Menu
Close
By Andries van Oers
Nymark Strategist
This piece comes from a discomfort I have felt in many leadership conversations about technology. The people in the room are usually smart, serious and aware that the stakes are high. And yet the conversation often feels strangely detached from the thing it is meant to govern. Not because anyone is careless, but because the machinery of executive work has already turned the reality into something cleaner and easier to discuss.
After almost fifty years of executive education to close the gap between leadership and the technology their companies run on, the same challenges keep appearing. One reading of this is about effort, or pace, or the difficulty of teaching people new ideas. A more interesting reading asks what kind of problem survives that much sustained attention to closing it.
In most leadership rooms, technology arrives quite late in the agenda. Someone presents a deck. The deck is competent, the questions polite, the engagement uneven compared to other topics in a way no one names. A decision is taken and the conversation moves on.
The consequences surface slowly. A platform that was meant to integrate everything turns out to integrate less than the sales engineer suggested. A contract renews on terms no one read carefully when they were the purchase terms. A dashboard that the company has quietly stopped trusting stays in next quarter's budget anyway, because pulling it would mean admitting the original decision was wrong.
One decision at a time, these are unremarkable. Across enough of them, in enough companies, they start to look less like bad individual calls and more like a structural feature of how the calls are made.
One would think that after nearly half a century, with all the smart people, CIOs, academia, it would move further down the list. It hasn't.
The conventional reading, and the one almost every industry conference reaches for, is that executives lack technical literacy and need to acquire it. This reading has been in circulation, in roughly its current form, for almost fifty years. The vocabulary updates with each wave. In the late eighties, leaders were told to understand the personal computer, and the books promised revolutions in white-collar productivity.
A decade later it was ERP, on covers featuring stern photographs of factory floors. Then the internet, showing binary code dissolving into faces. Then cloud, then mobile, then data, then SaaS, then AI, each wave dressed in a different livery, each promising that this one would be the one to close the gap.
After almost five decades of upskilling, the gap is still there. Which suggests the diagnosis has been looking in the wrong place. The gap is structural rather than personal. It lives in the architecture of leadership work, more than in what any individual leader has yet to learn.
The longitudinal data is unusually clean for a management claim. Senior IT leaders have been surveyed almost every year since 1980, and alignment between technology and the business has appeared in the top concerns of those surveys across the entire span.
The Standish CHAOS reports (a long-running benchmark series on IT and software project outcomes, first published by The Standish Group in 1994) tell the same story from a different angle. Across tens of thousands of project autopsies since 1994, executive sponsorship is the single most-cited success factor, and its absence the most-cited failure mode.
But sponsorship is not endorsement from a distance. It is executive involvement and accountability: the structural role that keeps a decision connected to power, trade-offs, resources and consequences.
It is about who owns the decision, over what timeframe value is judged, and what the company is actually paying for. Technical knowledge is something else.
Technology used to be something the company had. It is now something the company is.
In the early sixties, Alfred Chandler walked through the archives of Du Pont, General Motors, Sears and Standard Oil, and noticed something the companies themselves had not quite articulated: they had been reorganising into separate divisions, with a thin layer of administration on top.
He called it the multidivisional form and the name gave it permanence. A few years later Peter Drucker described the work of that thin top layer as knowledge work, which was a polite way of saying it was the work of deciding what other people should do without doing it yourself.
At MIT in the late seventies, John Rockart invented the executive information system: a way of translating the running of a business into a small set of pre-digested critical success factors, delivered on a manageable cadence. Around the same time the title Chief Information Officer first appeared in print, designed from the start as a translator between the technologists and a senior team that would never quite know the systems first-hand.
By 1992 the balanced scorecard had completed the picture. The quarterly board pack you sat through last month is a direct descendant.
The architecture was right for its moment. When the operating reality of a company was made of human work, organised into functions, executing on cycles measured in months, the rituals of leadership were calibrated for that operating reality.
You could parachute into a factory, sit in on a regional sales meeting, walk a warehouse, and form a usable picture of how the company actually ran. Abstraction was a feature of executive work, because the underlying reality was still visible enough to be checked directly.
That has stopped being true, and the change happened so gradually that most leadership structures did not register it. What runs the company now is not what runs the company in the way most people in the room think of it running.
Software releases on Tuesdays. Integrations bend reality between Salesforce and SAP in ways no one present has parsed. A line manager has signed a contract that quietly reshapes the unit economics of a product line, and the relevant document is on a laptop and not in any board pack. The substrate moved, but the architecture stayed where it was.
What I keep returning to is the inversion itself. Technology has been new in every decade; that part has always been true. What has changed is the relationship between technology and the company.
Technology used to be something the company had. It is now something the company is. The first kind of thing belongs to a delegate, but the second kind cannot.
So what looks like a competency gap in individual leaders is often something else: the predictable result of the system around them. The further information travels, the more it loses its edge. Context gets compressed and signals get tidied up. By the time the message reaches the people meant to act on it, it no longer carries the friction that made it useful.
The quarterly rhythm moves more slowly than the reality it is supposed to govern. By the time the board pack arrives, it is full of activity, when what is really needed is a clearer view of the conditions beneath that activity. Each part may make sense on its own. But together, they create a pattern that works against the people operating inside it.
The translator trap
Translation is not just transfer; it is interpretation: a decision about what matters, what can remain implicit, and what the audience on the other side needs to hear. The CIO role often performs that same compression for the executive team, turning complex technical and organisational realities into a version the room can absorb.
That work is necessary, but over time the compressed version becomes the executive understanding of the system, and what does not make it into that version is, in practice, absent. This is not a failure of any particular CIO. It is a property of the structure: each CIO is asked to perform the same translation, so the same kinds of information keep disappearing.
The fluency illusion
The phone in your pocket has done a remarkable job of teaching you that technology is intuitive. Every day it serves a vast underlying complexity to you in the form of a quiet, helpful interface, and the experience trains you to expect the same thing from every system. Enterprise technology declines to provide it.
Enterprise technology is the underlying complexity, dressed in interfaces that were built for administrators rather than for you. The fluency that develops at the consumer end of the substrate is the wrong fluency for making structural decisions about the enterprise end, and the gap is widest at exactly the moment leaders feel most confident closing it.
The procurement reflex
There is a particular feeling, late in a difficult quarter, when buying something starts to feel like doing something. The path of least resistance is to add a tool. A new platform, a new vendor, a new licence. The portfolio expands. The average company is now running over a hundred SaaS applications, a substantial share of which were bought by line managers without anyone in IT ever quite knowing.
The reflex is rational inside the architecture. Buying is what the architecture is built to do. The deeper response, which is to change the structure that keeps producing the strain, calls for instruments the architecture itself does not contain. So the reflex wins, another tool is added, and the underlying condition continues.
Being a digitally savvy board by the 2019 standards had become table stakes by 2024. It was no longer a differentiator.
The strongest counter-position is worth sitting with, because it comes from researchers who have done careful work and would likely challenge the argument I have made so far. In 2019, a team at MIT’s Center for Information Systems Research found that companies with “digitally savvy” boards significantly outperformed their peers on revenue growth, return on assets, and market capitalisation.
The methodology was rigorous and the sample was substantial. For a moment, the digital-literacy argument seemed vindicated.
However, when the same researchers returned to the question in 2024, the finding complicated the thesis. Their original criteria for board savviness, the same keyword set that had produced the performance premium five years earlier, now classified 72 percent of boards as savvy. The premium had disappeared. To find an effect again, they had to raise the bar. The boards had moved, and the definition had to move with them. What remained was the more important point.
The premium was not simply in what individual directors knew; it was in how boards organised their attention around technology: committee structure, agenda time, and the decision to treat technology as part of the strategic conversation rather than as a delegated topic.
The names changed and the technologies changed, but across both studies the highest-performing leadership teams were those that had built the technology substrate into the work of leadership, rather than placing it around the edges.
That is the thing worth copying. The literacy is downstream.
A prescription would close this too neatly. These instruments will not be redesigned in an article. They will be redesigned in the leadership rituals of actual companies, by people willing to stay with a harder question:
What should executive work consist of when the operating layer of the company has changed underneath it?
That is the question worth answering. If a company’s operating reality is now made largely of software, and that software changes faster than the rituals built to govern it, then leadership’s connection to what it leads becomes an architectural problem before it is an educational one.
The architecture of executive work is the last place most leadership teams look, because it is the most invisible and the hardest to change. But on the available evidence, it is also where the pattern resolves.
So the question is not whether leaders understand technology well enough. It is how leadership stays connected to what it is leading.
