"Catch design errors before construction" is advice everyone in commercial real estate has heard and almost nobody has turned into a checklist. It sounds obvious right up until someone asks what, specifically, to look for in a 400-sheet set before it goes to bid — and the honest answer, on most projects, is nothing specific at all. The architect's QA/QC ran. The GC did a buyout-level pass. Nobody read the full set, every discipline against every other discipline, looking for the contradictions that turn into RFIs once the set is priced and built.
Why "before construction" is the only window that matters
The timing isn't a nice-to-have. A conflict found before the set is priced gets corrected as an addendum — a clarification every bidder receives on equal footing, with no contract price attached to argue over. The identical conflict, found after the contract is signed, becomes additional scope relative to what was bid, and additional scope gets negotiated as a change order. That mechanism is covered in more detail in why change order risk mitigation starts before the bid, not after — the short version is that the same fix costs nothing on one side of contract execution and a negotiated change order on the other.
Academic research on construction cost-of-quality backs up why this is worth the effort. Studies on design deviations — including the frequently cited Lopez & Love research on rework costs — have found design-related errors and omissions accounting for the large majority of increased project costs, with construction-phase deviations a distant second. Design mistakes don't stay design mistakes; they become field costs, and the earlier they're caught relative to bid, the cheaper the correction.
The checklist
This is what "catching design errors before construction" actually breaks down into — not a philosophy, a list of specific things someone needs to check, sheet by sheet, before the set goes out.
1. Cross-discipline overlay, not single-discipline QC
Confirm someone has read the structural set against the mechanical set against the architectural set — not that each discipline's drawings are internally consistent with themselves. A duct routed through a beam, a ceiling plan that doesn't match the mechanical layout above it, a structural detail that doesn't leave clearance for a specified pipe run: none of these fail any single discipline's own QA/QC, because QA/QC checks a set against its own firm's standards, not against a different discipline's sheets. That distinction — and why a passed QA/QC review doesn't mean a coordinated set — is covered in why "it passed QA/QC" doesn't mean the set is coordinated.
2. Spec sections read against the drawn details
Specs and drawings are written by different people, often at different times, and frequently disagree with each other without anyone noticing until a sub is standing in front of a detail that contradicts the spec section governing it. Every spec section that touches a coordination-sensitive assembly — fire-rated wall types, clearance requirements, finish schedules — needs to be checked against the actual drawn detail it's supposed to describe.
3. Every detail callout resolves to a real detail
On a set this size, missing or mismatched detail references are common and easy to miss individually. A callout pointing to a detail number that doesn't exist, or that exists but describes something else, isn't cosmetic — it's a gap the field has to guess at, usually under time pressure, usually not in the owner's favor.
4. Written notes checked against what's actually drawn
A general note buried on a cover sheet assuming coordination that was never verified on the plans is one of the most common sources of field conflict, because it reads as resolved on paper while the drawings underneath it were never actually reconciled to match.
A written note and a drawn detail can both be individually correct and still contradict each other — the note says one thing, the detail shows another, and neither discipline's own QA/QC is scoped to catch the mismatch, because each is only checking its own document against itself.
5. Code compliance checked across disciplines, not just within one
Each discipline typically confirms its own scope meets the code sections it's responsible for. What's less consistently checked is whether the combination is still compliant — a fire-rated assembly that's correct on the architectural sheet but penetrated by an unsleeved duct on the mechanical sheet, for instance. That gap lives between disciplines, which is exactly where single-discipline code review doesn't reach.
6. Constructability, not just correctness
A detail can be technically buildable in isolation and still create a sequencing or access problem once it's read against the trades working around it — a duct that's correctly sized and correctly routed but can't physically be installed after the structural steel it depends on is already in place. Catching that requires reading the set with field sequencing in mind, not just checking each sheet for internal accuracy.
7. Findings documented with severity and dollar exposure, not just flagged
A list of "coordination notes" competes poorly for attention against a schedule that's already tight. A findings list that ranks each item by sheet, severity, dollar exposure, and schedule impact gets acted on differently — it's the difference between a note that gets deferred and a number that gets a decision made against it before bid.
Who should be running this checklist, and when
Every item above needs to happen before the set goes to bid, not after. Once a contract is signed, the same conflicts still get found eventually — they just stop being free to fix. On timing specifically, when to hire a drawing review consultant covers the tradeoffs of running this check at different phases, but the short version matches item one above: earlier is cheaper, and pre-bid is the last point where a fix is a redline instead of a negotiation.
None of these seven items require replacing the architect's QA/QC or the GC's buyout review — both are doing real, necessary work, just work that's scoped to a different question than "does this set agree with itself across every discipline." What counts as a "set" ready for bid covers the broader readiness question this checklist feeds into; this list is specifically the cross-discipline layer that tends to fall through the gap between the two reviews that already happen.
Key takeaways
- Catching design errors before construction means checking specific things — cross-discipline overlays, spec-vs-drawing agreement, detail callouts, written notes against drawings, cross-discipline code compliance, and constructability — not a general sense that "someone reviewed it."
- None of the seven checklist items are covered by a single discipline's own QA/QC, because QA/QC checks a set against its own firm's standards, not against a different discipline's sheets.
- The window where these conflicts are free to fix closes at contract execution — after that, the identical conflict becomes a negotiated change order instead of an addendum.
- Findings need severity and dollar exposure attached to compete for attention against a tight pre-bid schedule.
- This checklist is the cross-discipline layer specifically — it doesn't replace the architect's QC or the GC's buyout review, it fills the gap between them.
A checklist doesn't run itself. Someone still has to sit down with the full set, every discipline against every other discipline, and work through all seven items before the set is priced — which is a different job than any review already happening on most projects, and the only one scoped to catch what falls in the gap between them.
Frequently Asked Questions
What's the single most common design error that gets missed before construction?
Cross-discipline conflicts — a duct routed through a structural beam, a written note that assumes coordination the drawings never actually show — because they can clear every individual discipline's own QA/QC while still contradicting a different discipline's sheet. The conflict only exists in the gap between documents, which no single firm's internal review is scoped to check.
Isn't this what the architect's QA/QC process already does?
No. QA/QC checks a discipline's own drawings against that firm's internal standards, code scope, and detail library — it verifies internal consistency, not agreement between disciplines. A set can pass QA/QC completely in every discipline and still carry cross-discipline conflicts that only surface once someone reads the full set against itself.
When is the best time to run this checklist?
Before the set goes to bid. A conflict caught pre-bid is typically corrected as a no-cost addendum, applied equally to every bidder pricing the set. The identical conflict caught after the contract is signed is treated as additional scope and gets negotiated as a change order, which is a more expensive and more adversarial way to resolve the exact same design mistake.
Does BIM clash detection cover this checklist?
Only partially. Clash detection is strong at catching modeled geometry that physically overlaps, but it doesn't catch a written note that contradicts a drawn detail, a spec section that disagrees with the drawings, or a missing detail callout — conflicts that live in the documents themselves rather than in the 3D model.