Why design in a small practice can’t be delegated.

Every small architecture practice reaches the same inflection point. Work increases, the director is stretched, and the logical response is to delegate. But delegation stalls, work comes back wrong, and the director ends up redoing what they handed off in the first place.

The instinct is to blame capacity, or team capability, or poor communication habits. In most cases, the real problem is simpler and more structural: the design concept hasn’t left the director’s head yet, and until it does, delegation has nowhere to land.

This is not a management failure.

It is a knowledge transfer problem – one that small practices are particularly vulnerable to because design authority, client relationships, and accumulated project intuition tend to sit with a single person rather than being distributed across a team or embedded in shared systems.

Understanding why this happens, what it costs, and how to resolve it is one of the most important things a small practice director can work on.

The Nature of Design Knowledge in Small Practices

Design thinking in a small studio operates differently from how it functions in a large firm. In a practice of fifty or more people, design intent gets codified early – through stage reports, internal reviews, and multiple layers of sign-off. In a practice of three to eight people, it rarely does.

Instead, the director carries the project. They attend the client meetings, conduct the site visits, do the early sketching, and hold the evolving logic of the scheme in their own mind. This works, until it doesn’t.

Tacit vs. Explicit Knowledge in Architecture

The philosopher Michael Polanyi described tacit knowledge as the things we know but cannot fully tell. Architecture is saturated with it. The sense that a threshold should feel compressed before it opens.

The instinct that a material choice is wrong without being able to immediately say why. The understanding of what a client actually means when they say they want something “calm”.

Explicit knowledge, by contrast, is documentable and transferable – dimensions, specifications, regulatory requirements.

Most of architectural practice sits somewhere between the two, but architectural concept development sits firmly in the tacit category, especially at the early stages when ideas are still forming.

This matters because tacit knowledge cannot simply be delegated. It has to be translated first.

Why the Director Becomes the Concept’s Custodian

It is rarely a conscious choice. The director is involved in every client conversation, so they absorb nuances the team never hears. They do the initial site visit and register things intuitively that they never write down. They sketch over breakfast and refine ideas in the shower.

By the time a project reaches the point where junior staff could usefully contribute, the director has already built an internal model of what the design should be – and they assume, without realising it, that the team shares that model.

They don’t. And that gap is where delegation fails.

What “The Concept Living in Your Head” Actually Means

When a concept hasn’t been externalised, there are predictable signs. The director struggles to brief the team with any specificity. Junior staff make design decisions that seem reasonable in isolation but consistently miss the point. Work comes back looking competent but feeling wrong.

This creates a feedback loop that is exhausting for everyone involved. The director corrects, the team resubmits, the director corrects again – and the original concept drifts further from the work being produced.

The Invisible Brief Problem

Junior architects and technicians working without a fully communicated design intent are not working from the brief – they are working from their best guess at the brief. They fill the gaps with their own assumptions, their previous experience, and what seems logical given the information they have.

This is not a failure of ability. It is a structural problem: they cannot make the right decisions because they haven’t been given the right information. Understanding how a design brief should be constructed and communicated is fundamental to avoiding this situation.

The result is rework that consumes time, erodes morale, and leaves both parties frustrated without either fully understanding why.

How Directors Unknowingly Withhold Information

Most directors who do this are not withholding deliberately. They are assuming shared context that doesn’t exist. They say “make it feel more considered” without explaining what considered means in this project’s specific terms.

They mark up a drawing with corrections but don’t explain the design logic behind each change.

They give the team the output of their thinking without the thinking itself. And so the team learns to produce work that gets corrected, rather than developing the judgment to produce work that doesn’t.

Why Premature Delegation Fails

Delegation works when the person receiving the task has enough context to act on it independently. In design, that means understanding not just what needs to be drawn, but why the design takes the form it does and what principles should govern decisions when new problems arise.

Without that context, premature delegation doesn’t save the director time. It creates work.

The Correction Loop and Its Cost

The correction loop is the defining symptom of premature delegation. The director hands off a task, the team produces something, the director identifies problems and corrects them, the team revises, and the cycle begins again.

Each iteration takes time. Each correction requires the director to re-engage with work they thought they’d delegated.

By the third or fourth round, it would have been faster for the director to have completed the task themselves – and they know it, which makes the next delegation attempt even less confident.

The design process is genuinely iterative by nature, but iteration driven by unclear intent is not productive iteration. It is just repetition.

The Effect on Junior Staff Confidence and Growth

Repeated correction without conceptual framing is damaging in ways that go beyond project efficiency. Junior architects who consistently receive work back marked up without explanation learn to be cautious rather than creative.

They produce safe, minimal interpretations of the brief because taking design risks has proven unreliable.

Over time, this stunts the development of design judgment – the very thing the director needs the team to have in order to delegate successfully. The practice becomes more dependent on the director, not less, even as the team grows.

What Externalising the Concept Actually Requires

Externalising a concept is not the same as explaining it once in a meeting. It means creating a stable, accessible, shared reference that the team can return to independently when they need to make a design decision.

This takes more time upfront, but it is time that gets paid back across every stage of the project that follows.

Design Intent Documents and Reference Sets

The most practical tools for externalising design intent are low-tech and immediate. An annotated sketch that explains not just the spatial arrangement but the reasoning behind it.

A precedent board that communicates atmosphere, material logic, and spatial hierarchy rather than simply illustrating stylistic references.

A short written design narrative – even two or three paragraphs – that articulates the concept’s drivers, its key moves, and the hierarchy of priorities that should govern subsequent decisions.

A decision log that records why key choices were made, so that when conditions change, the team can problem-solve within the same conceptual framework.

These documents don’t need to be polished. They need to be honest and specific. A rough sketch with annotations is more useful than a beautifully formatted presentation that says nothing about intent.

The Concept Conversation as a Handover Tool

Before any significant delegation, a structured verbal walkthrough of the concept – covering its origin, its constraints, its non-negotiables, and the decisions still open – can transfer a remarkable amount of tacit knowledge in a short time.

The key is structure. Without it, concept conversations tend to cover what the design looks like rather than why it looks that way.

The director should be able to articulate: what the concept is trying to achieve, what constraints shaped it, what is fixed and what is still open, and how the team should handle decisions that fall outside the briefed parameters.

A well-run concept handover conversation is not a presentation. It is a dialogue, with the team asking questions and the director refining their own articulation in response.

Knowing When the Concept Is Ready to Leave Your Head

There is a practical test. Ask a team member to explain the design rationale back to you without prompting. If they can – not word for word, but accurately and in their own terms – the concept has been sufficiently externalised. If they can’t, it hasn’t.

Other signals include: design decisions being made consistently without escalation, the director’s correction rate reducing over successive iterations, and team members being able to identify when a proposed change conflicts with the concept’s intent.

None of this requires perfection. It requires the concept to be stable and shared enough that the team can carry it forward without the director being present at every moment of decision.

How to Build a Delegation-Ready Design Culture in a Small Practice

Shifting from a director-as-sole-custodian model to one where the team can genuinely carry design intent forward is not a single action. It is a series of embedded habits that, over time, reduce the gap between what the director knows and what the team knows.

For directors considering the broader structure of how a small practice operates, understanding the responsibilities that come with practice ownership is a useful frame for approaching this kind of long-term change.

Creating Repeatable Handover Rituals

Rituals make handover consistent. A design review at the end of concept stage that requires the team to present their understanding of the intent – not the drawings, but the thinking behind them – builds a habit of shared articulation.

A structured briefing format for each new delegation – however brief – that always covers intent, constraints, hierarchy, and escalation triggers removes the ambiguity that causes most handover failures. These don’t need to be elaborate. They need to be repeated.

Developing Team Design Literacy Over Time

The fastest way to accelerate future delegation is to bring the team into the moments where design intent is formed.

Shared site visits, client meetings where appropriate, and sketch review sessions are not inefficient use of team time – they are investments in the background knowledge that makes future delegation faster and more reliable.

A junior architect who has heard the client describe their needs directly, who has stood on the site in the same weather as the director, and who has watched the concept develop through iterative sketching has a fundamentally different understanding of the project than one who only ever receives the finished brief.

Case Studies

Caruso St John Architects

Caruso St John operate as a small London practice with an unusually particular design sensibility – one rooted in material specificity, historical reference, and a scepticism of formal novelty.

Maintaining this sensibility across projects requires close principal involvement at key design stages, particularly where material and detail decisions carry the most conceptual weight.

What distinguishes their approach is that their design philosophy is not secret. It has been articulated through writings, lectures, and published work over many years – giving the team a rich body of explicit reference material that supports tacit knowledge transfer.

The concept can live beyond the principal’s head because it has been externalised, in part, through consistent public articulation.

Lacaton and Vassal

Anne Lacaton and Jean-Philippe Vassal have developed one of the most communicable design philosophies in contemporary architecture: never demolish, always add, maximise inhabitable space, use the most generous budget of free space possible. These principles are not vague values – they are operational instructions.

This clarity means that collaborators working within the studio can understand and apply the principles to new problems without requiring constant direction.

The concept framework is genuinely externalised – not as a style to be imitated, but as a set of priorities that can guide decision-making independently. This is what a well-externalised concept looks like in practice.

Studio Mumbai (Bijoy Jain)

Bijoy Jain faces one of the most demanding versions of this challenge. His design process is deeply intuitive, rooted in material investigation, craft knowledge, and site-specific sensory response – forms of knowing that resist documentation almost entirely.

His solution is to externalise knowledge through making rather than writing. Craftspeople at Studio Mumbai learn the design intent through direct collaboration on material samples, 1:1 mock-ups, and extended site presence.

The concept is communicated through shared physical experience rather than documents – a model that is labour-intensive but highly effective for communicating non-verbal design knowledge to a large team.

Assemble

Assemble operate as a collective, deliberately resisting the model of the single design author. Concept ownership is distributed across the whole team from the outset, which means the question of “getting the concept out of someone’s head” is structurally pre-empted.

For small practices, Assemble’s model is instructive even if not directly replicable.

Their approach demonstrates that shared concept development – where multiple people are present for the formation of design intent rather than being briefed after the fact – produces teams that require less correction and can problem-solve with greater autonomy.

The cost is decision-making speed; the benefit is genuine collective understanding.

Peter Zumthor’s Atelier

Peter Zumthor represents the opposite extreme. His design philosophy is intensely personal, and his practice is structured to keep concept ownership tightly held by the principal.

Work proceeds slowly, and Zumthor’s direct involvement extends into details and decisions that most principals of equivalent stature would have delegated long ago.

The output is exceptional. The trade-off is clear: the practice produces relatively few projects, and the studio’s design capability is structurally dependent on Zumthor’s continued involvement.

As a model for practice structure, it demonstrates that tightly held concept ownership can produce extraordinary work, but only if the principal has both the capacity and the intention to remain the primary design intelligence across every project indefinitely.

Common Mistakes Directors Make When Trying to Delegate Design

Most of the mistakes directors make when delegating design stem from the same root cause: they hand off the task rather than the thinking. Understanding the specific forms this takes makes it easier to catch and correct.

Delegating Tasks Instead of Intent

There is a fundamental difference between asking someone to draw the west elevation and asking them to develop the west elevation in a way that expresses the building’s relationship to the street – and then explaining what that relationship is and why it matters.

The first is a task. The second is a delegation of design intent. Only the second gives the team member enough context to make good decisions when the task turns out to be more complex than anticipated – which, in architecture, it always does.

Using Correction as a Substitute for Briefing

A common pattern: the director briefs briefly, work comes back wrong, the director corrects it. The correction contains more design information than the original brief.

The director treats the correction loop as a normal part of the process, not recognising it as evidence that the upfront communication failed.

Correction is necessary. Using it as the primary vehicle for communicating design intent is expensive, demoralising, and – because corrections are often reactive rather than comprehensive – tends to produce work that fixes the immediate problem without developing the team’s understanding of the broader concept.

Delegating Too Late in the Process

Directors who hold the concept through early design and only externalise it when detailed design or technical production begins leave the team in an impossible position.

They are being asked to execute decisions they don’t fully understand, in a phase where the consequences of misunderstanding are most expensive to fix.

Effective delegation starts at concept stage, not at the point when the director runs out of time. The RIBA Work Stages offer a useful structure for thinking about when different types of knowledge transfer should happen across a project’s life cycle. Externalising intent at Stage 2 makes every subsequent stage more reliable.

Conclusion

The central problem of delegation in small architecture practices is not organisational. It is epistemological. The director knows things the team doesn’t, and until those things are translated into a form the team can access, understand, and act on independently, delegation will keep producing disappointing results.

The answer is not to delegate less, or to hire more experienced staff, or to work harder at giving feedback. The answer is to treat the externalisation of design intent as a distinct and necessary part of the design process – one that requires time, attention, and a set of deliberate habits.

When a concept genuinely exists outside the director’s head – when it can be explained by a junior architect, referenced in a document, illustrated through a precedent board, and used to make consistent decisions without escalation – delegation becomes possible in a way it wasn’t before.

The moment to delegate is not when the director is too busy. It is when the concept is ready.