Twelve Weeks in Spring — What Emergency Delivery Taught Us Before We Forgot It
What they proved is that a great deal of what organisations do in peacetime is not delivery at all — it is the elaborate maintenance of structures that exist to make the absence of delivery feel orderly.
Executive Summary
Between late March and mid-June 2020, programme delivery in organisations across the economy underwent an unplanned experiment. Constraints that had been presented as structural — approval chains, dependency management ceremonies, multi-stage governance — dissolved under emergency conditions and did not return. What replaced them was not chaos but something closer to the opposite: sharper scope, faster decisions, clearer accountability. This essay is a field catalogue, written while the evidence is still fresh, of the specific emergency practices that proved our delivery constraints were largely matters of preference rather than physics. It is also a prediction — offered with a practitioner’s certainty — of exactly how each practice will be killed as normality reasserts itself. The argument is not that emergency conditions should be permanent, but that we owe ourselves an honest reckoning with what the emergency revealed about the machinery we maintained in peacetime. That reckoning has a window. It is closing.
The Board That Finished Early
In the first week of March, a programme board I was involved with ran to its full two-and-a-half hours. There were thirty-seven slides. Dependencies were presented as a network diagram that no one in the room could read and no one was willing to admit they could not read. Three items were escalated for decision; none were decided. The board was scheduled to reconvene in four weeks.
By the end of April, the same programme had replaced that board with a forty-minute call. Five slides. Decisions were made on the call or within twenty-four hours. The dependency diagram had been replaced by a list of four names — the people who actually controlled the critical path — and those people were on the call. Scope had been cut by roughly a third, not because resources had shrunk but because someone had finally been forced to say which third mattered most.
No one designed this. No methodology mandated it. The programme did not hire a consultancy to redesign its governance. What happened was simpler and more uncomfortable: an emergency removed the organisational permission to defer, and the programme discovered that a great deal of what it had been doing was deferral dressed as diligence.
This essay is about that discovery. It is written in September 2020, six months into a disruption that is still unfolding, because the window for honest observation is already narrowing. The institutional immune system is waking up. The practices that proved themselves in spring are already being reclassified as temporary expedients rather than permanent revelations. If we do not catalogue them now — name them, explain why they worked, and predict how they will be killed — we will have learned nothing from the most expensive experiment in organisational delivery that most of us will ever witness.
What Turned Out to Be Optional
The spring revealed a pattern so consistent across programmes and sectors that it deserves to be stated plainly: a remarkably large proportion of programme delivery infrastructure exists not to manage complexity but to distribute accountability so thinly that no single person bears the consequence of delay.
Consider what actually dissolved.
Multi-stage approval. Before March, a typical change to programme scope or budget required sign-off through three or four layers, each adding between two and six weeks. In April, the same decisions were made in one meeting or, more often, in a single email thread. The constraint was not legal or regulatory — it was procedural, maintained because the effort of dismantling it exceeded the pain of living with it. The emergency inverted that equation.
Dependency governance as theatre. The pre-pandemic dependency management process in most large programmes is a form of performance. Dependencies are logged, categorised, RAG-rated, and presented in boards where the act of presentation is confused with the act of management. What spring demonstrated was that when you reduce the number of parallel workstreams — because you have been forced to prioritise — and put the relevant owners in the same conversation, the dependency register becomes redundant. The dependencies do not vanish; the intermediary layer that existed to track them does.
Consensus as decision-making. We discovered, under duress, the difference between consultation and consensus. Pre-pandemic governance treated them as identical: a decision was not made until every stakeholder had registered their comfort. What the emergency produced was consultation on a timeline — views were heard, but someone decided, and the decision stuck. The quality of decisions did not decline. In several cases it visibly improved, because the decision-maker could no longer hide behind the collective.
Scope permissiveness. Perhaps the most revealing casualty of the spring. When the world narrowed to one priority — keep running — programmes discovered that a significant portion of their scope had been accumulated through accretion rather than design. Features, capabilities, and deliverables that had survived every prioritisation exercise suddenly fell away when prioritisation became real. No one missed them. In one programme, a workstream that had consumed roughly fifteen per cent of the delivery budget for two years was paused in March and by June no stakeholder had asked when it would resume.
Ceremony for its own sake. Stand-ups that existed because the methodology said so. Retrospectives conducted as liturgy. Status reports written for audiences who never read them. Spring 2020 did not kill useful ceremony — teams that needed daily coordination still coordinated daily — but it exposed how much ceremony was performed for compliance rather than value.
The Mechanism Was Clarity, Not Crisis
It would be easy — and it is already becoming the preferred narrative — to attribute all of this to the heroism of people working under extraordinary pressure. The story flatters everyone: leaders were decisive, teams were resilient, the organisation rose to the occasion. It is not wrong, exactly, but it obscures the mechanism.
What drove the change was not motivation. People were not suddenly more competent or more committed in April than they had been in February. What changed was the information environment. Specifically, three things became true simultaneously that are almost never true in peacetime programme delivery.
The mechanism was not heroism but clarity. When there was one priority and everyone knew it, when consequences were visible within days rather than quarters, and when the person who bore the risk was the same person who controlled the remedy — delivery accelerated not because people worked harder but because the system stopped working against them.
There was one priority, and everyone knew what it was. Not a portfolio of twelve strategic objectives ranked in a matrix that no one had validated since the last planning cycle, but one thing: keep operating, deliver what matters, stop everything else. The cognitive load of navigating competing priorities — which in normal times consumes an astonishing proportion of programme management energy — dropped to near zero.
Consequences were visible and immediate. In peacetime, the consequence of a delayed decision is another delayed decision. The feedback loop runs in months or quarters. In spring 2020, the consequence of delay was visible within days: a service not restored, a process not adapted, a team left without direction while the world shifted underneath them. Leaders who had comfortably deferred for years found that deferral now had a face.
Authority and accountability converged on the same person. The distributed accountability structures that characterise most large organisations — in which the person who bears the risk is separated by several layers from the person who controls the remedy — compressed. Not everywhere, and not perfectly, but enough to demonstrate what happens when the person who must live with the outcome is the same person who can act on it.
These three conditions are not inherent to a pandemic. They are inherent to clarity. And clarity, unlike a pandemic, is something an organisation can choose to maintain. The question — the one we are already failing to ask — is whether we will.
A Field Catalogue of What Worked
What follows is not a theoretical framework but a practitioner’s inventory — the specific practices that emerged or were revealed between March and June, catalogued here before institutional memory begins its quiet work of reclassification.
The decision-maker-in-the-room principle. Governance meetings were restructured, often by accident, so that the person with authority to decide was present when the item was discussed. This sounds obvious. It was not the norm. The pre-pandemic pattern was to present to a board, which would escalate to an individual, who would consult, and then respond — a cycle measured in weeks. The emergency practice of putting the decision-maker in a direct conversation with the people who understood the problem cut this to hours.
Scope triage, done once and enforced. The emergency forced genuine prioritisation — not the annual exercise of ranking everything as high priority, but a binary cut: this continues, this stops. The discipline was maintained because the rationale was undeniable. Programmes that cut scope by thirty to forty per cent in March discovered that delivery velocity on the remaining scope increased by more than the reduction would predict, because the overhead of managing the deferred work had itself been a drag.
Small-team, high-trust delivery. Large programme teams fragmented into smaller units with clearer mandates. Co-location was replaced not by identical structures working remotely but by smaller groups given a bounded problem and the authority to solve it. The pattern is not new — it echoes everything we claim to believe about agile delivery — but it had been suppressed in most large organisations by the gravitational pull of centralised control.
Governance by exception. Rather than the comprehensive reporting cycle — green, amber, red, every workstream, every fortnight — several programmes shifted to exception-based governance: report when something is off track, otherwise proceed. The noise reduction was dramatic. Leaders who had spent hours reviewing status packs found they could govern more effectively by being available for the exceptions than by consuming the comprehensive view.
Direct stakeholder contact. Intermediary roles — those whose primary function was to carry information between two groups who could have spoken directly — lost their monopoly. When everyone was on the same video call, the information passed directly. The intermediary’s value shifted from transmission to synthesis, and where it did not shift, the role quietly emptied.
Rolling planning horizons. Annual and quarterly plans, already suspect, became impossible to maintain. What replaced them in the best cases was a pattern of short-horizon planning — two to four weeks forward, with a lighter six-to-eight-week outlook — that was more honest about uncertainty and more responsive to change. The irony is that this is precisely what iterative delivery methodologies have advocated for over a decade. It took a crisis to make it permissible.
How Each Will Be Killed
I write this section not as speculation but as prediction, grounded in having watched organisations recover from previous disruptions. The institutional immune system does not attack these practices directly. It does not convene a meeting and vote to restore the old governance. It operates through four mechanisms, and each of the practices catalogued above is vulnerable to at least two of them.
The comfort of precedent. As the acute phase recedes, the argument shifts from “what works?” to “what did we used to do?” Precedent requires no justification; it is self-authorising. A senior leader does not need to argue that multi-stage approval is better — they merely need to observe that multi-stage approval is “how we do things,” and the burden of proof shifts to the person advocating the emergency practice. This is how scope triage will die: not through a decision to restore the cut scope, but through an accumulation of small additions, each individually reasonable, none individually worth the fight.
Risk-aversion masquerading as prudence. The practices that worked in spring carry an inherent vulnerability: they were adopted under conditions of acknowledged emergency, which means they can be framed as appropriate only to emergencies. Governance by exception will be killed by this mechanism. Someone will point out — correctly — that exception-based governance carries the risk of missing a slow-moving problem that no one escalates. The fact that comprehensive governance also misses slow-moving problems, merely with more paperwork, will not feature in the argument.
The restoration of intermediary power. Every emergency practice that bypassed an intermediary layer created someone whose organisational relevance was diminished. Those individuals are not malicious — they are rational actors whose value depended on the information asymmetry that the emergency eroded. They will rebuild it, not through obstruction but through helpfulness: offering to “take the burden” of direct stakeholder contact, proposing to “add structure” to the rolling planning horizon, volunteering to “ensure nothing falls through the cracks” in the decision-maker-in-the-room model. Each offer is genuine. Each restores the intermediary layer. The decision-maker-in-the-room principle and direct stakeholder contact will be quietly re-intermediated within months of the acute phase ending.
Methodology compliance as proxy for quality. As the immediate pressure lifts, the methodology office — whatever it is called in a given organisation — will reassert the practices that the emergency suspended. Stand-ups will be restored to their mandated form. Retrospectives will return to their prescribed cadence. The argument will be that these practices were temporarily suspended, not permanently disproved, and that the emergency is not a valid test of their value. This is how small-team, high-trust delivery will be pulled back into centralised reporting structures: not by banning autonomy but by wrapping it in enough mandatory ceremony that the autonomy becomes notional.
“The institutional immune system does not convene a meeting and vote to restore the old governance. It operates through helpfulness, precedent, and the quiet gravity of the familiar.”
These four mechanisms are already visible. I can see them in the language shifting around me: “temporary measures,” “returning to business as usual,” “lessons learned” — a phrase that in organisational life almost always means lessons logged and then quietly lost. By the time a formal lessons-learned exercise is conducted — probably in early 2021, if the pattern holds — the institutional immune response will have advanced far enough that the lessons will be captured in their neutralised form: “we learned the importance of agility” rather than “we learned that most of our governance was waste.”
What Survival Would Require
I do not think most of these practices will survive in the organisations where they emerged. The forces arrayed against them are structural, and the constituency for preserving them — primarily the delivery teams who experienced the relief of working without unnecessary overhead — holds less organisational power than the constituency for restoration.
But if survival were the objective, it would require three things that most organisations will not choose to provide.
A named owner for each practice, with explicit authority to protect it. Not a working group, not a “transformation lead” with a broad mandate, but a specific individual accountable for preventing the reversion of a specific practice. Distributed ownership is how we lost these arguments before the pandemic forced them. We will lose them again the same way unless someone’s name is next to the line.
An honest accounting of what the emergency revealed about peacetime overhead. Not a celebration of resilience — we have had enough of those already — but a forensic examination of the work that stopped in March and was not missed by June. Every programme that cut scope should be required to justify restoring it. The burden of proof should sit with the addition, not the cut. This will not happen voluntarily; it would need to be mandated, and the people with the authority to mandate it are frequently the same people whose decision-making comfort depends on the overhead being restored.
An acceptance that clarity is a choice, not a circumstance. The single most important lesson of the spring — that the mechanism was clarity, not crisis — is the one most likely to be lost, because it implicates leadership in a way that the heroism narrative does not. If the reason things worked was that priorities were clear, consequences were visible, and accountability was concentrated, then the reason things did not work before was that leaders permitted ambiguity, tolerated invisible consequences, and distributed accountability to the point of dissolution. That is a harder truth than “our people rose to the occasion,” and it is the truth most likely to be quietly buried.
The Window Is Closing
We are, as I write this in early September 2020, approximately six months from the point at which the emergency practices of spring will be fully reclassified as aberrations rather than revelations. The institutional immune system works on roughly that timescale — I have seen it after major restructurings, after technology failures, after regulatory shocks. By the time we reach the anniversary of the first lockdown, the dominant narrative will have been rewritten: the organisation adapted heroically, learned what it needed to learn, and has since returned — wisely, prudently — to its proven ways of working.
This essay is written against that clock. Not because I believe a single practitioner’s catalogue can overcome structural incentives — it cannot — but because I believe the reversion should at least be visible. We should know what we are choosing to forget. We should name the practices that worked, record the mechanisms that will kill them, and note the moment at which we decided that the comfort of the familiar was worth more than the evidence of the spring.
The twelve weeks between March and June 2020 did not prove that organisations can deliver in a crisis. We already knew that. What they proved is that a great deal of what organisations do in peacetime is not delivery at all — it is the elaborate maintenance of structures that exist to make the absence of delivery feel orderly.
We are about to choose those structures again. We should do so with our eyes open.