Delivery Methods Are Risk Allocations, Not Org Charts
Owners often encounter delivery methods as diagrams — boxes and arrows showing who contracts with whom. The diagrams are accurate and almost entirely beside the point. What a delivery method actually does is allocate three things between the owner and the parties building the project: risk, control, and information. Every method gives the owner more of one by taking away another, and the trade is permanent once the contracts are signed.
This is why the question "which delivery method is best" has no answer, while the question "what does this project need the owner to keep, and what can it afford to give away" always does. The three dominant methods — design-bid-build, construction manager at risk, and design-build — are best understood by what the owner surrenders in each.
Design-Bid-Build: Full Design Control, Full Gap Exposure
Design-bid-build is the traditional sequence. The owner contracts separately with an architect to complete the design, then bids the finished documents to contractors and awards to the lowest responsive bidder. It remains the default for many public owners, and for good reason: it maximizes design control and produces a transparent, price-competitive award that procurement statutes were written around.
What the owner trades away is integration. The contractor prices documents it had no hand in producing, and the seam between design and construction belongs entirely to the owner. Every gap, ambiguity, or error in the drawings is the owner's to pay for through change orders — and a low-bid contractor operating on thin margin has every structural incentive to find those gaps. The owner also trades schedule: nothing overlaps, so the project takes as long as design plus procurement plus construction, sequentially.
The method also gives the owner no price information until bid day. If bids come in over budget after two years of design, the owner redesigns and rebids, on the owner's time and money. Design-bid-build works well when the design is genuinely complete, the scope is conventional, and the owner's team can administer the seam. It punishes incomplete documents and optimistic schedules.
CM-at-Risk: Early Builder Input, Blurred Accountability
Construction manager at risk brings the builder in during design. The CM advises on cost, constructability, and schedule while the architect works, then converts to a contractor by committing to a guaranteed maximum price, typically before design is fully complete. The owner keeps a direct contract with the architect and gains a professional partner instead of an adversarial low bidder.
The trades are real, though quieter. The owner gives up the clean price competition of a hard bid: the GMP is negotiated with one firm, built on estimates and allowances for work not yet designed, and its "guarantee" is only as strong as the assumptions and exclusions written into it. Owners routinely discover that a GMP set at sixty percent design contained enough qualifications to function as an estimate. Contingency ownership, savings splits, allowance reconciliation, and self-performed work pricing all become negotiated terms the owner must actively manage rather than market outcomes the owner passively receives.
Accountability also blurs. When the CM helped shape the design it later builds, the crisp design-bid-build answer to "whose fault is this gap" softens. The method rewards owners who audit the GMP line by line, hold the CM's preconstruction advice to a professional standard, and keep independent cost estimating alive throughout design — the core of disciplined procurement and financial management.
Design-Build: Speed and a Single Throat to Choke, at the Price of Control
Design-build places design and construction under one contract. The owner writes performance requirements, selects a design-builder, and receives a single point of accountability for the finished project. Schedule compresses because design and construction overlap, and the design-error change order largely disappears — the entity that drew it also built it.
What the owner trades away is design control, and the trade happens earlier than most owners expect. Once the design-build contract is signed, the owner's design influence is limited to what the bridging documents and performance criteria captured. Everything unstated becomes the design-builder's choice, and the design-builder's incentive is to satisfy the criteria at the lowest cost. Owners accustomed to iterating with their architect find that in design-build, every refinement after award is a change order, and the architect now works for the builder rather than for them.
The owner also trades information. Cost and design decisions happen inside the design-build entity, visible to the owner only through the reporting the contract demands. The method works brilliantly when requirements can be fully specified up front — which is why it dominates transportation and infrastructure work, where performance criteria are objective — and struggles when the owner's needs will evolve, as they do in complex civic and institutional buildings.
The Public Owner's Added Constraints
Public owners choose among these methods inside a procurement framework they do not control. Statutes may require competitive sealed bidding, restrict qualifications-based selection, mandate specific alternative-delivery approval processes, or prohibit certain methods outright. Protest exposure disciplines every evaluation decision. Transparency obligations mean scoring criteria, negotiations, and price build-ups may become public record.
These constraints are often described as burdens, but they are better understood as design parameters. A public owner selecting CM-at-risk must build an evaluation process that can survive a protest; one selecting design-build must invest heavily in bridging documents, because the procurement rules will limit its ability to negotiate its way out of vague criteria later. The delivery decision and the procurement strategy are one decision, and they should be made together, early, with counsel and owner-side advisors at the table.
The Method Only Works If the Owner Works It
The least discussed truth about delivery methods is that each one assumes an owner behaving in a particular way. Design-bid-build assumes an owner who enforces document quality and administers claims firmly. CM-at-risk assumes an owner who negotiates the GMP skeptically and manages open-book costs continuously. Design-build assumes an owner who invests in requirements before award and holds performance criteria rigidly after it.
When projects delivered under any method go wrong, the post-mortem usually finds not that the method failed but that the owner's side was not staffed to play its role. Contract administration, change management, schedule oversight, and claims posture — the substance of contract administration and risk management — are what keep the chosen allocation of risk from silently reallocating back to the owner one change order at a time.
Choosing with Eyes Open
The delivery decision deserves to be made deliberately, before schedule pressure makes it by default: an honest assessment of the project's scope certainty, the owner's tolerance for design compromise, the procurement rules in play, and the owner-side capacity available to work the method chosen. If your organization is approaching that decision, independent owner's advisory can map the trade-offs against your specific project and help structure the procurement so the method you choose performs the way its diagram promises.
