Remote Delivery Worked at Scale — but It Also Exposed What Offices Were Quietly Holding Together
The office was doing less than its defenders claimed and more than its critics noticed. Remote working is doing the same.
The Week the Programme Left the Building
On a Friday in March, a 240-person programme still had three offices, twenty-seven meeting rooms, a wall of delivery plans and a weekly rhythm built around physical presence. By the following Wednesday, almost everyone was working from home.
The programme did not stop. Within two weeks, planned releases were moving again. Daily meetings were shorter. Attendance improved. Decisions that once waited for a room were made on calls. Travel disappeared, specialists became easier to convene and teams discovered that much of what they had described as collaboration was proximity.
By June, output looked reassuring. The programme completed 91 per cent of its planned work, compared with 87 per cent in the previous quarter. Senior leaders began to say that remote delivery had been proven at scale.
Then different evidence appeared. Cross-team defects had risen by a third. Four new joiners had waited more than three weeks to become productive. A change to a customer-notification process passed through three teams without anyone recognising that the wording contradicted an operational policy. Experienced people were delivering; the system through which people became experienced was weakening.
Both stories are true.
Remote programme delivery has worked far better than many expected in 2020. It has also borrowed heavily from relationships, knowledge and trust accumulated in offices before the disruption. The mistake is to turn an emergency success into a simple verdict on location.
What Worked Was More Than Technology
The visible enablers were obvious: portable equipment, secure remote access, video calls, shared documents, digital work boards and messaging. Organisations that had already invested in them moved more easily. Those relying on fixed desktops, paper approvals or local file stores experienced a painful first fortnight.
But technology explains connection, not delivery.
The programmes that recovered fastest had three less visible assets.
Work was already legible
Teams with clear backlogs, defined ownership, testable outputs and current documentation could relocate the work because the work had a shape independent of the building. Teams that coordinated through memory and hallway conversation lost their operating system when the hallway disappeared.
Remote delivery did not create good management. It amplified whether good management already existed.
Relationships carried ambiguity
People who had worked together knew whose brief answer meant confidence and whose meant concern. They understood which architect to call, which operational manager held the real process knowledge and which supplier commitment needed careful interpretation. A ten-minute call could replace a meeting because months or years of context sat behind it.
This is why established teams often reported increased productivity while new or reconfigured teams struggled. The call was efficient because the relationship had already paid the setup cost.
The crisis simplified priority
During the early weeks, programmes were given clearer outcomes and fewer competing demands. Decisions moved because delay had visible consequences. Leaders who had tolerated a monthly escalation cycle were suddenly available daily. Functions advised together rather than in sequence.
Remote working received the credit for speed that was partly produced by simplified priorities and concentrated authority.
We did not merely send the programme home; we temporarily changed what the organisation was willing to decide.
What Did Not Work Was Easy to Hide
The first months have favoured measures that flatter remote delivery. Completed tasks, releases, meeting attendance and short-term service continuity are visible. Erosion is quieter.
Coordination became scheduled
Within a team, remote practice often improved. Daily calls were disciplined. Written decisions were easier to find. Work in progress became more visible.
Between teams, coordination became a calendar problem. The integration lead who previously overheard a conversation now joined only when invited. The operations expert who could correct an assumption in thirty seconds saw the design after it had hardened into a decision.
In the composite programme, individual team throughput increased by 11 per cent. Defects found during end-to-end testing increased from 18 to 24 per release. The teams had become locally efficient while the programme became less coherent.
That pattern matters because programmes exist precisely where work crosses boundaries. Optimising the team while weakening the joins is not delivery at scale. It is deferred integration.
Meetings multiplied to replace visibility
Many organisations responded to distance by adding calls: daily team meetings, cross-team coordination, leadership updates, welfare sessions and virtual drop-ins. Calendars filled because nobody could see whether work or people were drifting.
This created a new form of presenteeism. Instead of occupying the office, people occupied the call. The working day fragmented into thirty-minute units, leaving difficult thinking for early mornings and evenings.
The strongest remote teams did the opposite. They wrote more, reduced routine attendance and reserved live time for ambiguity, conflict and decisions. They treated synchronous attention as a scarce resource.
Unequal circumstances became operational facts
A remote programme does not operate in one environment. It operates in hundreds of homes with different space, care responsibilities, connectivity, privacy and equipment. A policy that assumes uniform conditions transfers the cost of delivery invisibly to individuals.
The people most able to absorb that cost may appear most committed. Those who protect boundaries may appear less responsive. Without care, performance judgement begins to reward domestic circumstance rather than contribution.
This is not a peripheral welfare issue. It shapes availability, retention, error and who receives consequential work.
Learning became extraction
Established staff could continue to draw on what they knew. New joiners received scheduled introductions, documents and task lists. What they did not receive was the low-risk exposure through which judgement develops: hearing a difficult conversation, watching a senior colleague reframe a problem, asking a question without booking time.
One team onboarded twelve people remotely. All completed the formal induction within ten days. After six weeks, only five could resolve a routine exception without escalation. The induction had transferred information; it had not transferred situated knowledge.
Programmes can consume their accumulated social capital for months while reporting stable output. By the time the loss becomes visible, the people who notice and repair weak judgement are already overloaded.
The Case for the Office Has Been Overstated — and Understood Too Narrowly
The strongest argument for returning is that physical presence enables creativity, trust, learning and rapid coordination. There is truth in it. Co-location carries information that formal systems omit.
But the office was never uniformly collaborative. It also produced interruption, travel, status signalling, exclusion from informal networks and a false preference for people who were visible. Distributed specialists were routinely treated as secondary participants. Deep work competed with noise. Meetings existed because rooms were booked.
A return to the old pattern would preserve benefits and restore dysfunction together.
The opposing claim — that remote delivery has proven offices unnecessary — is equally thin. It confuses continuation with renewal. An established programme can continue remotely by drawing on relationships and routines formed elsewhere. That does not prove it can indefinitely create new teams, develop judgement, negotiate difficult boundaries and renew trust in the same way.
The useful question is not office or remote? It is what kind of work is being done, and what conditions does it require?
| Work | Remote strength | Remote weakness | Deliberate response |
|---|---|---|---|
| Focused production | Fewer interruptions, wider access to specialists | Isolation and fragmented calendars | Protect meeting-free time |
| Routine team coordination | Visible boards and short calls | Activity can replace outcome | Use written status and live exceptions |
| Cross-team integration | Easier to convene distributed roles | Weak signals and uninvited expertise disappear | Schedule interface reviews and rotate observers |
| Complex decision | Rapid access to evidence and advisers | Private influence can become opaque | Record advice, dissent and authority |
| Learning and onboarding | Repeatable material and broader access | Judgement and context transfer poorly | Pair people on real work and create observation |
| Conflict and trust repair | Time to prepare and reduce theatre | Nuance is harder to read | Prefer richer, sustained conversation; meet physically where possible |
What We Forgot About Programme Work
Programmes are often described as collections of plans, projects, risks, benefits and governance. That description misses their social nature.
A programme is also a temporary network of commitments. People agree to make decisions with incomplete evidence, expose problems before they are safe, lend scarce expertise and accept short-term inconvenience for a shared outcome. These commitments cannot be reduced to tasks.
The office supported some of this network accidentally. A delivery lead learned that another team was struggling. A sponsor’s concern became visible before it became an instruction. A disagreement continued after the formal meeting until the real issue emerged.
Remote delivery has forced these interactions into designed channels. That is valuable because accident was an unreliable operating model. It is also dangerous because we tend to design what we can count: calls, messages, documents and tasks. Trust, apprenticeship and weak signals do not fit easily into a dashboard.
The programme manager’s task therefore changes. It is no longer enough to ensure that every team can work remotely. The programme must deliberately maintain the connective tissue between teams.
That means asking:
- Where does essential tacit knowledge sit?
- Which relationships carry the most consequential ambiguity?
- Which interfaces have lost informal observation?
- Who is producing while becoming progressively more isolated?
- Which new joiners can complete tasks but cannot yet exercise judgement?
- Where has written clarity improved, and where has documentation become a substitute for conversation?
- Which decisions require richer human interaction than a short call provides?
These are delivery questions, not cultural decoration.
Productivity Is the Wrong Argument
Much of the debate is becoming a contest over productivity. One side points to sustained output and reduced travel. The other points to fatigue, weak collaboration and the difficulty of innovation.
Productivity is too blunt to settle the matter. It compresses different kinds of work and different time horizons into one number. A team can produce more this month while consuming the relationships that make next quarter possible. Another can spend time building shared understanding that looks unproductive until a difficult integration decision arrives.
The better distinction is between production, coordination and renewal.
- Production turns known work into output.
- Coordination aligns work across boundaries.
- Renewal creates the relationships, capability and judgement needed for future work.
Remote delivery has often performed strongly in production, unevenly in coordination and weakly in renewal. That is not a permanent verdict. It is a warning about what must be designed.
“A programme can meet every short-term milestone while quietly spending the trust and knowledge on which the next milestone depends.”
A More Honest Operating Model
The answer is unlikely to be a universal policy. Programmes need an explicit operating model built around the work.
Preserve what has improved:
- Written decisions and visible work
- Shorter, outcome-led meetings
- Equal access for distributed participants
- Reduced travel and broader specialist participation
- Greater autonomy for clear, bounded work
Repair what has degraded:
- Create named forums for cross-team interfaces, not more general status calls.
- Pair new joiners with experienced practitioners on real exceptions.
- Protect blocks of uninterrupted work.
- Make workload and domestic constraints discussable without turning them into tests of commitment.
- Bring people together, when feasible, for formation, difficult design, conflict and trust renewal.
- Review the operating model monthly using integration defects, decision delay, onboarding time and concentration of knowledge — not task completion alone.
The point is not compromise for its own sake. It is precision. Use distance where distance improves the work. Use presence where presence adds information, trust or judgement that the programme cannot otherwise create.
What the First Half of 2020 Has Actually Proven
Remote delivery at scale is possible. That is no longer a serious question.
What remains unsettled is whether organisations can sustain it without mistaking inherited social capital for a renewable resource, local team productivity for programme coherence, and continuous availability for commitment.
The most impressive achievement of recent months is not that people remained productive outside offices. It is that programmes adapted while operations, households and public life were under extraordinary strain. That achievement deserves respect, not a simplistic mandate for one permanent model.
We should keep the autonomy, written clarity and freedom from unnecessary travel. We should stop the calendar saturation, invisible domestic subsidy and extraction of experienced people’s knowledge. We should remember that delivery is not only the movement of tasks but the formation of collective judgement.
The office was doing less than its defenders claimed and more than its critics noticed. Remote working is doing the same. The programme leader’s responsibility is to see both before emergency practice hardens into unexamined doctrine.