A change that costs an hour at concept stage can cost weeks once construction drawings are issued. Knowing what's supposed to be decided when is how a client avoids paying twice for the same decision.
Concept design: the decisions that are cheap to change
At concept stage, the real deliverables are the building's footprint, massing, general room layout, and its relationship to the site, orientation, entry points, and how it sits against neighbouring structures. Nothing is drawn to a buildable level of detail yet, which is exactly why this is the cheapest stage to make major changes.
This is the point to resolve the big trade-offs, view versus thermal comfort, open plan versus separated rooms, single storey versus two, because changing any of these later means redrawing structure, not just aesthetics.
Clients sometimes want to skip ahead to material and finish decisions at this stage, but that's premature: a finish selected before the room it's in has a settled layout is a decision that often has to be revisited anyway once the space is actually resolved.
Preliminary design: structure and systems get their shape
This stage is where the structural system, the mechanical and electrical strategy, and the material palette are broadly fixed, not final in every specification, but fixed enough that the building's basic construction logic is set.
Structural sizing at this stage still has some flexibility, but a column grid or a floor-to-floor height agreed here is expensive to revisit later, because everything downstream, permit drawings, quantity estimates, contractor pricing, starts building on top of it.
This is also the last comfortable point to change the building's footprint or overall form without meaningful rework, and the point where a realistic cost estimate first becomes possible, since the major cost drivers, area, structural system, finish tier, are now known even if not every detail is.
Construction drawings: precision becomes mandatory
Construction drawings translate the fixed design into instructions a contractor can price and build from, dimensions, junction details, material specifications, and coordination between structural, mechanical, and architectural drawings all have to be resolved and internally consistent.
A change requested here almost always cascades: moving a wall means checking the structural drawing, the electrical layout, the mechanical ductwork, and the finish schedule, all of which reference each other. This is why a change that would have taken an hour at concept stage can take days here, purely in redrawing and re-checking, before a single thing has moved on site.
Once construction drawings are issued for permit or tender, they should be treated as the point of no easy return. Every project we run makes this explicit to the client in writing, precisely because the cost asymmetry between stages is the single biggest lever a client has over their own budget.
Why front-loading design time pays for itself
Clients under schedule pressure often want to compress concept and preliminary design to reach construction drawings faster, reasoning that drawings are the 'real' progress. In practice this usually moves problems downstream rather than eliminating them, an unresolved layout question doesn't disappear when construction drawings start, it just gets discovered later, at a stage where fixing it costs far more.
The projects that run smoothest are consistently the ones where the client resisted that pressure and let concept and preliminary design run their full course, even when it felt slow at the time. The schedule saved at the front rarely survives contact with a change order at the back.
This is a genuine trade-off worth naming to a client early: a longer design phase with a more stable construction phase, versus a shorter design phase with a higher chance of costly mid-construction changes. Both are legitimate choices, but only one of them should be a surprise.
A simple rule for when to lock a decision
The practical test we use with clients: if a decision affects the structural grid, the building envelope, or the position of a wet room (kitchen, bathroom, anywhere with plumbing), it needs to be locked by the end of preliminary design. If it affects only a finish, a fixture, or a colour, it can reasonably wait until construction drawings or even into the tender stage.
This test also explains why some 'small' late changes are expensive and some 'big' late changes are cheap: swapping a floor tile late is genuinely low-cost, while moving a bathroom two metres late is expensive regardless of how minor it looks on a plan, because it touches plumbing, waterproofing, and often structure.
Sharing this test with a client at the start of a project, rather than explaining it after an expensive change request, is one of the simplest things that improves how a project actually runs.