The Board Wants a Successor in Twelve Months. The First Ninety Days Decide Whether You Get the Right One.

The board meeting ends and the chair asks the CEO to stay behind. The message is polite and clear: we want a named successor within twelve months. Maybe the CEO has talked about stepping back. Maybe the investors want a plan before the next raise or an exit. Maybe a director watched a peer company lose its chief executive with no warning. The reflex is to call a search firm that afternoon, or to quietly anoint the obvious internal candidate and call it done. Both skip the decisions that make the next twelve months work, and almost all of those decisions land in the first ninety days.

Two numbers explain why this is harder than it sounds. Russell Reynolds Associates' Global CEO Turnover Index for 2025, published in January 2026, found that 86% of incoming CEOs worldwide were first-time CEOs (79% in the S&P 500). Whoever you name will very likely be doing the job for the first time, which makes the twelve months a development window as much as a selection window.

The same report found that CEO departures in the 30-to-36-month window rose 79% year over year. More new chief executives are leaving around their third year, which is roughly when the strategy they were picked for has either worked or stopped working. The decision that points to is the profile: pick the successor for the company's next three years, not for the last three. These figures cover large listed companies, so read them as direction for a mid-market board rather than a measured rate.

Four doors for a succession mandate

Our Seat Decision framework uses four doors for any open leadership seat. For a CEO succession they look like this:

1. Name. An internal candidate is ready now, and the twelve months are a structured handover. Rarer than boards hope, but real.
2. Develop. An internal candidate has one or two defined gaps, such as no full P&L or no board exposure. Give them a stretch seat with a review date, which is the case we made in The Number-Two Seat.
3. Benchmark. Run a confidential retained search that brings in outside candidates and measures the insiders against the same scorecard. An insider who wins a real comparison starts with more authority than one who was simply next in line.
4. Bridge. If the timeline breaks because the CEO has to leave early, a deal closes, or a health issue arises, an interim or fractional executive holds operations while the process finishes. A fractional COO often carries the operating load so the board is not forced into an early appointment.

Days 1 to 14: decide who knows, and answer three questions

The first decision is the circle. It should be small: the chair or a succession committee of two or three directors, the CEO, and usually one trusted executive such as the head of HR or general counsel. Who must not know yet is longer: the internal candidates themselves, the wider executive team, employees, customers, and lenders unless a covenant requires notice. Candidates who learn they are being assessed start performing for the assessment. Those who learn they are not on the list start taking recruiter calls. We covered where these processes leak in The Confidential Search.

Then the three questions, answered in writing:

1. What will the company need from its CEO in the first three years after the handover? A turnaround, an integration, a sale process and steady compounding each call for a different person, the distinction at the heart of Turnaround or Compounder. The board owns this answer. If the room cannot agree on it, no candidate will look right, inside or out.

2. Who inside could do that job, and what exactly is missing? The CEO knows the team best and should give the first assessment. The board should test it through direct exposure: candidates presenting at board meetings, dinners, a site visit. Write the gaps down as specific experiences, not traits.

3. If the CEO were gone tomorrow, who runs the company? Name an emergency successor in writing now, separately from the long-term choice. It is often a different person, and having that answer removes the panic that forces bad appointments.

Days 15 to 45: split who decides what

Succession goes wrong when the CEO and the board blur their roles. A clean split looks like this. The board decides the profile, the process, the scorecard and the final choice. The CEO decides how internal candidates are developed and exposed, what knowledge is handed over and in what order, and their own path: leaving, staying as chair, or advising for a set period. A CEO who picks their own successor tends to pick someone who resembles them, which is the risk we described in What Boards Owe the Next CEO. The board holds the pen so the CEO does not have to.

Days 46 to 90: choose the door and start

By day ninety the board should hold five things: a written profile tied to the three-year plan, a named emergency successor, an honest assessment of each internal candidate against that profile, a chosen door, and a decision date, usually month nine or ten, which leaves two or three months for the handover. If the door is Benchmark, the search starts here under a code name, with outreach that describes the opportunity without naming the company until a candidate has signed a confidentiality agreement. Insiders are assessed on the same criteria, discreetly, through the work they are already doing.

A twelve-month deadline is really a ninety-day deadline with a long handover attached.

What this is not

This is not an argument against promoting from within. The internal candidate is often the right answer, and benchmarking them against the market usually makes the appointment stronger rather than weaker. It is also not a rule that every succession needs a search. Some boards will finish the ninety days, look at a ready successor, and choose Name. That is a good outcome when it is chosen deliberately.

The process rule is the one we use in every seat: name the door, write down why, and set a review date. For a succession that date is the day-ninety board session, and nothing about the choice should be left to drift until month eleven.