The Blind Spot in the Most Studied CEO Succession Process Ever Conducted
In the late 1990s, General Electric ran what became the most studied CEO succession process in corporate history. Jack Welch devoted six years to it. He narrowed a field of internal candidates to three finalists and evaluated them across business cycles, through crises, and under competitive pressure. The board was deeply engaged, as the governance community watched closely. When Jeff Immelt was named successor in 2001, the process was heralded, before the outcome was even known, as a gold standard every company should emulate.
Welch himself later said: "If you want to pin failure on me, I missed it."
Immelt's tenure ultimately coincided with a dramatic decline in GE's market value and years of strategic restructuring, leading many observers to revisit what had once been celebrated as the model succession process.
The gap no one measured
The GE board had all the information it required, and the directors had spent years observing the finalists. They had data on business performance, leadership assessments, peer evaluations, and Welch's own considered judgment. The governance structure was sound and the commitment genuine.
The problem was that no governance tool existed to measure whether board confidence aligned with organizational reality. The board was not missing information; it was missing visibility into whether its collective assumptions reflected organizational reality. Some accounts of the succession process suggest that not every director shared Welch’s final assessment of the candidates. Because the process relied heavily on the departing CEO's judgment, rather than explicitly testing the board's collective assumptions about what the organization would need next, those differing perceptions never emerged as a measurable governance signal.
The board's confidence in the process gradually became confidence in the decision. Those are not the same thing.
Confidence in the process confirmed the process. It revealed nothing of whether the board's collective perception of the right candidate matched the organizational reality that would follow.
This is the successor version of the Thin Ice Problem, which I had previously written about. The surface looked solid because the process was rigorous. The confidence was real, but beneath it, the assumptions that were going untested would only be validated under the weight of the actual transition.
Succession is not unique. It is simply vivid.
GE is an unusually visible example in scale, but the dynamic it illustrates of board confidence diverging from organizational reality, with no mechanism to surface the gap, appears with uncomfortable frequency across domains and company sizes.
Recent data from Heidrick & Struggles quantifies what many directors already sense. Boards are fourteen to sixteen percentage points more optimistic than C-suite executives when assessing confidence in CEO succession planning. The board believes the pipeline is stronger than management does, and the board believes the process is more rigorous than management experiences it to be.
That gap is evidence that perception itself can become a governance risk. Two groups are evaluating the same organization, reaching materially different conclusions, and neither knows the divergence exists. The question is whether your board's confidence in its succession readiness reflects the reality your leadership team is living. If a gap existed, what would it take to surface it?
Why this matters beyond succession
Succession is a useful lens precisely because the stakes are high and the timeline is long. The gap between board perception and organizational reality has years to accumulate before a transition forces it into view.
But succession is not the only domain where this dynamic operates. There are many: cybersecurity, AI, culture, strategic execution, and capital allocation. Each of these areas, boards are forming perceptions of organizational capability, and the management teams are living a different version of that reality, and that drift is quietly increasing.
The supply chain article that prompted me to think about this series made the case that boards need to pay closer attention to operational resilience. That argument is correct. However, the deeper question, which the GE succession story makes impossible to ignore, is whether attention is enough.
Because if the board's perception of an organization's capability can diverge from reality even after six years of rigorous process, the question is no longer about effort or engagement. It is about what governance tools are designed to measure, and what they miss.
What boards should measure
Most governance tools are designed to measure process. They assess whether succession plans exist, candidates have been evaluated, meetings have occurred, and oversight responsibilities have been fulfilled. Far fewer tools measure something different:
whether directors hold materially different views of organizational readiness,
whether management's assessment diverges from the board's,
whether confidence is supported by evidence,
or whether governance drift is increasing beneath otherwise healthy performance.
These perception gaps are often invisible until a leadership transition, strategic failure, or external shock forces them into the open. By then, the opportunity to surface them early has already passed. Governance failures rarely begin with a lack of oversight. More often, they begin when confidence quietly outpaces evidence, and no one is measuring the distance between the two.