The succession conversations that get scheduled are the comfortable ones. The chief executive will retire in three years; there is time to groom a successor, stage the handover, hold the retirement dinner. What almost never gets rehearsed is the other kind: the leader who collapses on a factory floor, resigns under investigation on a Friday, or does not come back from a weekend. That is the succession that actually tests an organization, and it is the one nobody wants to imagine.
Ricardo Semler has told his own version of the sudden exit. At twenty-five he collapsed in a factory and was told by a doctor that he was the healthiest sick person the doctor had ever seen, that nothing was wrong except the way he was living. That afternoon he began dismantling the machine that had produced the collapse. His was a warning rather than a departure, but it carries the point exactly: the crisis does not schedule itself, and the organization that has quietly made itself dependent on one person's presence is the one least able to absorb that person's sudden absence.
The shock is real even when the exit is planned
Start from what the evidence establishes about ordinary, orderly succession. A meta-analysis by Schepker and colleagues, covering 13,578 CEO successions, found that a change of leader depresses short-term performance and has no reliable direct long-term effect. Even the smooth handover, with a named successor and a farewell tour, costs something while the organization re-forms around a new center.
Now remove the preparation. An emergency succession is the same shock with none of the cushioning: no transferred relationships, no staged authority, no time for the successor to be tested before the weight lands. The short-term cost the meta-analysis describes is the floor, not the expectation, for a transition nobody got ready for.
Preparation is the variable
The reason to rehearse is that preparation demonstrably changes the outcome. Zhang and Rajagopalan, studying 204 successions, found that relay successions, where an heir is designated and genuinely prepared before taking over, led to better post-succession performance, with the largest advantage precisely when performance had been weak and the environment was unstable. Instability is the defining feature of an emergency. The relay finding says that readiness pays off most in exactly the conditions an unplanned exit creates.
The uncomfortable implication is that emergency succession cannot be improvised well. The plan that works in a crisis is the one that was built before the crisis: someone who already holds real authority in part, already knows the critical relationships, already understands the decisions on the table. You cannot manufacture that on the Friday afternoon the news breaks.
What a real emergency plan contains
A plan nobody rehearses is a name on a page. A plan that would survive contact with a real emergency looks different. It identifies not one heir but the two or three people who could hold the organization together for ninety days, because the sudden vacancy is rarely the moment to pick the permanent leader. It ensures those people already have access to the relationships, the numbers, and the decisions, rather than discovering them cold. It clarifies who decides, who communicates, and who holds authority in the interval before a permanent choice is made. And it is tested, at least on paper, against the specific question everyone avoids: if the leader disappeared tonight, what breaks first, and who catches it?
There is also a communication dimension that improvisation handles badly. In a sudden vacancy, the people who most need steadiness, employees, customers, lenders, are the ones most rattled, and a leadership shock the meta-analysis already links to short-term performance loss is made worse by a vacuum of information. A rehearsed plan settles in advance who speaks, to whom, and how fast, so that the organization projects control it has actually prepared rather than control it is faking. The absence of that is why unplanned exits so often spiral from a personnel event into a confidence event.
The organizations that handle sudden loss well are not lucky. They are the ones that treated the unlikely, unpleasant scenario as worth an afternoon of honest planning, precisely because it was the scenario they most wanted not to think about. The plan you never rehearse is the plan that fails on the one day it is asked to work.
Sources
- Schepker, Kim, Patel, Thatcher & Campion (2017). CEO Succession, Strategic Change, and Post-Succession Performance: A Meta-Analysis. Across 60 samples and 13,578 successions, a change of CEO hurts short-term performance and has no significant direct long-term effect; internal successors give continuity, external ones bring more strategic change. https://doi.org/10.1016/j.leaqua.2017.03.001
- Zhang & Rajagopalan (2004). When the Known Devil Is Better Than an Unknown God: Relay CEO Successions. In 204 successions, relay (a designated, prepared heir) produced better post-succession performance, especially when prior performance was weak and later instability high. https://journals.aom.org/doi/10.5465/20159598