The cheapest version of a construction defect is the one you find on a PDF. A conflict caught during plan review costs a markup and a coordination meeting. The same conflict caught in the field costs demolition, rework, a change order, and schedule — industry experience puts that ratio at roughly ten to one.
Pre-construction QA is the discipline of finding those conflicts before ground breaks: reading the full document set, verifying that the disciplines agree, and confirming that what the drawings show can actually be built. It is not glamorous work. It is also not optional if you are the one who absorbs the cost when things go wrong.
This checklist runs through what pre-construction QA actually covers, discipline by discipline, with the coordination seams and completeness checks that matter most. It is organized for a practitioner — a GC chief estimator, an owner’s rep, a project manager doing a final review before bid — not for a consultant’s deliverable.
Who owns pre-construction QA — and when it runs
Pre-construction QA does not belong to one role. Different parties run it at different times for different reasons:
- The design team runs internal QA before issuing for permit or bid — verifying that their own disciplines are internally consistent and that the spec matches the drawings.
- The general contractor runs it at bid and again at pre-construction — looking for the constructability issues, coordination gaps, and scope conflicts that affect schedule, subcontractor buyout, and contingency sizing.
- The owner or owner’s rep runs it (or should) before committing to a GC price — because the conflicts that get built into the structure get built into the change orders, not into the bid.
The right answer for when to run it: as early as possible, and again when the documents are updated. A conflict found during design development costs a revision. A conflict found after permit is issued costs a reissue. A conflict found after mobilization costs a change order.
Checklist by discipline
Architectural
Structural
Mechanical / HVAC
Plumbing
Electrical
Civil / Site
Cross-discipline coordination checks
This is where the money is. Individual disciplines can each be internally correct while still contradicting each other — and those inter-discipline conflicts are the ones that cost $28,000–$45,000 and two to three weeks of schedule when they surface in the field. (That figure comes from a real project in Flikt’s dataset, documented on the evidence page.)
The seams to check:
- Structural vs. mechanical: shear walls, beams, and column caps in the paths of duct mains, pipe risers, and conduit runs. Verify that duct routes have actual clearance to run where the mechanical drawings show them running, accounting for structural member depths.
- Architectural vs. structural: slab-to-slab heights on architectural plans match the floor-to-floor dimensions in the structural sections. Discrepancies show up as conflicts between ceiling height, mechanical depth, and structural member depth.
- Mechanical vs. electrical: ceiling plenum space is shared. HVAC ductwork, lighting fixtures, sprinkler mains, and electrical conduit are all competing for the same zone. In tight plenum conditions this requires explicit coordination — not just independent plans that fit on their own.
- Plumbing vs. structural: horizontal waste runs below slab or between framing require structural penetrations. Where they’re not detailed, they become field RFIs.
- Spec vs. drawings: the project manual and the drawings are separate documents prepared by separate parties and sometimes updated separately. Product types, fire ratings, finish levels, and performance specs should be cross-checked — disagreements become scope gaps at buyout.
On one real project, more than 40 change orders traced back to conflicts between 2D plan sheets alone — not latent site conditions, not owner changes. Just sheets that didn’t agree. That’s the baseline case for treating cross-discipline coordination as its own review category, not an afterthought. The MEP coordination failures in multifamily construction post documents what this looks like in practice.
Completeness checks
A drawing set can be internally consistent and still be missing things. Completeness checks are a separate pass:
- Sheet index vs. actual sheets: every sheet listed in the index is present; no sheet is present but unlisted. Missing sheets at bid create scope gaps and contingency guesswork.
- Referenced details: every note or callout that says “see detail X on sheet Y” should actually have a detail there. Orphan references are either omissions or revision artifacts that didn’t get cleaned up.
For the broader document-review workflow this feeds into, see the construction document review checklist.
Constructability and sequencing checks
Not every conflict is a documentation error. Some are legitimate design choices that create construction problems:
- Can the structure be built in the sequence the drawings imply? Are there access constraints for crane picks, formwork, or precast erection?
- Do underground utilities need to be installed before foundation work begins — and are their locations confirmed?
- Are there long-lead materials (structural steel, custom glazing, equipment) that need to be ordered before design is fully resolved? If so, which details are still open?
- Are any details shown in a way that’s technically correct but practically difficult to build or inspect? Standard connections should be preferred where the drawings leave ambiguity.
- Is there a concrete or masonry scope that requires shop drawings before construction? Is that lead time in the schedule?
For a more detailed treatment, see constructability review: what it is and how to run one.
Where automated review fits in this process
Manual pre-construction QA is thorough when time and personnel allow. In practice, a drawing set of several hundred sheets cannot be fully cross-checked sheet against sheet by a reviewer working against a bid deadline. The cross-discipline seams and spec conflicts described above are exactly the ones that fall through.
AI construction plan review works on the 2D PDF set you already have — no BIM, no model, no additional software licensing. It reads the full set, cross-references disciplines and specs, and surfaces the conflict categories this checklist describes: spec mismatches, cross-sheet dimension conflicts, coordination gaps, missing scope. It runs at design review, before bid, before mobilization — not after a coordination-ready model has been built.
Flikt’s benchmark for drawing quality is conflicts per 100 sheets: how many real conflicts a set contains, normalized by size. It’s a consistent, comparable signal across projects and design teams. You can see what it looks like in practice at conflicts per 100 sheets.
The checklist above reflects what a thorough manual review covers. AI review is not a replacement for the judgment it takes to prioritize and triage what it finds — but it is a coverage layer that doesn’t get tired, doesn’t skip the spec, and doesn’t run out of time before bid day.
If you want to connect the findings of a pre-construction QA to the downstream risk it addresses, start with how to reduce RFIs before construction starts. The RFI is what an unresolved conflict becomes once the project is underway; the checklist is what prevents it.
Flikt reviews 2D PDF drawing sets for cross-discipline conflicts before construction — no BIM required. If you want to see what conflicts actually look like on real projects, visit the evidence page. For owners and owner’s reps managing design quality across multiple projects, see how Flikt works for owner’s reps.
Ready to catch conflicts early?
Upload your plan set and get AI-powered conflict detection — no BIM required.