Cost escalation is not a reason to wait. Across the 140 projects we priced this year, the owners who stayed on budget were not the ones with the most optimistic estimates — they were the ones who made a small number of deliberate decisions early and stuck with them.
Below are the five decisions we expect to separate on-budget projects from overruns in 2027. None of them require a bigger budget. All of them require early commitment, clear sequencing and the discipline to freeze scope. If you want the full data behind our view, our 2026 Construction Cost Report covers the methodology in detail.
1. Choose your delivery method before design
Most owners leave money on the table because the delivery method is picked by habit, not by evidence. Drawings are finished before a builder sees them, bids come back over budget, and nobody owns the gap between the design and the number. That is exactly the kind of gap an integrated team can close in a single quarter.
- Map every major risk — design, cost and schedule — to the party best placed to carry it.
- Compare the methods on what matters to you, not on habit alone:
- design-build, when speed and a single point of responsibility matter most;
- CM at risk, when you want your own architect and early pricing;
- design-bid-build, for simple, fully drawn scopes with little room for change.
- Give one person on your side clear authority over design decisions.
Where to start
Start with the projects you already have planned. Feasibility is the lowest-risk moment to bring in a builder, because nothing has been drawn that needs redrawing. At Brightway Logistics, early contractor involvement took 9% out of the first estimate before design development began.
A quick check for your project team
- Do we know our cost per square foot against comparable recent projects?
- Who can approve a design change above $50,000, and how often does it happen?
- When did we last check the estimate against current market pricing?
- If it was more than six months ago, start there.
- If you value-engineered across the board, check what it did to running costs.
2. Coordinate the model before you build
Every duct, beam, sleeve and hanger on your project is a decision about how the building will actually go together. Yet on most of the projects we are asked to rescue, that information lives in a dozen disconnected drawings and is only reconciled when the trades collide on site.
The owners we build for treat the coordinated model like inventory: they know where it is, who owns it, and what it costs to keep it current. They invest in a single, clash-free federated model before they invest in renderings.
That sounds unglamorous. It is also the foundation for every other decision on this list — you cannot price, prefabricate or schedule what you have not coordinated.
“We stopped arguing on site about whose drawing was right and started arguing about how to build it faster. That shift alone paid for the whole model.”
— Marcus Okafor, COO, Brightway Logistics
A practical first step is to agree on ten coordination rules and publish them. For example, we define clash_tolerance as the minimum clearance between any two systems before a clash is flagged — and we write that rule down where everyone can see it:
# bim/coordination-core.yml — single source of truth
clash_tolerance:
owner: vdc
formula: "min_clearance_mm >= 25"
refresh: weekly
rfi_turnaround:
owner: design
formula: "rfis_closed_in_7_days / rfis_issued"
target: 0.963. Secure long-lead equipment early
The disruptions of the last five years taught most owners that the cheapest supplier is rarely the cheapest option. Switchgear, transformers and steel now set the critical path on many jobs — and they are far cheaper to secure early than in the middle of a delay.
Here is how the five decisions compare on timing, payback and risk, based on outcomes across 1,240+ Corwell projects:
| Decision | Horizon | Typical payback | Execution risk |
|---|---|---|---|
| Delivery method | 3–6 months | < 6 months | Low |
| Model coordination | 6–12 months | 9–14 months | Medium |
| Long-lead procurement | 6–18 months | 12–18 months | Medium |
| Prefabrication | 4–9 months | 6–12 months | Medium |
| Field leadership | 12–36 months | 24+ months | Low |
A schedule is not a list of things you hope will happen. It is a short list of things that must happen before anything else.
Daniel Whitmore
4. Make prefabrication a production line, not a pilot
The fastest returns from prefabrication we see today are not in signature facades. They are in bathroom pods, MEP racks, wall panels and headwalls — places where work is repetitive, tolerances are clear and the baseline is easy to measure.
- Pick three assemblies with high repetition and a stable design.
- Measure the baseline for installed cost, labor hours and rework rate.
- Pilot on one floor with a cross-functional crew:
- one superintendent with authority to change the sequence;
- one VDC engineer;
- one estimator to sign off on the savings.
- Scale or stop — and publish the decision either way.
For a deeper look, read Priya Raman’s practical BIM playbook for owners.
5. Invest in the people who run the site
Every one of the decisions above depends on people who can run a site without waiting for permission. In our experience, the constraint is rarely the executive team — it is the 20 to 40 superintendents and foremen who carry the day-to-day decisions.
Give them real cost and schedule ownership, a peer group from other regions and a mentor who has run a major site before. It is the slowest decision on the list, and the one with the longest tail of returns.
The bottom line
You do not need to make all five decisions at once. You do need to choose — deliberately — which two or three you will commit to in 2027, and which habits you will drop to make room for them. If you would like a second opinion on your project, talk to our preconstruction team.
