Comparing the Performance of External Successors

When Mattel's directors went shopping for a new CEO, they found their man in processed cheese. When IBM's board needed to replace John Akers, it picked an executive who had worked in financial services and cigarettes. Were these companies going out on a limb by opting for outsiders who knew little about toys or computers? Yes and no, according to a 2001 working paper, “External Successors, Transferable Skills and Firm Performance,” that will be published in a forthcoming issue of Managerial and Decision Economics.

Authors Elizabeth E. Bailey, John C. Hower Professor of Business and Public Policy at the Wharton School, and Constance E. Helfat, associate professor of business administration at Dartmouth's Tuck School of Business, analyze and compare the performance of three groups of externally hired CEOs. One group is composed of executives who came to their new jobs from completely different industries and brought only generic skills, such as the ability to lead. IBM chief Lou Gerstner is such an example. Executives who took the helm already having experience in the same industry and the ability to transfer industry-specific skills to their new roles constitute the second group. Robert Morrison, who moved from Kraft Foods to Quaker Oats, falls into this category.

A third set of executives, including Mattel CEO Robert Eckert, also a Kraft veteran, lies between the first two. These CEOs' portfolios include what Bailey and Helfat call “related-industry” skills. Building on work by Carnegie Mellon professor of business strategy Jeffrey R. Williams, the authors define related industries as sectors that produce different products but share similar challenges and require similar capabilities. For example, slow-cycle industries, such as pharmaceuticals, typically value executives who are skilled at building long-term relationships. Fast-cycle industries, such as semiconductors, put a premium on the ability to speed new products to market. And in standard-cycle industries, which range from fast food to cars, successful executives learn to build brand loyalty and manage economies of scope and scale.

Do new CEOs with only generic or related-industry transferable skills perform as well as external successors with industry-specific experience? In order to answer this question, Bailey and Helfat collected data on the 300-plus CEOs who appeared in Forbes's annual surveys of executive compensation from 1978 through 1987. This process identified 36 external successors, defined as top executives who had been with the firm for no more than two years before becoming CEO. Of this group, nearly 60% had recent experience in the same industry as the hiring company, whereas 40% had only generic or, more commonly, related-industry transferable skills. The researchers then constructed a matching sample of firms that had internal successors throughout this period and analyzed the external successors' performance relative to these controls.

The results? Mean performance, measured as the difference between a company's presuccession and postsuccession return on assets, less the change in ROA for control companies over the same period, was the same for both groups of external successors for up to three years following a CEO's appointment. (Beyond that point, at least one-third of the external successors had lost their jobs, potentially biasing further analysis.) However, the variability of company performance was considerably higher among CEOs with only generic or related-industry transferable skills. The interquartile range — the range of values covered by the middle 50% of observations — was roughly six times as large for this group as it was for CEOs with industry-specific experience.

“We were really surprised to find you couldn't see much difference in performance,” says Bailey, who explains this finding by suggesting that companies hire different types of CEOs because they have different needs. Some firms require CEOs with strong relationships within their industry, while others need innovative leaders who can shake things up. If boards in both situations are equally effective at choosing CEOs that match their companies' needs, there is no reason to expect performance, on average, to depend on the extent of a CEO's transferable skills.

But even when a board has correctly diagnosed the need for an industry outsider, it still faces greater risk than companies that look for a track record within the industry. In part, says Bailey, the reason may be that directors generally have less information about candidates from outside their industry. However, it may also be that industry outsiders simply do what they were hired to do: They bring a fresh perspective to the job, which encourages them to attempt greater strategic and organizational change. As a result, sometimes they deliver stellar performance; sometimes they fall far below the mean.

Despite this risk, companies are increasingly looking at a larger pool when hiring a CEO. Over the past five or ten years, “there's been a discernible change in openness to talent from outside the industry,” says Tom Neff, U.S. chairman of executive search firm Spencer Stuart. Many companies already have people in place to handle their day-today operations, he observes, “but what is missing is someone who can provide overall leadership and develop the vision and a competitive strategic plan.”

Neff counsels boards looking for that ideal candidate to start close to home. “You work your way out from the industry to related industries that might serve the same markets or use the same distribution channels or have similar technologies,” he explains. Ultimately, however, industry or even related-industry experience shouldn't be the deciding factor when it comes to choosing a CEO. “I personally would prefer a star from a related industry or a super athlete from farther afield to an average player from the industry,” says Neff. “A good leader can ensure he is surrounded by the right resources to do what needs to be done.”