Digital Transformation Needs the CEO, Not Just the CIO

There is a specific moment I recognize now in most digital transformation conversations. Somebody senior says the words "digital transformation" and means "we should use the cloud more" or "we need to automate some of this." The CIO nods. The project gets scoped. A vendor gets selected. And six months later, the organization has a new tool layered on top of the same processes, run by the same structure, with the same incentives that produced the old problems. The transformation part never happened.

I have read enough of the IS literature now to understand why this pattern is so durable. The Wessel et al. (2021) framing is the most precise I have found: genuine digital transformation means redefining the value proposition and changing the organizational identity. Not improving how you deliver the existing value proposition. Not making the existing identity faster or cheaper. Redefining what you offer and to whom, and becoming a different kind of organization in the process. That is a much harder thing than adding a cloud app, and it requires a very different kind of leadership.

The CIO cannot do this alone. That is not a knock on CIOs. It is a structural observation. A CIO controls IT infrastructure and, depending on the organization, has influence over IT-adjacent business processes. What a CIO does not control is the business model, the organizational culture, the incentive structure, the governance of business units, or the strategic direction of the firm. All of those things have to change in a genuine digital transformation, and none of them change without the CEO's active involvement and sustained commitment. Not endorsement. Not sponsorship in the passive sense of appearing at a kickoff event. Active involvement in the hard decisions about what the organization is willing to give up, who bears the transition costs, and how long the organization will tolerate the discomfort of changing while continuing to operate.

David Rogers, writing about digital transformation strategy, has argued (and I hedge this because I am working from secondary discussion of his work rather than reading his book directly) that the core challenge is not adopting new technologies but rethinking strategic assumptions. Organizations are often so deep inside existing assumptions about how they create value that they cannot see which assumptions are the ones the market is about to stop rewarding. This framing aligns with what Wessel et al. are describing: the transformation is not a technology event, it is a strategic and cognitive event, and it requires leadership that can challenge assumptions at the top of the organization, not just deploy tools at the operational level.

Culture change is slower than technology change. This seems obvious when stated plainly, but it is not how transformation programs are typically scoped and budgeted. A software implementation has a go-live date. The training program has an end date. The new platform is "live" in some sense by a defined point. Culture does not have a go-live date. The informal norms, the actual decision-making patterns, the real incentives that govern how people behave when nobody is watching, those things change over years, not quarters. Organizations that budget eighteen months for a digital transformation and measure success by technology deployment completion have misunderstood what they are trying to change. The technology was the easy part.

Organizational politics are where DX initiatives most commonly die, and this is where the CEO's authority is irreplaceable. Every significant digital transformation creates winners and losers inside the organization. A process that gets automated eliminates work somebody was doing. A digital channel that replaces a physical one restructures the distribution that somebody managed. A shared data platform that gives the center visibility into business unit performance feels like surveillance to the unit leaders who preferred opacity. These political conflicts are real, and they surface during implementation as resistance, delayed approvals, scope negotiations, and strategic non-compliance. A CIO cannot resolve these conflicts because the conflicting parties often have organizational rank equal to or higher than the CIO. Only the CEO can make the call that a particular business unit cannot veto the transformation because the transformation is the strategy.

This is also why digital transformations that work tend to be organized differently from the ones that stall. The successful ones treat the transformation as an organizational change program with technology as the enabling layer, not a technology deployment program with organizational change as the follow-on. The governance structure is different. The metrics are different. The failure modes are anticipated differently. Instead of measuring platform adoption rates, you measure whether the value proposition is actually changing. Instead of measuring training completion, you measure whether decision-making patterns have shifted. These are harder to measure, which is partly why most organizations default to measuring the technology deployment instead.

The other pattern I keep seeing is the pilot problem. Organizations run a pilot. The pilot succeeds. The board is excited. Then nothing scales. This is not because the pilot was dishonest or the technology was wrong. It is because pilots are structurally protected from the organizational politics that kill scale deployments. A pilot exists at the edge of the organization, in a friendly business unit, with hand-picked participants, dedicated support, and executive attention. When the pilot succeeds and the time comes to roll out broadly, all of those protections disappear. The technology meets the full weight of the organization's existing culture, politics, and inertia. And the organization discovers that it did not run a pilot of the transformation. It ran a simulation of what transformation would look like if the organization were already different from what it is.

The sociotechnical systems view is useful here. Trist and Bamforth's coal mine research, which Sarker et al. (2019) bring back into focus, showed that technical optimization without social optimization degrades overall performance. The longwall method was technically superior. The social system it destroyed was not optional. Digital transformations that optimize the technology and treat the cultural and organizational change as secondary are running the same experiment, and the coal mine result keeps showing up in enterprise transformation data as stalled initiatives, expensive technology that nobody uses, and transformation announcements followed by quiet abandonment.

The CEO cannot delegate the authority required to resolve political conflicts about the organization's future. The CIO cannot supply it. The board can mandate transformation in the abstract, but the specific decisions about which assumptions to abandon, which business units bear the transition costs, and how long the organization will sustain discomfort before reverting, those require executive leadership that is willing to be present for the uncomfortable parts, not just the announcement and the celebration.