Why Doesn't CEO Succession Work?

These issues dominated a panel discussion on succession at the Agenda 07 Forum. Throwing down the gauntlet was Jim Benson, former CEO of John Hancock Life Insurance and now CEO of Clark Benson, a unit of Clark Consulting. "I actually don't think there's much seriousness at all, despite the rhetoric about succession planning," argued Benson, who also has held top positions at MetLife and New England Financial. "In each case, I thought there was lip service paid to the whole idea," he said. "Frankly, as a CEO, what you want is to have some very highly competent technicians around you, but slightly flawed, so that there was always something slightly askew with all of the internal candidates and you could not ever be truly replaced."

Tom Plaskett, who is nonexecutive chairman of Novell and who also serves on the board of RadioShack, both of which have had CEO transitions this year, argued that boards simply aren't spending enough time on succession. "With the average tenure for CEOs today of 5.3 years, it's probably a topic that should be on every board's agenda at every meeting," said Plaskett. "I think it is viewed in many places as a rather simplistic process that you do once a year. A nice book is put together by the CEO, and you look at the pictures of the key executives and you talk about perhaps a couple of options. And then you revisit it a year later."

The transition at Novell went smoothly, but Plaskett said directors at RadioShack were surprised. "It began as an orderly transition with an enlightened CEO, who desired to retire early and who had been developing a successor for 10 years," Plaskett recounted. The succession process was reviewed annually by the board and updated regularly. Everything seemed to be right.

But then after the new CEO took over, a credibility issue emerged—there was false information on his resume. "When it comes down to a point of a representation to the board that this information is accurate and the board takes the position of standing behind the CEO, and then the next day we find out that it's not true, that's it," Plaskett said. Process, it seems, does not always work perfectly.

A mistake that some boards make is limiting the responsibility for CEO succession planning to a single committee, said Tom Clarke, CEO of TheStreet.com. "I think we have to involve the full board," he said. "I think it's too difficult to lay this on the compensation committee, which is what typically happens. I think this is a full board involvement because there are different criteria that a lot of board members want to bring to the equation."

Defining those criteria is the heart of the dilemma. "It's not as simple as saying 'Here are the metrics and deliver on the metrics,'" Clarke said. "It's important that you get the full board to opine on the issues."

The sharpest dispute among panel members revolved around term limits for CEOs. Benson and Steve Mader, vice chairman of the board practice at Christian & Timbers, both argued that a term limit of perhaps five years is reasonable. Both Plaskett and Clarke opposed the notion because it would turn a CEO into a "lame duck" and prevent him or her from being effective. "The lame duck issue is really a great point," Mader argued, "but I believe the board should embrace the philosophy that there is an effective length of time, and that should be an ongoing discussion at all times with the CEO." Boards, he added, should time a CEO's departure to be best for the business, not best for the individual.

That issue aside, statistics show that fewer than 50 percent of boards in the Standard & Poor's 500 are "very confident" about their succession plans, said Fred Steingraber, chairman of Board Advisors.