Why Boards Quietly Lose Confidence in Data Programs

Boards rarely lose confidence in data programs overnight. What I see instead is a slow, almost imperceptible shift. Early curiosity turns into tolerance, and tolerance eventually becomes quiet concern. Nothing dramatic happens, the program doesn’t collapse and funding often remains in place. Updates continue to be presented and on the surface, everything looks stable.

But beneath that stability, something important changes and that is the nature of the questions being asked.

In the early stages, boards tend to ask forward-looking questions. What’s next? What capabilities will this unlock? How will this help us compete? These are questions driven by optimism and strategic intent. Over time, however, I’ve seen the tone shift. The questions become more grounded, more cautious, and more financially oriented. What are we actually getting for this spend? Which parts of this program are delivering value today? How confident are we that this investment is justified?

When those questions cannot be answered clearly and consistently, confidence begins to erode. Not because the work is wrong, and not because teams are failing, but because the value story is weak. Boards are not anti-data or anti-AI. They are, however, deeply uncomfortable with ambiguity, particularly when it is attached to significant and ongoing spend.

The challenge is that data investment rarely presents itself as a single, visible line item. Instead, it accumulates quietly over time and across initiatives. Platforms and tooling are procured to modernize the estate, specialist teams are hired to build capability, external partners are brought in to accelerate delivery, with AI pilots and proofs of concept launched to explore future opportunity. Each of these decisions is usually reasonable in isolation. Each one has a defensible rationale at the time it is made.

The problem emerges when these decisions are viewed collectively. Very few organizations can articulate, with confidence, which data initiatives matter most, which outcomes are being deliberately prioritized, and which elements of spend would be actively defended under scrutiny. As a result, data programs often become broad, well-intentioned portfolios of activity and outputs, rather than a small number of sharply prioritized bets.

This is where boards start to feel exposed.

From a board perspective, the risk is not technical failure, but reputational and financial. When a significant area of investment cannot be explained in simple, defensible terms, it becomes vulnerable. Not vulnerable because it lacks potential, but because it lacks clarity; and when clarity is missing, trust starts to thin.

It is at this point that boards intervene, often quietly. They may request additional reporting, seek independent reviews, tighten governance, or change sponsorship. In some cases, leadership roles are reframed or replaced. Rarely is the program “cancelled” outright. More often, it is constrained, re-scoped, or put under heavy control. By the time these actions are taken, confidence has already been lost.

A useful way to surface this risk is to ask a deceptively simple question: If the board asked tomorrow which three data initiatives you would personally defend, and why, could you answer immediately and with evidence? Not with aspiration, or future potential, or references to effort and activity, but with a clear articulation of value realized or value confidently expected.

In many organizations, that question exposes the real issue. The gap is not delivery capability, not talent and not even strategy in the abstract. It is the absence of a disciplined, shared view of value, one that allows leaders to make trade-offs, stop work deliberately, and explain decisions under pressure.

This is the gap I spend most of my time working on with executive teams. Not how to deliver more, but how to make choices clearer. Not how to build momentum, but how to make investment defensible.

So if the board asked that question tomorrow, would you feel confident standing behind the answer?