Very early in a project, a small number of people sit with a site plan and decide, provisionally, which way a building faces and roughly what shape it is. There are usually three or four of them. They have less information than anyone who will work on the project afterwards, and less time. What they settle in those weeks will govern the building's efficiency, its running cost and much of its character for the whole of its life.
That is the paradox of concept design: the decisions with the longest consequences are made when the least is known and the fewest people are involved. It is also, reliably, the phase clients compress, because at that point the project feels open, nothing has been committed, and moving quickly seems free.
It is not free. By the end of concept design the building's massing, orientation, structural logic, plan depth and efficiency are effectively settled. Everything after that is development. A project that arrives at technical design with a weak concept does not get better; it gets documented.
What concept design actually decides
The phase produces drawings that look provisional and decisions that are not.
Massing determines the building's relationship to its plot, its neighbours and the sun, and it sets the envelope area that will have to be paid for and cooled. Orientation determines how hard the building works against the climate for its entire life. Plan depth determines daylight, ventilation and the ratio of usable area to circulation. The structural strategy, even in outline, determines the grid, which determines unit layouts, which determines what can be sold. Core position determines circulation efficiency on every floor.
Each of these is taken with limited information, under time pressure, by a team at its smallest. None of them looks like a commitment at the time.
Why compression is expensive later
The relationship between when a change is made and what it costs is not linear.
Revising massing during concept design is a design exercise. The same revision at spatial coordination invalidates the work of every other discipline, because structure, services and façade have all been developed against the previous arrangement. At technical design it invalidates documentation. During construction it becomes a variation, with cost and programme consequences visible to everyone including the client's financiers.
A week saved at concept can therefore be repaid many times over, and the repayment falls at the point in the project with the least tolerance for it. This is well understood in principle and routinely disregarded in practice, because the cost of compression is invisible at the moment the decision is taken and obvious eighteen months later.
The commercial argument for a separate appointment
If concept design carries this much weight, there is a reasonable case for treating it as a distinct commission rather than the first slice of a larger one.
A separate concept appointment does several useful things. It gives the phase its own scope, fee and programme, so it is resourced for what it decides rather than absorbed into an overall fee where it competes with production work. It lets a client test a practice on the part of the work where judgement matters most, before committing to a multi-year relationship. And it produces a defined output that can be taken to market, to a lender or to an authority, and on which subsequent decisions can be based.
It also permits a genuine option: appointing one practice for concept and another for delivery. That is a common structure on large developments, where the identity of the building matters to a client who also needs delivery capacity at a scale a design-led practice may not hold. It works when the transition is planned. It fails when the design intent is assumed to transfer with the drawings, which it does not, because much of the intent lives in judgement about why things are the way they are.
We hold pre-concept and concept design as a distinct capability for these reasons, separate from architectural design, rather than as the opening stage of a single undifferentiated appointment.
What good concept work involves
The phase is not a search for an image. It is a structured narrowing of options against criteria the client has agreed.
That means testing several massing strategies rather than developing the first plausible one. It means understanding the site properly, including orientation, prevailing wind, adjacent development, access and whatever the local framework permits. It means establishing the commercial logic, because a residential concept that cannot deliver the required unit mix is not a concept. And it means bringing structural and services input in early enough to know whether the idea stands up, without letting engineering constraint drive the architecture before it has been formed.
It also means being explicit about what has been decided. A concept package that presents an attractive proposal without stating the assumptions behind it hands a client a picture rather than a position.
Pre-concept: the stage before the stage
There is useful work available even before concept design, on projects where the brief is not yet settled.
Pre-concept work tests what a site can support, what a brief implies and whether the project as imagined is viable in the form proposed. It may examine several development options, test densities against the plot, or establish whether the client's requirements and the site's capacity are compatible at all.
This is the least expensive work in a project and among the most consequential, because it is the last point at which the fundamental proposition can change cheaply. A client who discovers at concept stage that the brief does not fit the plot has already spent more than they needed to in order to learn it.
Common misconceptions
"Concept design is the phase where nothing has been committed yet"
By the end of it, massing, orientation, structural logic, plan depth and efficiency are effectively settled. Everything after that is development, not creation.
"Compressing concept design saves time"
It buys a decision deferred, not avoided. The cost of compression is invisible at the moment it is taken and obvious eighteen months later.
"A concept package that looks attractive has done its job"
A package that presents an appealing proposal without stating the assumptions behind it hands a client a picture rather than a position.
"Pre-concept work is a cost that can be skipped to save fees"
It is the least expensive work in a project and among the most consequential, because it is the last point at which the fundamental proposition can change cheaply.
What a client should ask for
- How many distinct options will actually be tested, rather than presented?
- What site and climate analysis will inform the massing?
- Which other disciplines will be involved, and when?
- What commercial criteria will the concept be tested against?
- What assumptions will be documented alongside the drawings, so a later team knows what was decided and why?
The last of these matters most where concept and delivery sit with different practices. A concept package that records its reasoning survives the transition. One that records only its output does not.
The proportion worth remembering
Concept design typically occupies a small fraction of an architectural programme and a small fraction of the fee. It determines a substantial majority of the building's performance, efficiency and character. That asymmetry is the entire argument, and it does not require exaggeration to be persuasive.
For where concept design sits within the wider process, see the RIBA stages, which sets out what each phase decides and locks. For how scope translates into fee, see architect fees in Dubai, and for the appointment structures a concept commission can sit within, choosing an architecture firm in the UAE. The beginning of a project is exactly when this is worth discussing. Start a conversation.