Digital Transformation Restarted with the Same Flaws — Because Organisations Mistook Pause for Learning
A programme can resume its activity without recovering its capacity to learn.
Digital Transformation Restarted with the Same Flaws — Because Organisations Mistook Pause for Learning
The longer lesson from 2021: emergency speed changed delivery conditions, but not the assumptions governing the programme
The transformation programme returns to the board after a six-month pause. The cover slide is new. The dates have moved. The delivery model now includes remote workshops, cloud capacity and a renewed emphasis on digital channels. Yet by page twelve, the programme is recognisable as the one that stalled before: the same workstreams, the same monthly reporting cycle, the same supplier boundaries, the same benefits written as aspirations, and the same decisions deferred to a steering group too distant from the work to make them well.
Everyone agrees that the world has changed. Almost nothing in the programme’s governing assumptions has.
This is becoming one of the defining contradictions of 2021. Organisations have proved, under extraordinary pressure, that they can move quickly when an outcome is unmistakable and authority is clear. Services have been redesigned in weeks. Teams have worked across locations without waiting for the perfect collaboration model. Infrastructure that once required months of debate has been provisioned at emergency speed. Customers and employees have adopted digital routes far faster than many forecasts allowed.
And yet, as paused transformation portfolios restart, much of that evidence is being treated as an interruption rather than a lesson. The emergency is being archived. The programme manual is being reopened.
The result is not merely a missed opportunity to “work more agilely”. It is institutional amnesia: the organisation has experienced a different way of deciding and delivering, but has not converted the experience into a different operating logic.
The pause changed the calendar, not the theory
A restart is often managed as a scheduling exercise. Teams revalidate milestones, refresh business cases, confirm resources and issue a revised plan. This work is necessary, but it quietly assumes that the programme’s original theory remains sound.
That theory usually contains several beliefs:
- stable requirements can be established early enough to organise delivery around them
- specialist workstreams can optimise their own plans and integrate later
- governance creates control by receiving more complete information
- suppliers can be managed principally through scope, milestones and contractual obligations
- benefits can remain at programme level until delivery is sufficiently advanced
- business participation can be requested when design decisions require it
Those assumptions were already fragile. The experience of the past year has made the fragility harder to excuse.
During the emergency, organisations did not succeed because they produced better versions of their existing programme paperwork. They succeeded where they shortened the distance between operational need, technical judgement and executive authority. The purpose was concrete. Trade-offs were visible. Decision-makers were close enough to the consequence to act. Teams accepted imperfect information because delay itself carried a cost.
None of this means that emergency delivery should become normal. Heroic hours, suspended controls and exceptional risk acceptance are not a sustainable operating model. But it does mean the old claim—that slow, layered decision-making is the price of safety—can no longer pass without examination.
When the programme restarts unchanged, it does more than return to familiar governance. It discards evidence about which parts of that governance were useful and which parts were ceremony.
What organisations remembered—and what they forgot
Institutional memory is selective. We preserve artefacts more readily than judgement. A programme can retain its risk registers, stage-gate papers and lessons-learned logs while losing the practical understanding of why delivery moved differently under pressure.
The facts that survive are often the easiest to record: how many users moved online, how many laptops were deployed, how quickly a service was launched. The harder lessons fade because they implicate the operating model.
The organisation may remember that a team delivered in twelve weeks, but forget that one accountable executive settled competing priorities every forty-eight hours. It may celebrate remote delivery, but overlook the fact that the same five people attended every cross-functional decision because nobody else could join the pieces. It may praise supplier flexibility, but omit that commercial boundaries were temporarily relaxed and problems were solved before responsibility was apportioned.
This distinction matters. If the outcome is remembered but the mechanism is forgotten, leaders conclude that the organisation simply “rose to the occasion”. They convert a design lesson into a morale story.
A programme can resume its activity without recovering its capacity to learn.
That is why the same flaws reappear. The documents are updated, but the causal account is not. The organisation has not asked which normal constraints disappeared, which decisions accelerated, which risks increased, which controls genuinely protected the outcome, and which people compensated for structural weakness.
A familiar restart, seen close enough to matter
Consider a composite £38 million service transformation restarted in the spring of 2021. Before its pause, the programme had four principal workstreams: customer channels, core technology, operating model and data. Each had a credible plan. The integrated plan contained 2,700 activities and reported an overall completion rate of 41 per cent.
The programme had nevertheless failed to resolve a basic question: whether the first release was intended to reproduce the existing service digitally or redesign the service around different customer journeys. Technology assumed the former because it protected the delivery date. Operations assumed the latter because it justified the investment. The customer team tried to accommodate both. The benefits profile depended on the redesign but did not make that dependency explicit.
When work resumed, the programme spent seven weeks rebaselining. Every workstream achieved a new green plan. The unresolved proposition remained unresolved.
At the first integration checkpoint, the customer journey required data that the legacy platform could not provide without additional design. Operations had not allocated people to handle exceptions created by the digital route. The supplier correctly noted that both changes sat outside the agreed scope. A decision that had been avoided before the pause returned as three separate delivery problems.
The programme had restarted efficiently. It had not restarted intelligently.
The figures make the mechanism visible. Eighty-five per cent of planned activity could proceed without resolving the service proposition. That activity created reassuring movement. The remaining fifteen per cent contained the decisions that determined whether the parts would form a coherent service. Progress reporting rewarded the former and concealed the latter.
This is how old flaws survive a new context. They are embedded not only in method, but in what the organisation chooses to count.
The serious defence of the old methods
There is a reasonable opposing view. The last year was exceptional. Crisis arrangements relied on concentrated authority, unusual tolerance for risk and levels of discretionary effort that cannot be sustained. Large transformations still need financial control, architecture, assurance, procurement discipline and a plan that coordinates hundreds of people. To discard established governance in the excitement of emergency delivery would replace bureaucracy with improvisation.
That argument deserves respect. Many apparent examples of rapid delivery transferred cost into later support, security remediation, manual workarounds or exhausted teams. Speed alone is not proof of effectiveness. Nor is every delay caused by governance; some problems are genuinely difficult, and some dependencies cannot be wished away by a daily meeting.
But the choice is not between the old programme and permanent crisis mode. That is the convenient binary that protects both from scrutiny.
The useful question is: which disciplines reduced uncertainty or protected value, and which merely delayed the moment at which someone had to exercise judgement?
A sound architecture decision is not bureaucracy. Requiring the same issue to be summarised for three forums is. Independent assurance can reveal optimism and blind spots. A monthly meeting that notes an unresolved dependency without assigning authority does not. Commercial discipline protects public and shareholder value. Using contract boundaries to postpone a cross-supplier design decision destroys it.
The lesson of emergency delivery is not that controls are unnecessary. It is that controls should be judged by the decisions and protections they create, not by the evidence that the process occurred.
The opposite of bureaucracy is not speed. It is governance that converts uncertainty into an accountable decision.
The deeper flaw is separation
Much digital transformation has been organised by separating things that must ultimately work together.
Strategy is separated from delivery through a business case handed over after approval. Technology is separated from operations through workstreams. Customer experience is separated from process design through specialist teams. Benefits are separated from products until late-stage adoption. Suppliers are separated by contracts even when the customer experiences one service.
The programme then builds an integration layer—boards, plans, reports and assurance—to reconnect what its own structure has divided.
That layer struggles because it receives abstractions. A workstream reports percentage complete; it does not expose the design assumption on which its progress depends. A supplier reports against scope; it does not own the end-to-end outcome. A steering group sees risks once they have been formatted, moderated and assigned colours. By then, the live tension has often become a presentational issue.
The past year temporarily collapsed some of these separations. Operational leaders sat beside technical teams because the service had to work. Executives made narrower, more frequent decisions because waiting for complete information was itself dangerous. Teams organised around outcomes that ordinary governance had previously fragmented.
As transformation restarts, the old separations are returning because they are institutionally comfortable. They align to budgets, professions and contracts. The cost appears later, as integration risk.
Learning must change the next decision
A lessons-learned exercise will not cure institutional amnesia if its findings are stored beside the old governance. Learning becomes real only when it changes who decides, what is measured, how work is grouped or which commitment is refused.
A credible restart should therefore force several uncomfortable moves.
- Re-state the outcome in operational terms. Not “deliver a digital platform”, but which service changes, for whom, with what measurable consequence.
- Expose the few decisions that make the plan coherent. Put these ahead of activity volume and give each a named owner with authority across workstreams.
- Organise integration around the service, not the reporting calendar. Bring operations, technology, data, customer and commercial judgement together where choices are made.
- Carry benefits into design. Every major design choice should show which benefit assumption it supports, weakens or makes impossible.
- Retire controls that cannot explain their protective purpose. If an artefact has no decision, risk reduction or assurance function, familiarity is not a sufficient reason to keep it.
- Revisit supplier boundaries at points of shared consequence. Contracts may divide accountability, but the customer journey will not.
These are not invitations to remove discipline. They are ways to relocate discipline nearer to the work and its consequences.
The board’s role also changes. Instead of receiving a comprehensive description of progress, it must engage with the small number of choices that shape the programme’s direction. That requires less theatre and more judgement. It may also require admitting that a green rebaseline is less valuable than a red decision made early.
The longer view
It is tempting to describe the current moment as a once-in-a-generation acceleration of digital transformation. That phrase flatters organisations because it suggests movement was inevitable and success came from embracing the future.
The more demanding interpretation is that the emergency exposed how much delay had been designed into normal delivery. It revealed that some organisations were not short of methods, technology or ambition. They were short of permission to decide across boundaries.
As offices reopen unevenly and hybrid working becomes a practical management question, transformation programmes will absorb new tools and rituals. Workshops will remain online. Cloud investment will continue. Customer expectations will not return to their earlier pace. Yet these visible changes can coexist with the same underlying programme: fragmented ownership, distant governance, benefits detached from design and progress measured as completed activity.
That is why the restart matters. The test is not whether the portfolio resumes. It is whether the organisation can identify what its own experience disproved.
Institutional amnesia is rarely a failure to remember events. It is a failure to let those events disturb the system that produced them. Organisations will preserve the story of how quickly they moved under pressure. The stronger ones will preserve the mechanism—and redesign ordinary delivery so that clarity, proximity and accountable judgement no longer require a crisis.
Digital transformation does not restart when the plan is approved. It restarts when the organisation is willing to change the assumptions that made the old plan fail.