Why the AE Industry Is Shortchanging Schematic Design — And What It Should Look Like
The AE industry has a quiet consensus: allocate 15 to 20 percent of the design schedule to Schematic Design, then move on to the “real work” in Design Development and Construction Documents.
I think that’s backwards — and in practice, it’s even worse than the numbers suggest.
Before getting into why, it’s worth being precise about what Schematic Design actually is — because it is consistently underestimated. SD is not a sketch phase. It is not a placeholder where a few bubbles get drawn while the “real” design waits. SD is where the decisions that govern everything downstream are made: building massing and systems approach, structural concept, MEP strategy, site organization, and program validation. These are not details. They are the load-bearing choices of the entire project. Once locked in at DD and CDs, they are extraordinarily expensive to revisit.
The principle is straightforward: changes made early in design are cheap. As a project moves through DD and CDs, the cost of any given change rises sharply. By construction, a single design change can trigger cascading costs across redesign fees, change orders, material restocking, and schedule impacts. The logical response is to invest deeply in the phase where decisions are cheapest to test and refine. That phase is SD. The industry’s default schedule does the opposite.
Here is what that looks like in practice — and why the 15 to 20 percent figure is already an overstatement.
The schedule is already lying to you
When a client approves a project and the contract is executed, most AE firms are already running at 80 to 90 percent utilization on existing work. There is an unavoidable transition period before the full team is actually engaged on the new project: existing commitments need to be wrapped up, staff need to be ramped on, subconsultants need to be mobilized. That time does not come from thin air. It comes directly out of SD.
In practice, a project with a six-week SD phase may have two or three weeks of genuinely productive design work before the full team is oriented and engaged. The schedule says 15 percent. The effective reality is closer to 8.
This is not a failure of individual project managers — it is a structural problem with how the industry resources new projects. The solution is not to accept it. It is to build a schedule that accounts for it: allocate enough time to SD from the outset that the actual productive design work can happen at the depth the project requires, even after the ramp-up period.
For renovation projects, the burden is even heavier
For existing building renovation projects, the compression problem is compounded by a burden that new construction simply does not carry: before meaningful design can begin, you need to understand what you are working with.
That means field verification of existing conditions — drawings that are frequently incomplete, outdated, or simply wrong. It means investigation of concealed systems and spaces. And it means the enabling work of producing accurate as-built documentation that every discipline on the project will rely on. On a complex renovation, developing that existing building record for use by all trades can consume more time than the entire SD phase allows at 15 to 20 percent of schedule.
And that enabling work is just the entry cost. Layer on top of it the technical due diligence specific to renovation: confirming that existing horizontal and vertical distribution spaces can physically accommodate new systems; verifying that utility connections have adequate capacity for new loads; confirming that the structural system can support new equipment; assessing the condition of existing elements planned for reuse. These are not items that can be deferred to DD or CDs without significant risk.
When they are deferred — and they frequently are, because the SD schedule does not leave room for them — the consequences follow a predictable pattern. A structural adequacy question that should have been answered in SD becomes a change order when a mechanical contractor hits a conflict in the field. A utility capacity issue that should have been investigated early becomes a project-threatening finding at 60 percent construction documents. An existing element assumed to be reusable is found deficient during construction, triggering emergency redesign. These are not hypotheticals. They are the field realities of renovation projects where SD was treated as a formality rather than a phase of genuine technical investigation.
Stakeholder alignment is a process, not a meeting
One of the most consistently underestimated elements of SD is the stakeholder engagement work required to arrive at a design direction that genuinely reflects what owners and user groups need.
This is not a matter of holding a kickoff meeting and collecting a program. Real stakeholder alignment requires multiple iteration cycles: surfacing what different groups need, identifying where those needs conflict, developing design options that attempt to satisfy as many priorities as possible while making explicit which lower-priority items were set aside, and documenting clearly which tradeoffs were made and why. That documentation is critical. Owners and user groups need to understand not just what was decided, but what was considered, weighed, and deliberately set aside. Without that transparency, design decisions made in SD become contested the moment a stakeholder who wasn’t fully engaged raises objections in DD — at the worst possible time, when changes are far more expensive.
This process is iterative by nature. Each round of feedback requires time to synthesize, time to develop revised options, time to present comparisons and build consensus. A compressed SD phase handles this by truncating the iterations: fewer options evaluated, less stakeholder input gathered, less of the decision trail documented. The result is an SD deliverable that looks complete but is built on assumed alignment rather than confirmed alignment. That fragility surfaces in DD — and it is entirely predictable.
The estimating problem compounds everything
Perhaps no single issue causes more downstream damage in the SD phase than deferred estimating — and it is among the most common failures in current practice.
On many projects, the cost estimate is not completed until after SD documents are issued for owner review. The rationale is that the design needs to be sufficiently developed before it can be estimated. But this sequencing has a critical flaw: it means the owner is being asked to approve a design direction without knowing whether it is within their budget.
When the estimate arrives as an addendum to the review package rather than as an integral part of it, there are two possible outcomes. Either the estimate confirms budget alignment and the project proceeds, or it reveals a gap that requires scope reduction and value engineering. That VE process, if needed, gets pushed into DD — a phase that was not designed or scheduled to absorb it. The design team is now simultaneously developing more detailed drawings and revisiting scope decisions that should have been settled weeks earlier. The schedule slips. Costs go up. The owner, who believed the project was approved and moving forward, finds themselves mid-DD with significant decisions still unresolved.
SD is not complete until the design and the budget are both confirmed. A complete SD package presented for owner signoff means program validated, stakeholder tradeoffs documented, systems coordinated, existing conditions verified, and estimate reconciled — with scope decisions made in full awareness of cost. No loose ends. Anything less is not a completed SD phase. It is a deferred problem.
These allocations were built for pencil and paper
Here is the dimension of this problem that rarely gets examined: where the current phase schedule allocations actually came from.
The conventional SD/DD/CD breakdown was established during an era when construction documents were produced by hand. Drafters at drawing boards. Pencil, ink, and mylar. Every plan, detail, and schedule drawn manually, revised manually, and reproduced manually. Under those conditions, CD production was genuinely and necessarily the most time-intensive phase of the design process. The schedule reflected that reality.
It no longer does.
BIM and modern production tools have fundamentally changed the time equation for CD production. Coordination that once required manual cross-referencing between drawings now happens parametrically. Details replicate. Schedules auto-generate. Drawing sets update globally from a single model change. A competent, well-trained design team working in BIM can produce construction documents substantially faster than their predecessors working in hand drafting — or even early CAD.
That time has not been redistributed to SD, where it is most needed. The old percentage allocations simply persisted — inherited from a workflow that no longer exists, encoded into proposal templates, and passed down to the next generation of project managers who never had reason to question them.
The industry is not just underinvesting in SD. It is doing so based on a schedule framework designed for a technology that has been largely replaced.
A more rational breakdown
If we take seriously both the case for front-loading SD and the efficiency gains that BIM has introduced into CD production, a rational phase allocation might look something like this:
| Phase | Allocation | Rationale |
|---|---|---|
| Schematic Design (SD) | 35% | Owner alignment, stakeholder engagement, existing conditions investigation, systems coordination, integrated estimating, and iteration |
| Design Development (DD) | 25% | Refinement of a well-developed and confirmed design direction — not problem-solving from scratch |
| Construction Documents (CD) | 20% | Production against a resolved design; BIM reduces the time burden relative to hand drafting |
| Construction Administration (CA) | 20% | Sufficient oversight to protect owner interests and design intent through construction |
This is not a universal prescription. Project type, complexity, and delivery method all affect the right balance. But as a starting framework for how AE firms should be thinking about phase allocation, it reflects a more honest accounting of where the work actually lives — and where the consequences of underinvestment are felt most acutely.
The current default is a legacy artifact. The phase allocations the industry inherited from the pencil-and-paper era are not a constraint. They are a choice. And they are a choice that can be made differently.
I’m curious whether others in the AE and construction community have pushed back on this convention — or have found ways to make the case to owners and clients for a more front-loaded design schedule. What has worked? What has gotten in the way? I’d welcome the discussion.