The cost of design errors and omissions is often discussed as if it's a fixed percentage that gets applied to a budget regardless of how the project is delivered. It isn't. The dollar figure a coordination conflict or a spec-vs-drawing contradiction eventually generates is largely the same mistake either way — but who absorbs that cost, how quickly it surfaces, and whether anyone was positioned to catch it before it hit the field changes substantially depending on whether the owner self-performs the work or hands delivery to a general contractor.
What "self-performed" and "GC-led" actually mean here
The distinction matters because the two terms get used loosely. In this context, self-performed means the owner is acting in the general contractor's role directly — holding trade contracts itself, whether through a single multi-prime structure or an internal construction management team, rather than passing that responsibility to one GC that then holds all the subcontracts. GC-led means a traditional single point of contract responsibility: the owner signs one agreement with a general contractor (lump sum, GMP, or CM at-risk), and that GC holds every trade subcontract beneath it.
The delivery method changes who's contractually in the middle of a design conflict when it surfaces. It does not change where design risk originates.
Design risk starts with the owner either way
Under the Spearin doctrine — the implied warranty, dating to a 1918 U.S. Supreme Court decision, that plans and specifications an owner provides are accurate and adequate for their intended use — a contractor who builds according to the owner's design isn't responsible for the consequences of defects in that design. That principle holds under both delivery structures. Whether the owner has one GC standing between them and the trades, or is holding every trade contract directly, the design documents themselves — and the errors baked into them — are still the owner's.
What changes is what happens next.
GC-led: the GC becomes the buffer, not the fix
On a GC-led project, the general contractor becomes the first party to encounter a design conflict in the field — coordinating trade sequencing, fielding RFIs, and typically processing the change order request before it reaches the owner. That creates a buffer, but not a smaller number. The GC has no incentive to absorb a cost caused by the owner's design team, and under a lump-sum or GMP contract, a legitimate design-error change order gets priced and passed through, sometimes with markup, sometimes with schedule impact attached. The owner still pays for the design error; they just pay for it through a formal change order process that a GC's project management team is built to run, with the GC's own margin protected on the way through.
That buffer is administrative, not preventive. A GC's field team is built to process what a bad set produces — RFIs, submittals, change orders — not to catch what a bad set will produce before it's issued for construction. Preconstruction estimating and constructability input from a GC typically happens against pricing and buildability, not full cross-discipline document coordination, which is a different and narrower check than what a document review actually covers.
Self-performed: no buffer, and the coordination duty sits with the owner
On a self-performed or multi-prime project, that administrative buffer doesn't exist. The owner (or the owner's internal construction management arm) holds each trade contract directly, which means the owner also holds the coordination and sequencing responsibility that a GC would otherwise carry. Courts have found that owners on multi-prime projects have an implied duty to actively coordinate the separate prime contractors, precisely because the owner — not a GC — retains control over how each contract performs. When a design conflict surfaces between two trades on a self-performed job, there's no GC in the middle absorbing the first round of RFIs and field coordination; it lands on the owner's team directly, in real time, often during active construction.
The Spearin doctrine puts design risk on the owner regardless of delivery method. Self-performing doesn't create that risk — it removes the GC layer that would otherwise process and buffer it before it reaches the owner's desk.
That has two effects worth separating. First, cost exposure isn't smaller because there's no GC markup layered on a design-error change order — a self-performing owner can, in principle, resolve a trade-to-trade conflict for closer to the raw cost of the fix. Second, and working against that, the owner absorbs coordination and disruption risk directly: if one prime contractor is delayed or disrupted because a design conflict wasn't caught before it reached the field, the owner — not a GC — is the party exposed to that contractor's delay claim, because the owner is the one who retained the coordination duty in the first place.
The one thing that doesn't change: neither structure catches the conflict before bid
This is the part that gets lost in a self-performed-vs-GC-led comparison: neither delivery method, on its own, includes a step that checks the design documents across disciplines before the set goes out for pricing or construction. A GC's preconstruction team is checking buildability and pricing the set as drawn, not verifying that the mechanical, structural, and electrical sheets agree with each other and with the specs. An owner's internal team on a self-performed job is typically staffed for procurement, scheduling, and field coordination — not for a discipline-by-discipline document audit either. Both structures are set up to process a set once it's issued, not to catch what's wrong with it beforehand — the same gap that shows up in why "it passed QA/QC" doesn't mean the set is coordinated, regardless of who's holding the contracts once construction starts.
That's also where the two paths converge on the same conclusion. A self-performing owner has more to gain from catching a conflict before bid, because they're the one exposed to the coordination-duty claim if it surfaces mid-construction instead. A GC-led owner has more distance from the day-to-day fallout, but still pays for the fix either way, and still owns the design risk under Spearin regardless of who's standing between them and the field. In both cases, the fix is the same: a document-level review that surfaces the conflict while it's still a redline, not a change order — the difference E&O insurance coverage is built to respond to only after the fact.
Key takeaways
- The cost of design errors and omissions originates with the owner under the Spearin doctrine regardless of delivery method — self-performing or hiring a GC doesn't change who owns the design.
- On a GC-led project, the GC buffers a design conflict administratively — processing RFIs and change orders — but doesn't absorb the cost, and typically doesn't catch the conflict pre-construction either.
- On a self-performed or multi-prime project, the owner holds the coordination duty directly, which means design conflicts surface to the owner's team in real time, with no GC layer in between.
- Self-performing removes GC markup on the fix but adds direct exposure to delay and disruption claims from other prime contractors if a conflict wasn't caught before it hit the field.
- Neither delivery structure includes a step that checks the set across disciplines before bid by default — that's a separate, deliberate review either way.
Frequently Asked Questions
Does self-performing a project increase the cost of design errors and omissions?
Not the underlying cost of the error itself — that's set by the design documents, not the delivery method. What changes is who absorbs it and how directly: a self-performing owner avoids GC markup on the fix but takes on direct exposure to coordination-duty and delay claims from other prime contractors, since there's no GC buffering that risk.
Who is legally responsible for a design error on a self-performed project?
Under the Spearin doctrine, the owner impliedly warrants that the plans and specifications it provides are adequate, regardless of whether the owner is self-performing or has hired a GC. On a self-performed or multi-prime job, courts have also found owners have an implied duty to actively coordinate the separate prime contractors — adding a coordination obligation on top of the underlying design warranty.
Does hiring a GC protect an owner from design error costs?
No — it changes the process, not the exposure. A GC-led structure means the GC typically processes the RFI and change order first, but a legitimate design-error cost still gets passed through to the owner, since the GC has no obligation to absorb costs caused by defects in an owner-provided design.
Is a cross-discipline document review more valuable on a self-performed project?
It's valuable on both, but the case is often sharper on a self-performed job, because the owner has no GC layer to catch or buffer a coordination conflict before it becomes a field issue — a review that surfaces the conflict pre-bid removes the owner's direct exposure to the coordination-duty and delay-claim risk that comes with holding the trade contracts.