Companies Are Running Out of CEOs Who Have Already Failed Before
The statistic seems technical until you understand what it is actually measuring: the average executive who takes the helm of a public company today is 55 years old, seven years older than in 2000. A study by the National Bureau of Economic Research that analysed more than 50,000 executive career trajectories arrived at that figure and explained it with an apparently reasonable logic. Boards of directors, confronted with greater economic uncertainty and growing regulatory and geographic complexity, prefer to bet on executives with broad, accumulated experience over executives with unproven talent. The problem is that this pursuit of safety is inadvertently building a long-term structural trap.
Because what that number does not say directly is this: 86% of incoming CEOs in 2025 had never previously held the role at a public company. And the average tenure of an outgoing CEO fell from 8.3 years in 2021 to 7.1 years in 2025. Two trends that, read together, describe a market that is consuming its leadership resources faster than it is replenishing them.
---
The Scissors Effect Nobody Wanted to Name
Boards of directors are doing something paradoxical: they seek experience, yet they hire primarily those who do not have it — at least not the kind they are looking for. And those who do have it are left in the role for less time. The mathematical result of that combination is a contraction of the pool of CEOs who have led an organisation through complete transformation cycles.
This is not an abstract succession problem. It is a problem of institutional capital being squandered faster than it can be accumulated.
Executive recruiting firm Russell Reynolds Associates, which tracks CEO movements across 13 global stock indices, documented 11 mandates that lasted less than a year in 2025. That speaks to something more than bad luck or personality clashes. It speaks to selection processes that are not properly calibrating the gap between what a board expects from a CEO and what that CEO can actually deliver within the available time. Eleven interrupted mandates in a single year is not statistical noise: it is a signal that the conditions of the role are changing faster than the criteria used to choose who fills it.
The logic of boards has its apparent rationality: if there are no candidates who have previously run a public company, the next best candidate is someone who led a critical function in depth — operations, finance, talent. But governing a function, even at the highest level, is not the same as bearing full responsibility for capital allocation, organisational culture, market expectations and long-term strategy, all at the same time. The gap between those two types of experience is not closed by seniority or by age.
---
What Artificial Intelligence Is Making Visible
The artificial intelligence context adds a layer of pressure that did not exist with the same intensity in previous cycles of technological transformation. Not because AI is qualitatively different from other disruptions — it is, but that is not what matters most here — but because the pace at which it forces organisations to rethink operating and revenue models is considerably shorter than the average CEO tenure.
If an executive steps into that role for the first time today and has a statistical margin of 7.1 years before being replaced, their window for learning the organisation's patterns, articulating a coherent transformation strategy and executing it with measurable results is narrow. Narrow enough that, in the best case, they manage to set the direction but do not harvest the results. In the worst case, the organisation ends the mandate without having completed the pivot.
Someone who has already been through that process at another company — with their mistakes, their adjustments, their learning about which decisions cost more than expected — does not need that initial calibration time. They arrive already knowing where the organisational bottlenecks are that never show up in reports. They know that resistance to transformation rarely comes from the market: it comes from the internal incentive structure.
But that profile — the CEO who has already done it, learned from it and is willing to do it again — is disappearing as an available option. Not because it does not exist, but because the current design of the role structurally disincentivises repetition. Tenures are shortening, pressure is intensifying and, for those with an established reputation, the risk-benefit calculation of taking on a second or third role becomes less favourable with each passing year.
---
The Talent Pipeline Model Companies Abandoned Without Realising It
For decades, the logic of leadership development in large companies operated on an implicit architecture: executives accumulated progressive exposure to decisions of increasing consequence. Cross-functional assignments, high-risk projects with board visibility, operations in difficult markets — all of this functioned as a testing system in which candidates demonstrated the capacity for decision-making under uncertainty before reaching the top role.
That model was never perfect, but it had a concrete operational virtue: it produced candidates who had failed in controlled environments before failing where the cost is systemic. The relevant question for any board should not be whether the candidate has sufficient functional experience, but whether they carry specific scars from having managed a crisis they did not ask for, a strategic pivot that met internal resistance, or a restructuring that cost more than budgeted.
What the Russell Reynolds Associates data suggests is that this system of progressive formation is producing fewer prepared candidates than the market needs, and that the dominant response — hiring older candidates with deep functional track records — is not solving the problem but deferring it. More age does not automatically equal a greater diversity of situations managed. In many cases, it equals more years doing one thing very well.
The profile the moment demands — proven adaptability across different organisational contexts, not depth in just one — is not formed by waiting. It is formed by deliberately designing conditions in which high-potential executives face decisions that force them to reason outside their domain of expertise, are accountable to parties with divergent interests and learn to operate under constraints they do not control.
That is not a leadership development programme. It is deciding what kind of real work, with real consequences, is assigned to someone five years before they become a CEO candidate.
---
Succession as an Operational Function, Not a Contingency
The true cost of this problem does not appear in the income statement for the year in which the CEO change takes place. It appears in the two or three years that follow, when the organisation absorbs the learning curve of the new leader, renegotiates its strategic priorities and recalibrates market expectations. That period of friction carries a price that standard valuation models rarely capture with precision, because it is not labelled as a "succession cost."
The reduction in average tenure — from 8.3 to 7.1 years in four years — means that this friction cycle is being triggered more frequently. And if the observed trend of tenures not exceeding three years continues, the accumulation of those costs begins to affect organisations' ability to execute strategies whose horizons extend beyond the average tenure of the person who designed them.
What this implies for boards of directors is not only a need to revisit the selection process. It means redefining succession as an operational function with its own metrics, assigned budget and a five-to-eight-year horizon — not as a contingency plan activated when a vacancy arises. Because when there is a vacancy and no deliberately built pipeline exists, the board chooses from candidates who are available, not from those it would have needed.
The executive talent market will not spontaneously produce more CEOs with multiple career trajectories if the design of the role continues to make a second or third assignment carry a personal and reputational cost that few are willing to pay. The companies that understand this first will change the structure of the problem, not merely the criteria for the next search process.
