The constraint most industrial boards miss

Ask how an industrial board is built and the honest answer is experience. The chair has run a business at scale. The senior independent director has sat through two downturns and a hostile bid. The audit chair has closed the books at a listed company and knows where the bodies tend to be. Every seat is a proven career, and the collective memory in the room runs to centuries.

That memory is the point of a board. Directors exist to have seen the film before. When a management team proposes a leveraged acquisition, someone in the room recalls how the last one unwound. When a plan assumes the cycle turns in the second half, someone remembers the years it did not. Pattern recognition is the service a non-executive sells, and it is priced accordingly. The composition of most industrial boards optimises for it precisely.

The numbers confirm the tilt. The 2025 UK Spencer Stuart Board Index puts the average age of a serving non-executive at around 60, and new appointments arrive at about 58. Directors are chosen late in a career, after the record is written. A board is a portfolio of completed careers, weighted towards the decisions that already resolved. It is, by design, a machine for reading the present against the past.

There is a second capability a board can hold, and industrial boards hold far less of it. Call it imagination. Not creativity in the marketing sense, and not enthusiasm for the new. It is the disciplined ability to reason about conditions that have not occurred yet. A technology that changes the unit economics of a plant. A competitor that did not exist five years ago. A policy that redraws the map of where things get built. Imagination is judgement about the counterfactual, and it is a different muscle from memory.

The two are not interchangeable. Experience answers the question of what usually happens. Imagination answers the question of what could happen this time, and why this time might differ. A board rich in the first and thin in the second reads every new situation as a version of an old one. Most of the time that instinct is right, because most situations are versions of old ones. The risk sits in the minority of cases where the pattern breaks, and those are the cases that decide the decade.

The industrial economy is running an unusual number of those cases at once. Artificial intelligence is moving off the screen and into machines and plants. The power system is being rebuilt around demand the grid was never sized for. The International Energy Agency expects data-centre electricity to roughly double between 2025 and 2030. Grid connection is now a binding constraint on where capacity can land. Supply chains are being reshored under policy pressure that reverses thirty years of the opposite. Each of these is a case where the last cycle is a poor guide to the next one. Experience alone will misread them, because experience is calibrated on a world that is changing underneath it.

This is where the composition question turns into a governance question. Boards do not lack imagination by accident. They filter for it out, gently, through the mechanics of how directors are found and chosen. The nomination committee writes a specification built from the incumbent profile: sector veteran, listed-company record, prior board service. Every clause on that list is reasonable, and the sum of them screens for people who resemble the people already in the room. The search firm optimises against the brief it is given. The candidate who has thought hardest about the next fifteen years, and has less of the last fifteen on paper, does not clear the specification. Governance filters out the exact capability the decade demands, and it does so while following good process.

The remedy is not to swap experience for imagination. A board that traded its memory for novelty would make different and worse mistakes. The remedy is to widen the aperture deliberately, and there are three levers that do it without disturbing the discipline that makes a board useful.

The first is committee design. The standing committees, audit, remuneration, nomination, encode a view of what a board is for, and that view is retrospective and controlling. It monitors what has been reported and rewards what has been delivered. A technology or strategy committee, or a standing agenda block owned by the whole board, changes what the board is structurally obliged to look at. Structure decides attention. A capability with no committee behind it gets discussed when time allows, which means rarely.

The second is agenda time. Board agendas are dominated by the backward-facing: results, audit, compliance, the report on the last quarter. This is necessary and it is measurable, which is why it expands to fill the day. The forward-facing item, the one about a shift that has not shown up in the numbers, is scheduled last and cut first when the meeting overruns. A board serious about the future protects that time before the meeting starts. The important-and-urgent will always take whatever it is allowed to.

The third is external lenses. A board does not have to carry every capability in its own seats. It can commission an outside read on a technology shift, a competitor, or a market it does not yet understand. It can bring that read into the room as a standing input, on the agenda every cycle. An external lens does not outsource judgement. It gives the board’s own judgement something to work on that its collective memory does not already contain.

None of this asks a board to bet on a future it cannot see. It asks the board to notice that its composition, its committees, and its agenda are all tuned to the same setting, and that the setting is memory. The industrial decade will reward the boards that keep their experience and add the other thing. The composition question is where that starts, because a board can only reason about what it is built to look at.

If your board is weighing where its own aperture sits, the Advisory runs board briefings and a fixed-scope diagnostic that maps the binding constraint. Start a conversation.