Insight · Board expectations

11 Transformation Mistakes Boards No Longer Tolerate

None of these are new. What has changed is that boards used to accept them on trust and now ask about them directly.

For chief executives All sectors 13 min read

In brief

Boards have not become more demanding. They have become more specific, because the last cycle of transformation spend did not produce what was promised.
The eleven below all used to pass on trust. Each is now the subject of a direct question, and the honest answer is often uncomfortable.
The position worth holding is not “we have a good relationship with the board”. It is “nothing in the paper depends on them taking our word for it”.

Boards have changed what they will accept about transformation, and most executive teams are still writing papers for the previous audience.

The change is not that boards have become adversarial. It is that they have become specific. A cycle of significant spend produced less than the cases promised, and the questions that follow are the ones nobody asked the first time.

What used to be sufficient — a credible sponsor, a coherent narrative, a benefits total that added up — is now the opening position rather than the answer. Non-executives have watched the same pattern at other firms and arrive with a sharper list.

The eleven below all used to pass on trust. Each is now the subject of a direct question, and the difference between a comfortable meeting and an uncomfortable one is usually whether the work was done before the paper was written.

01

A benefits total with no named owner behind each line

The case sums to a number that justifies the spend. Each line is plausible. What used to be enough was that the total was credible and the sponsor was trusted.

The question now is which director has personally committed to each line and how their delivery of it will be tracked. If the answer is the sponsor for all of them, nobody owns anything.

02

Benefits stated without the number they move

Improved efficiency. Better customer experience. Faster decisions. These used to be acceptable descriptions of value.

They are now met with a request to name the line on the management accounts that changes, by how much, and when. A benefit that cannot name its number cannot be tracked, which means it cannot be claimed later either.

03

A technology design with no matching design of how the business will run

Papers document the technology thoroughly. The description of how the business will operate afterwards is usually at vision level.

Boards have started noticing the asymmetry, because they have seen where it leads. The question is direct: show us the operating design at the same standard as the technical one. Most papers cannot.

Boards have not become harder to satisfy. They have become harder to satisfy without evidence, which is a different thing.

04

A single-point estimate rather than a range

Cases have historically presented one number for cost, one for benefit, one for timing. Everyone understood these were written to be approved.

Boards now ask for the range and the assumptions behind each end of it. A case that survives being halved is a different proposition from one that only works at the optimistic end, and non-executives increasingly want to know which they are looking at.

05

No baseline, so no way of proving anything afterwards

Programmes used to be approved without anyone measuring the current position. The improvement was assessed afterwards by judgement.

That no longer passes, because boards have been through the cycle where the improvement could not be demonstrated. The question at approval is now whether the baseline exists — and if it does not, whether the paper should wait the few weeks it takes to establish one.

Where do you sit?

The difference between a comfortable board meeting and an uncomfortable one is usually work done before the paper was written.

A free Executive AI Opportunity Assessment. Forty-five minutes, online, with your leadership team. We arrive with a model of how firms in your sector make money and where AI has moved those numbers — and you tell us where it is wrong.

Book a free assessment

06

Change budgeted as training

The change line has historically been small, and described in terms of communications and training. Boards accepted it because it looked proportionate.

They now ask what it was sized against. If the answer is a percentage of the total rather than the number of people whose jobs change, the line is usually wrong by a wide margin — and the board has seen what that produces.

What changed in the board conversation — a pattern, not a measurement

Trust

What used to carry a paper

A credible sponsor, a coherent narrative, a total that added up. All of it accepted in good faith.

Evidence

What carries one now

Named owners, stated baselines, ranges rather than points, and a design of how the business will run.

Why

The cycle that changed it

Significant spend that did not produce what the cases promised, at their firm or at one they know.

Gap

Where most teams are

Still writing papers for the previous audience, and surprised by the questions.

This describes the shift we see rather than a measured dataset. The test worth running is your own: take the last paper you submitted and mark it against these eleven. The ones you would rather not discuss in detail are the ones to fix before the next paper.

07

Directors described as engaged rather than committed

Papers say the executive team is aligned, consulted, supportive. That used to close the question of whether the organisation would adopt what was being built.

Boards now distinguish between consultation and commitment. Has each director signed for the specific changes their team will adopt, the timing, and what they will be measured on afterwards? Consultation gathers opinion. It does not create obligation.

08

Vendor selection presented as though it settled nothing

Papers often arrive after the platform has been chosen, with the technology presented as an enabler of whatever the business decides to do.

Boards have learned that the choice frequently closes options about how the business can work. The question is which decisions the technology has already made, and whether the board is approving a design or ratifying one.

For the chief executive

Three things to do before the next paper goes in

Mark your draft against these eleven and find the ones you would rather not discuss in detail. Those are the questions you will be asked.
For every benefit line, name the director who has committed to it in writing. If any cannot be named, the paper is not ready.
Check whether a baseline exists for what you are claiming to improve. If not, delay the paper by the few weeks it takes. That delay is cheaper than the conversation eighteen months later.

09

Only two options: approve, or do not proceed

Papers have traditionally offered a binary, with the alternative to approval being the cost of inaction. It framed the board into yes.

Non-executives increasingly want the middle options costed. Approve with conditions. Send it back to be reshaped. Do a smaller version first. Being offered only two choices now reads as the paper managing the board rather than informing it.

10

No plan for what happens if it slips

Risk registers list what might go wrong and how it would be mitigated. That used to be sufficient.

The question now is specific: if month eighteen arrives six months late and a quarter over budget, what happens? Which workstreams continue, which stop, who decides, and what brings it back to the board? “We would manage it” is no longer an answer.

11

Reporting that stops at go-live

The most structural of the eleven. Programmes have historically reported until delivery, then closed. The benefits arrive later, unreported, because nobody was accountable by then.

Boards have noticed that they approve outcomes and receive completion reports. The expectation now is that someone owns the benefits for at least a year after the programme ends, and reports on them against the case that funded it.

That single change makes most of the other ten self-correcting, because a paper written by someone who knows they will report against it in two years is a different paper.

What this means

None of these eleven are new mistakes. Every one has been made routinely for years and accepted in good faith. What changed is the tolerance, and it changed because a cycle of substantial spend did not produce what the cases described.

The common thread is evidence. Every one of the eleven is a place where a paper used to rely on the board taking something on trust: that the benefits would be owned, that the numbers were realistic, that the organisation would adopt it, that somebody would be accountable afterwards. Boards have stopped extending that credit, not out of suspicion but out of experience.

The economics of adapting are straightforward. Naming a director per benefit line takes a meeting. Establishing a baseline takes weeks. Stating a range rather than a point takes honesty. All of it is cheaper than the alternative, which is a paper that gets sent back or, worse, approved and then quietly missed.

Which is the argument this whole series has been making. Change the work before you change the technology — and be able to show what will change, what it is worth, who owns it, and how you will know. Boards are asking for that now. They will keep asking.

The next step

Would your next paper survive all eleven?

Forty-five minutes with your leadership team, online, at no cost. We bring a working model of how firms in your sector make money and where AI has moved those numbers. You tell us where it is wrong. Then we tell you what we think — including “don’t invest”, if that is the honest answer.