transition. The primer for them is totally different from the primer for GE, IBM, or GM. Then there are principles that transcend all of this. Like how do you think about the next decade versus the last decade? And what is the role of a non-executive chairman?

Donlon: The non-executive chairman is common in the U.K. and elsewhere, and was not uncommon in the U.S. until the post-war era.

Neff: Every situation is somewhat unique. We talked about different types of companies, but there are also different personalities. We've got to factor in the personality of the sitting CEO and the relationship with the board. What works for GE is not going to work for GM. I don't think there is any one model. I agree that an independent perspective is critical, be that out of the wisdom of a Fred Smith, or the lead director or non- executive chairman. I also think there is a role for a non-executive chairman in certain companies, particularly companies in transition.

About 15 percent of public firms in the U.S. have a chairman who is not the CEO. But most of them are in transition from the founder to the next generation, or have a new CEO who needs a mentor. For example, companies like Campbell Soup, where the family owned 67 percent of the stock, are entitled to have a non-executive chairman, which they did for years. We actually have assignments to recruit non-executive chairmen, particularly for spinoffs.

Kaufman: In the U.K., ex-CEOs become non-executive chairmen, often in two or three companies simultaneously.

Neff: That works for them, but we've found it's awfully tough to recruit an American into a British company under that model. That one-person model has worked in the U.S. for a long time, so why change it?

Lear: British companies have a dominant majority of inside directors, and that's to give them the power, by having a nonexecutive chairman.

Donlon: Ed and Jim, I want to get your reaction to the Fannie Mae model. Jim Johnson argues that rather than leave to the successor something that's contentious or messy, the outgoing CEO or chairman attends to it. Is there validity to that?

Kangas: Of the few things that I didn't get done, the reason was that I wasn't equipped to get them done, didn't want to get them done, or was afraid to face them. So why try to do something at the very end that I couldn't get done in the course of 10 years? He'd get them done a lot faster than I ever would.

The more valid reason for staying for a while in my case is to play off the positives. There's a lot of angst in our current process of going from a federation to integrated management, financially and managerially. And there are a lot of people that have got to take a very difficult trip. I have some relationships and can help some people with that comfort level to get to where we need to be in a year or so.

Copeland: There was more of a workload issue from our standpoint. We were undertaking enormous change at the same time we were trying to manage our operations as a global business for the first time, actually. I trust Ed absolutely, so it's great to have him around at this point in time to deal with those tidal waves of change in activity.

Ogden: One of the most fascinating things about the GE situation is the combination of the well-publicized horse race and Jack's intention to get out as soon as he's appointed his successor, or shortly thereafter. Because no matter who's picked by the board and Jack, there will be 14 potential CEOs of other companies running the key businesses.

That's a mess somebody will have to clean up. Jack has such incredible personal relationships with the top group, it strikes me that he could do a great service by helping his successor clean up that mess. Because there really could be quite a drain of brain power.

Donlon: How is the process this time around different from the process that selected Jack when Reg Jones was leaving?

Naft: When I was in Wharton grad school in the mid-'80s, Reg Jones spent a weekend with us. He said the minute he looked at the succession committee's list he realized his mistake. This was a committee of guys loyal to him, and they picked a list of six guys who were just like him. And he said, "Absolutely not what I need." He added one name, which was Jack Welch. But he said the tragedy of his mistake was what it did to the careers of those six guys that were on that list.

Donlon: They all left, didn't they?

Naft: Yes, and he wished he could have gone back and done it differently.

Tom Haggai: (IGA) I told our board members I don't want any input, because, contrary to their thinking, they need someone totally different. In my firm, I think the next person will definitely come from the outside. I may have a vested interest, because I own stock, but it's their responsibility. That's why they're on the board. If they can't do it, they shouldn't be on the board, it's that simple.

One of the things that's encouraging, even though it's played down a bit, is the [evolution of the] COO role. We now realize that the idea of a COO and a CEO is out, but we should make a professional out of the COO. One of the best people I've ever known was a COO of three different companies. And I said, "Did you ever regret that you're never a CEO?" He said, "No. When I was in school my favorite professor said, 'Remember, Bill, the better second fiddle you play, the better first fiddle player you sit beside.' "And he said, "I know I'm the second fiddle player, and I don't get blamed for the sour notes. I'm not ready to take that blame."

Sperber: Not every COO is equipped to move up in terms of the strategic planning that's needed. You need the meat-and-potatoes individual who can handle the implementation and the day-to-day responsibilities of somebody's strategic vision. And the person who is doing the meat and potatoes may not want to or may not be equipped to be the visionary to lead the company.

I've played the role of COO, CEO, and chairman over the last two years. And that has not done my company the service that it needed, but we were trying to effect a transition and it's an ongoing process. I'm looking for, or will be looking for, a CEO who's got the vision, can plan strategically for where the company has to be, and knows how to implement. But he won't be designated as needing the implementation tools. That has to be a COO.

Kangas: Companies your size often aren't quite large enough to groom a CEO, so have to look outside or to their boards.

Sperber: True. In 10 years we went from $50 million to $500 million. That's not dot-com stock growth, but in our business it's significant. As we grew we really didn't have the ability to have the infrastructure and do succession planning.

Donlon: Implicit in our conversation is that the preferred candidate should come from within, all things being equal.

Sperber: If you have the luxury and infrastructure to groom your successor over a two- or three-year period, you're much better off. But at this point in our growth, somebody from the outside might serve the company's future better.

Donlon: Is it that one comes from outside if something internally didn't work?

Neff: I wouldn't assume that anymore. On most CEO assignments that we do, we're benchmarking inside versus outside. We interview and evaluate insiders.

Pollard: More so than 10 years ago?

Neff: Yes, without question.

Ogden: There are firms that do a great job of producing large stocks of general managers. They produce great functional talent, great CFOs, great marketing people, great technical people. But very few - there's GE, obviously - produce a stockpile of proven good general managers.

Donlon: Where will the next generation of CEOs come from?

Kaufman: I'll say a lot of them will be from the other gender. It's interesting that we've used male pronouns. No one has used the indefinite pronoun "they" or "her."

Donlon: We've got a Fortune 100 CEO now at H-P. Is this the wave of the future or an anomoly?

Neff: It's just a matter of time. I think more will also come from outside the U.S. also. They may be from within the company, but they'll be non-Americans. There are a number of examples right here.

Weston: Sometimes after you go to the outside, you think, "Well, that person isn't going to land running, and I need somebody who's going to land running. And I have somebody on the inside who I really didn't appreciate enough until I saw what the outside brought in."

Copeland: You are looking to remove yourself enough from today, to have some perspective on what the future requirements will be, and to decide what characteristics the person who will carry the organization forward should have. If there is a common denominator for all of the companies around this table, it's how to search for objective perspective in this decision.

Ogden: The H-P assignment was fascinating, because H-P is one of the small number of companies who have traditionally been credited with doing a good job developing internal management and giving people responsibility for big chunks of business on an international scale. They had a woman, Ann Livermore, running a $20 billion business and they picked a fantastic woman from Lucent.

Neff: HP developed terrific executives, but most of them left the company.

Copeland: A chicken-or-egg question we didn't answer was what comes first, the strategy or the CEO? If you say, "We've got a great strategy, we want a CEO to drive the strategy," you're talking about wanting a visionary. If you get a visionary, he or she will want to drive the strategy.

Neff: It's strategy first, unless you have a situation where the strategy has failed, the CEO has been thrown out, and the board is not about to develop the strategy. In that case, you've got to bring in someone who can develop the strategy and implement it. But if you have a company that's doing well, I don't think you're looking for somebody to come in and radically change it.

Neff: I think you will see a growing number of unconventional choices, especially in rapidly changing businesses. They're looking for an agent for change; an explorer. That's when you see unconventional choices like the one H-P made.

Ed Drudge (Personnel Group of America): There's also the issue of whether that new CEO has to have a philosophic compatibility to the firm's history and foundation.

Donlon: How important is character?

Winn: Character counts. Experience is the easiest thing to look for, but personality and style are critical to us. Intelligence is also critical. I'd say it's character, attitude, intelligence, and lastly experience.

Kaufman: I like to get past the first barrier. All of the people coming to us, including the guy we hired to be COO and the next CEO, are interviewed by five or six other senior people, some of whom have a very different lens into personalities. Most anyone can keep up a front for two hours. With more than five meetings of two or three hours each, it's harder to maintain a front if it's not real.

Shalam: Besides the qualities that have been described, the new CEO should be able to integrate smoothly. We have a small organization, and the last thing we want to do is to bring in a new CEO who's going to disrupt things and create distrust among the other executives.

Carol Evans ICE): Arthur Levitt told me that when he was making decisions like this, he would imagine himself walking down a dark alley, and then pick the person that he would like to have beside him.

Donlon: And then watch where he walks. [Laughter]

Winn: The challenge is to replace management as opposed to strategy. If the strategy is sound, then the most important thing the outgoing CEO can do is to desire the change to be made, and to empower the new person. If strategy is the issue, then the most important thing the old CEO can do is get out of the way.

Copeland: The CEO has a responsibility for the development of internal candidates. The CEO also needs to have enough humility to get out of the way in the selection process, and let the board make an independent decision. And then the CEO has a responsibility for a successful transition to new leadership.

A WHO'S WHO OF ROUNDTABLE PARTICIPANTS

James E. Copeland Jr. is chief executive of Deloitte Touche Tohmatsu, a $9 billion global professional services firm, and its U.S. practice, Deloitte & Touche.

Edward P. Drudge Jr. is chairman and chief executive of Charlotte, NC-based Personnel Group of America, a $783.9 million provider of information technology and personnel staffing services to businesses, professional, and government organizations.

Mario Giacone Jr. is co-chairman and president of East Rutherford, NJ-based MicroAge Computing Concepts, a systems integrator providing network and systems technologies.

Lawrence E. Gloyd is chief executive of Rockford, IL-based CLARCOR, a $427 million manufacturer of filters, air-cleaning systems, and consumer product packaging.

Thomas S. Haggai is chairman and chief executive of Chicago-based IGA, a supermarket group with annual sales in excess of $19 billion

Edward A. Kangas is chairman of Deloitte Touche Tohmatsu, a $9 billion global professional services firm.

Stephen P. Kaufman is chairman, president, and chief executive of Melville, NY-based Arrow Electronics, an $8.34 billion distributor of electronic components and computer products.

Keith R. Morgan is chief executive of Bala Cynwyd, PA-based AAMCO Transmissions, a $425 million automotive services chain.

Barry N. Naft is president and chief executive of Potomac, MD-based Environmental International, a private company which manufactures products from recycled waste materials.

Thomas J. Neff is chairman, U.S., of Chicago-based Spencer Stuart, a $257 million international consulting firm focusing on senior-level executive searches and board director appointments.

Dayton Ogden is co-chairman of Chicago-based Spencer Stuart, a $257 million international consulting firm focusing on senior-level executive searches and board director appointments.

John Shalam is chairman, president, and chief executive of Hauppauge, NY-based Audiovox, a $616.7 million marketer of electronic parts and equipment.

Martin Sperber is chairman, president, and chief executive of Florham Park, NY-based Schein Pharmaceutical, a $523.2 million developer and marketer of generic pharmaceuticals.

Richard Vissers is president and chief executive of Wilmington, DE-based CTS Corp., a $500 million software company specializing in bulk fare, business-to-business travel.

Josh Weston is honorary chairman of Roseland, NJ-based Automatic Data Processing, a $5.54 billion computing services firm providing human resource management systems and benefits and payroll processing.

Craig Winn is chairman of Charlottesville, VA-based Value America, an Internet-based retailer.

RELATED ARTICLE: DANA CORP: SUCCEEDING AT SUCCESSION

Earlier this year, Southwood J. "Woody" Morcott stepped down as CEO of autoparts maker Dana Corp., while another company veteran, Joseph Magliochetti, took his place. Next year, Morcott will relinquish the chairman's title to Magliochetti as well. And if all goes accordingly, their 86,000 co-workers will hardly notice. "We're trying to make succession boring," quips Morcott, 61, an affable Georgia native who became CEO in 1989 and chairman the following year.

With many CEOs being shown the oak door before their time, or chasing career options in a quest for more lucrative stock options, corporate succession planning has never been so integral to smooth business operations. Toledo, OH-based Dana frankly hasn't had much experience with top-level succession - Morcott is only the fifth chairman since Charles Dana took the company public in 1916 - but it does have a highly organized company-wide strategy. "Succession at all levels is a process, not an event," Morcott explains. "In 1989, when i became CEO, I started planning for my succession. When the event occurs, you don't go scrambling around looking. That should all be done in advance."

Managers of each Dana business division create what Morcott calls "depth charts" with the names of three or more potential replacements that the board reviews each December. These snapshots highlight an employee's strengths, weaknesses, and corporate experience. People on the depth chart know who they are, and their supervisors work with them to build their capabilities.

Additionally, Dana's five-member policy committee has its own depth charts to track the top few dozen division managers and other high-placed operatives. Dana tends to promote from within, and gives continuous training, education, and managerial challenges that are designed to eventually produce several qualified candidates for the top job. Morcott, for instance, joined Dana in 1963; Magliochetti arrived three years later.

"Because we promote from within, we tend not to lose as many people," Morcott says. "They know nobody's going to be parachuted in above them. It's up to them to win it or lose it"

The succession system is so ingrained that Dana already is looking for people who can fill Magliochetti's shoes when the 57-year-old retires in about eight years - and beyond. "We're talking about who will be chairman in 2015," says Morcott. "I know the names of the players who have the potential if we move them around, train them, teach them, educate them."

There's also another name in the top circle, which Morcott genially refers to as his "hit by a truck" replacement, who would be able to take the reins in dire straits. For Morcott's first two years, that task went to his immediate predecessor, Gerry Mitchell. Morcott himself became Dana president in 1986 to fill a vacancy after a sudden death, setting the stage for his future role. "You have to tailor everything to circumstances," he advises.

Open communication and feedback are linchpins of Dana's succession process. Morcott insists that he's opposed to executive horse races, where intense personal rivalries can undermine teamwork. "Dana hasn't functioned that way," he asserts, noting that naming Magliochetti CEO made it clear to the entire organization that he also would become chairman. "So you take away all discussion," Morcott adds, "and you get on with preparing him, positioning him, and letting him absorb the job in a more orderly fashion than just dropping everything on top of him." At the same time, the process tends to be self-selecting. A hotshot executive who's big on titles and has difficulty conceding power would probably not feel comfortable around Dana's nonthreatening, unemotional succession philosophy.

Has Dana found the solution to poor succession planning? Morcott, who serves on three other boards, cautions that there's no perfect way. "We don't have the answer," he says. "We have the answer for us."

- Jonathan Burton