Board refreshment: building the right mix for what's next

Last year a chair we work with replaced three non-executive directors over an eighteen-month window. One reached the end of a third three-year term. One stepped down for health reasons. The third had to go for a quieter set of governance reasons that did not require a public statement. Three credible new directors were brought in, each through a properly run search, each on merit. The chair told me at the end of the cycle that the process had gone well.

Then he told me, without quite meaning to, that the board still felt the same. The skills it had two years ago, it still had. The skills it did not have two years ago, it still did not have. The geography it was missing, it was still missing. The committee chairs were unchanged. The dynamic in the room was unchanged. He had refreshed three seats and ended up, in his own private summary, with the same board plus three new names.

That conversation is the most common version I have of the failure mode this essay is about. It is not a failure of search execution. The search process worked. It is a failure of question. The chair refreshed three seats; he did not refresh the board.

Why one-out-one-in is structurally wrong

The seat-replacement framing is structurally wrong because it treats the board as a sum of its current parts minus retirements. The framing is: this seat is becoming vacant; what kind of person should fill this seat. The framing produces, reliably, a candidate who looks like the previous occupant of the seat with one or two updates, often the updates that were obviously missing the last time round.

This is fine if the board is already correctly composed and the only problem the chair is solving is the calendar problem of director term-end. But that is rarely the actual situation. The far more common situation is that the board's composition was last actively designed three or four years ago, the institution has changed since, the regulatory and capital-markets environment has changed since, and the strategy the board is now overseeing has changed since. In that situation, replacing a single seat with a like-for-like updated profile preserves a board shape that has already drifted out of fit.

The chairs who have done this best, in our experience, do not start with the seat. They start with the board.

The structural confusion is between personnel decisions and composition decisions. Personnel decisions are about which individuals occupy which seats. Composition decisions are about which seats should exist, what each one should contribute, and what the relationship between the seats produces in aggregate. A board that confuses the two collapses composition into personnel, and the composition question, because it is never asked, never gets answered.

This is not a small technical observation. It is the difference between a board that gets gradually stronger over a decade and a board that gets gradually less fit for the institution it is meant to govern.

A better starting point: the five-year mix question

The question we recommend chairs start with is the five-year mix question. It runs roughly as follows. What does this institution need its board to be able to do, well, in 2031. Not who is on it; what does it do. From that, what mix of skills, geographies, sector experience, regulatory exposure, generational fluency and personal conviction does the board need to embody to be capable of doing those things.

The answer is rarely the current board. The answer is also rarely a totally different board. It is usually a recognisable evolution: certain capabilities reinforced, others added, others gracefully removed.

The framework reframes what every subsequent decision is in service of. The next director the chair appoints is not the replacement for the seat that just opened. It is the next move toward the 2031 board the institution will need. Sometimes that means the next appointment looks nothing like the seat being vacated; sometimes it does; sometimes the question of which seat to vacate next is the more important question, because the chair is suddenly able to see that one of the long-tenured directors is not in the 2031 picture and the conversation with that director should start now rather than at term-end.

The data above is not subtle. The single-seat brief is the standard form a board refreshment conversation starts in. The five-year-mix conversation is, in the boards that get this right, where the conversation ends up. The window in which the widening happens is the first six weeks. After that the brief crystallises around the original seat and becomes hard to unwind.

This puts a particular burden on the chair, the search firm and, increasingly, the lead independent director or deputy chair if the board has one. The widening of the conversation is something one of those three has to be willing to push for. It is rarely the case that the widening happens organically.

The four mixes worth getting right

There are four mixes that, in our experience, repay the most attention in this conversation. They are not the only four. They are the four where most boards either get it materially right or materially wrong, and where a disciplined five-year reading produces the largest practical effect on appointments.

Operating experience mix

The most common gap is operating experience at the relevant seat level. Boards routinely add finance backgrounds, regulatory backgrounds and consultant backgrounds; they add senior operators less often, partly because senior operators are still in operating roles and partly because the search is harder. The boards that consistently outperform their peers tend to have at least one and ideally two directors who have actually run something at scale recently. Recently is the operative word. A retired CEO whose last operating experience was eight years ago is not the same input as a CEO who handed over the seat eighteen months ago.

A worked example, anonymised. A pan-regional industrials group we advised in 2024 had a board with strong finance representation, two regulatory backgrounds, an experienced family-side chair, and one academic appointment. The chair told us the board had become structurally cautious. The five-year read, when we walked through it, surfaced that the board had no recently-stepped-down chief executive on it. The next appointment, against the original brief which was for an audit committee chair, became a former regional CEO with audit committee fluency. The committee chair shape was preserved; the operating-experience gap was filled in the same hire.

Geography mix

The geography mix is the one most often misread because it is most often confused with diversity-of-passport rather than diversity-of-lived-experience. A board with three directors holding three different passports may, in lived professional terms, be a board with three directors who have spent their careers in the same global cities and the same global firms. That is not a geography mix; that is a passport mix.

The relevant question is whether the geographies the institution operates in, and the geographies it intends to expand into over the five-year horizon, are represented on the board by directors whose professional history actually sits in those places. For a Gulf-headquartered group expanding into East Africa, the question is whether anyone on the board has run something in Nairobi, Lagos or Addis Ababa for long enough to have an opinion. For a Singapore-headquartered logistics group expanding back into the Gulf, the question is whether anyone on the board has worked in Dubai or Riyadh long enough to have working relationships there.

Geography mix done well requires the chair to be honest about where the institution actually operates, where it actually intends to expand, and where it does not. Boards that ask the question well typically end up with fewer geographies represented at depth, rather than more, because the discipline of insisting on lived depth narrows the answer.

Functional mix

The functional mix is the one that is most likely to be already articulated in board nomination committee documentation but is most often articulated wrongly. The wrong articulation is the skills matrix that lists every conceivable functional area and asks each director to self-assess. The right articulation is the functional gap analysis that asks: at first-hand depth, which functional perspectives is this board capable of holding the executive accountable to.

Most boards have finance covered. Most boards have legal or regulatory covered. Operations is patchier. Technology, even in 2026, is patchier still, because too many boards continue to describe a director who used technology in their executive role as a "technology director", which is not the same as a director who has run a technology function at scale. ESG is becoming a real first-hand depth question rather than a compliance overlay; talent and people strategy is a real first-hand depth question that very few boards have a credible director on.

The functional mix, run honestly, almost always reveals one or two real gaps. The five-year read pushes those gaps to the front of the appointment queue.

Tenure mix

The tenure mix is the most uncomfortable of the four because it is the one that requires the chair to look at the directors already in the room and form a view about which of them should still be in the room in three years' time. A board that has not done this is a board with a tenure distribution that has accumulated rather than been designed.

The boards that do this well tend to plot their director tenures forward against the five-year horizon. The plot frequently reveals that two directors will be at twelve-plus years of tenure within the horizon, that one will reach the end of a third three-year term in eighteen months, and that the chair himself will be at his sixth year by 2031. The plot makes visible decisions that are otherwise made one at a time and almost always too late.

A worked example, anonymised. A GCC-listed group we worked with in 2025 had four directors at over nine years of tenure and one at over fifteen. The five-year tenure plot, when produced, made the conversation about which two of those directors would still be on the board in 2031 a conversation that the chair could lead constructively, rather than one that became a series of awkward individual conversations as each director's term expired.

Where this most often falls apart

The five-year mix question is, in our experience, the right question. It is not the question that gets asked easily. There are four common reasons for that, and the chairs who refresh boards well have explicit positions on each.

The first is the politics of refreshment. A serious five-year read, run honestly, will surface the fact that some sitting directors are not in the 2031 picture. Asking the question therefore requires the chair to be willing to have the conversation that produces the answer. Chairs who are not willing to have those conversations should not run the read; they will end up making cosmetic adjustments and keeping the underlying composition in place.

The second is family dynamics. In family-controlled groups, particularly Gulf-listed ones, the board includes family members, family-affiliated long-service directors, and independent directors whose appointment was originally a relationship appointment. The five-year read asks the family principal whether each of those directors, family or not, is in the 2031 picture. That is a culturally complex question, and it is harder to ask in some institutional cultures than others. The chairs who handle this best treat the question as institutional rather than personal, and they introduce the framework before any individual director's term-end has made it personal.

The third is the retiring director who is not yet ready to retire. The five-year read, when it identifies a director whose presence will not extend through the horizon, occasionally does so for a director who has not yet decided that themselves. The chair then has to have a structured conversation about what the runway looks like, and the conversation goes badly when the chair is not prepared for it. The conversation goes much better when the chair has asked the five-year question for the whole board, not for the individual director, and can frame the runway conversation as a board-level decision rather than a personal one.

The fourth is the principal who treats the board as a personal cabinet. This is the situation in which the chair, the principal, or both, see the board primarily as a circle of senior advisors who are loyal to the principal and to each other, rather than as a governance institution that exists to oversee the executive on behalf of the broader shareholder base. In this configuration the five-year mix question is structurally hard, because the answer the question produces is often that the board has been composed on a relationship basis rather than a capability basis. The chairs who navigate this well are the ones who use the listed-company discipline, where it applies, to professionalise the conversation; in unlisted groups, they use the family council as the institutional venue in which the question can be raised without personal cost.

What the right refreshment looks like

The right refreshment, in our experience, has three characteristics that distinguish it from the wrong one.

It starts before the seat opens. The chair who knows, eighteen months out, that two terms are coming up for renewal, has time to run the five-year read, to test the implications, and to brief the search firm on the wider mix question rather than the narrow seat question. The chair who starts the conversation when the seat is vacant has already lost most of the room to ask the bigger question.

It is run as a single conversation across multiple appointments rather than as a sequence of separate single-seat conversations. The same chairs who refreshed three seats over eighteen months and felt the board had not changed are the chairs who ran three separate conversations. The chairs whose three appointments produced a meaningfully different board ran one conversation that happened to land three appointments.

It includes director exits as part of the design, not as separate housekeeping. The five-year read identifies directors whose presence is not in the 2031 picture. The decision to manage their runways toward graceful exit, rather than to wait for term-end, is a design decision that the chair owns. Exits are part of the refreshment, not a separate problem.

The chairs who do this well make their boards, over a decade, materially better at the job of governing the institution. The chairs who do not do this well make their boards, over a decade, materially less fit for the institution they are meant to govern. The work of board refreshment is the slow accumulation of either trajectory, and the question that decides which trajectory a board is on is whether the chair refreshes seats or refreshes the board.

That is the only question that matters.

JOH Partners advises chairs and nomination committees on board composition, refreshment and chair-level mandates across the GCC, the UK and Singapore. For confidential conversations, contact the partners directly.