When Is an Outsider
CEO a Good Choice?
On average, CEOs recruited from outside
the company perform about the same as
those who come up through the ranks, the
authors’ research suggests. But there are
certain circumstances in which outsider
CEOs tend to do better.
BY AYSE KARAEVLI AND EDWARD J. ZAJAC
Former IBM CEO Lou Gerstner is an example of a high- profile outsider CEO who became an effective change agent.
In recent years, established corporations such as Siemens, Hershey, 3M and Wrigley recruited new CEOs from outside the
company for the first time in more than a century. Both General
Motors and Hewlett-Packard have had more than one CEO who
was brought in from outside.
But when is an outsider CEO the right choice? Outsider CEOs
such as William Perez of Nike and Jeff Nugent of Revlon were
hired by boards to lead their companies through major transformations but were not retained by the same boards through a
two-year post-succession period. This suggests the possible existence of a paradox in which boards choose outsider CEOs with a presumed mandate for
change but soon become dissatisfied with those same outsider CEOs.
We seek to elaborate and resolve this paradox of outsider CEO succession. Our research suggests that, on average, outsider CEOs are neither better nor worse than insider
successors within a three-year succession period. However, we also found that new external CEOs outperform insiders under certain pre- and post-succession circumstances.
Our results suggest that boards of directors can increase the probability of successful
external CEO hiring if they choose such CEOs under the conditions of poor performance and/or high industry growth. Furthermore, external CEOs outperform insider
successors when they replace the company’s senior management team with new executives. Our research also suggests that — contrary to conventional wisdom — company
performance usually suffers when new outsider CEOs rush to make strategic changes in
the early post-succession period.
We conducted a detailed empirical investigation of the performance and strategic change
consequences of CEO succession in 90 single-business organizations over 30 years (1972-
2002), using two contrasting environmental situations in terms of turbulence and growth/
munificence: the U.S. airline and chemical industries. The sample included all public com-
panies that reached $100 million in sales at any point between 1972 and 2002. We observed
163 successions. Some incompleteness in available data resulted in a final sample that in-
cluded 140 succession observations, with 80 insider successions and 60 outsider successions.
(We define outsider successions as successors who became CEO within two years of being
hired from outside the company.) As part of
the research, we also conducted a meta-
review of five decades of published
empirical research (1954-2002) on the con-
sequences of CEO succession events.
COURTESY OF IBM
SUMMER 2012 MIT SLOAN MANAGEMENT REVIEW 15