More than half of engineering firms in the US are turning down profitable work right now— not because the work isn't there, but because they can't deliver it. That's not a demand problem. That's a delivery ceiling.
According to the ACEC Research Institute's Q4 2024 Engineering Business Sentiment Study1, 51% of engineering firms are turning down work due to staffing shortages, with 26% declining profitable projects because they lack the necessary personnel. These are firms with winning track records. They're not failing at business development— they're choking on their own success.
"Firms are quietly turning away work, forgoing new bids, and shelving expansion plans because they simply can't staff up in time."2 — Newforma, 2025
This piece walks through the structural mechanism— using Theory of Constraints logic applied to AEC QA/QC— and explains why the fix isn't more headcount. It ends with a specific deployment model for where AI belongs in the workflow.
The frustrating part isn't the backlog. It's understanding why it jams.
The Structural Diagnosis: Senior Time Is the Throughput Ceiling
The senior engineer isn't the bottleneck because they're slow. They're the bottleneck because every deliverable in the firm must pass through their review before it goes out the door— and that review capacity doesn't scale the way production capacity does.
This is the Theory of Constraints (ToC), first articulated by Eliyahu Goldratt, applied to AEC project delivery. In any system, one constraint limits the output of the whole. The three moves are:
- Identify the constraint: QA/QC sign-off by a licensed senior engineer
- Subordinate everything else to it: junior workflows exist to serve the constraint, not compete with it
- Elevate the constraint: increase senior review capacity before adding anywhere else
The trap most firms fall into is adding junior staff to boost production without addressing the review bottleneck. More junior output means more deliverables queued for the same senior reviewers. Throughput doesn't improve. The queue grows.
Monograph's 2025 benchmarking data3 shows principals average 72% utilization— the lowest of any firm role. Job captains hit 91%. Project managers run 88%. That 28% non-billable gap is where QA/QC review lives, alongside business development, mentoring, and planning. Senior time is the scarcest resource in the firm, and the QA/QC obligation sits squarely inside it.
"The approval that gave a junior team confidence at 10 people becomes a queue at 50, and the architectural opinion that produced good decisions becomes a gatekeeping mechanism when the system is too large."4 — Leadership Garden
The constraint didn't change. The pressure on it did.
The Missing Middle Is Gone— and Won't Be Back Soon
The mid-career engineers who would normally sit between junior staff and principals— absorbing first-level escalations, running coordinated reviews, carrying institutional knowledge— aren't there. 76% of AEC firms identify this gap as a moderate or extreme concern5.
The missing middle isn't a retention failure. It's a demographic pipeline gap rooted in the Great Recession of 2008, which eliminated an entire cohort of AEC entrants who never fully returned to the field. The engineers who would be leading coordinated reviews today were the ones who left the industry fifteen years ago and never came back.
What the missing middle normally provides:
- Absorbs junior escalations before they reach the principal
- Coordinates cross-discipline review cycles
- Carries project history and institutional knowledge between engagements
Skills-shortage vacancies in civil engineering jumped 84% between 2022 and 20246. That's a signal of structural scarcity, not a temporary hiring market. And 40% of AEC firms worry about institutional knowledge loss when senior staff retire2— when the senior leaves, the knowledge walks out with them.
You can't hire your way out of this alone in the near term. The constraint needs a different solution. Firms experimenting with AI workflow automation are finding ways to extend what their current team can absorb without waiting for the pipeline to refill.
What the Bottleneck Actually Costs
When QA/QC review can't keep pace with project volume, the result shows up in rework— and the numbers are concrete. Firms with consistent QA/QC processes keep rework below 5% of project budget in 56% of cases. Firms without that standard manage the same only 37% of the time7.
| Firms With Consistent QA/QC | Firms Without Standards | |
|---|---|---|
| Rework below 5% of budget | 56% of cases | 37% of cases |
| Typical rework exposure | 4–10% of project cost | 4–10% of project cost (unmanaged) |
| US construction rework loss | — | $31 billion annually7 |
Run those numbers against a $2M project: 4–10% of budget is $80,000–$200,000 in rework exposure per project— and for a firm running fifteen active projects, that's a portfolio-level risk sitting in a backlog everyone calls a sign of success. Firms without QA/QC standards are also 50% more likely to face warranty exposure and 23% more likely to face subcontractor disputes7.
The bottleneck isn't just a throughput problem. It's a margin problem.
AI's Job Is to Return Senior Time, Not Replace Senior Judgment
AI doesn't eliminate the senior engineer bottleneck. It moves the constraint upstream— from clerical execution to problem framing and design judgment. In practical terms: that's the goal, if you deploy it correctly.
Here's a distinction I make with every AEC firm I work with8: the pre-QC pass and the formal QC review are two separate functions that most firms collapse into one.
- Pre-QC pass (designer-led): Catches clerical errors— missing dimensions, incorrect callouts, mismatched code references. Junior staff and AI handle this layer. It doesn't require a stamp.
- Formal QC review (senior-led): Judgment on design intent, constructability, professional liability. This is non-delegable. No tool substitutes for it.
A quick test: if the answer requires a licensed engineer's professional judgment, it's formal QC. If it requires looking something up or checking a match, it's pre-QC.
When you collapse these two roles, senior engineers spend their scarcest hours checking whether a dimension string matches the key plan. That's chasing pennies when they should be chasing dollars— defending design intent and catching the constructability problem a junior engineer won't see.
"AI does not eliminate constraints. It moves them upstream."9 — Dr. Lisa Lang
That's precisely the goal. When AI handles the deterministic pre-QC layer, the constraint shifts from clerical execution to the quality of problem framing and design judgment. Which is exactly where senior time belongs.
One engineering firm, Dunaway, recovered nearly 10,000 employee hours by deploying AI to automate regulatory compliance research— work that had previously required engineers to manually search documentation across 50+ cities10. That's one firm's reported outcome, not an industry benchmark. But the pattern it illustrates is real: remove the deterministic, searchable, repeatable work from the senior queue. Understanding what AI culture actually requires at the team level before the deployment pays off matters as much as the tool itself.
The caution worth holding: 93% of engineering leaders across industries expect AI to drive productivity gains, but only 3% are currently achieving "very high" impact11. The gap exists because most firms deploy AI without first diagnosing where the constraint lives. Diagnosing your constraint before you deploy anything is the right instinct— the AI decision framework for founders was built for exactly this.
The ambition is right. The solution isn't fewer projects. It's returning senior engineers to the work that earns the fee— while AI and junior staff handle the work that doesn't require a stamp.
The Ambition Is Right
The firms turning down work aren't failing. They're constrained— and constraint, unlike ambition, is fixable once you can see it. The diagnostic is the first step: where is your throughput ceiling?
The path is structural: identify the throughput ceiling (senior judgment at formal QC), subordinate junior workflows to serve it, and deploy AI where it belongs— the pre-QC pass that currently drains senior time on clerical work. A useful starting question to bring to your leadership team: which tasks in our QC workflow require a stamp, and which just require a lookup? That's where the diagnostic begins.
No matter the question, people are the answer. AI is the tool that protects your capacity to do the work only you can do. If working through this diagnostic feels like a full-time job on its own, that's exactly what AI implementation services are built for— the kind of diagnostic work that finds the constraint before you spend money on the wrong fix.
FAQ
What is the senior engineer bottleneck in AEC?
The senior engineer bottleneck occurs when all QA/QC review must pass through a small number of licensed senior staff, creating a throughput ceiling. As junior capacity grows without corresponding growth in senior review time, the queue at the constraint grows— not the throughput. The bottleneck isn't about pace or individual effort. It's about system structure.
Why are engineering firms turning down profitable work?
51% of engineering firms cite staffing shortages as the reason they decline projects, with 26% specifically declining profitable work due to insufficient personnel1. In most cases, the binding constraint isn't junior headcount— it's senior review capacity at QA/QC, where non-delegable sign-off creates the actual throughput ceiling.
Does hiring more junior engineers fix the senior bottleneck?
No. In a constrained system, adding capacity upstream of the bottleneck increases queue pressure without increasing throughput. More junior production means more deliverables waiting for senior sign-off— the bottleneck gets worse, not better. Theory of Constraints logic predicts this outcome in any professional services system where review is non-delegable.
Can AI eliminate the senior engineer bottleneck in AEC?
AI moves the constraint upstream, not eliminates it. The correct deployment targets the pre-QC pass: deterministic checks— dimensions, callouts, code references— that currently consume senior hours on clerical work. Senior time is then preserved for formal QC review: the design-intent and constructability judgment that only a licensed engineer can provide9.
What does rework cost AEC firms without strong QA/QC?
Rework accounts for 4–10% of total project cost across project types. The US construction industry loses $31 billion annually to rework7. Firms with consistent QA/QC processes keep rework below 5% of project budget in 56% of cases, compared to 37% of firms without established standards.
References
- ACEC Research Institute, "Engineering Business Sentiment Study Q4 2024" (2024) — https://openquire.com/recruiting-and-retention-in-the-aec-industry-navigating-the-pain-points/
- Newforma, "Labor Shortages in AEC: The Hidden Impact and How Smart Firms Are Adapting" (2025) — https://www.newforma.com/labor-shortages-in-aec-the-hidden-impact-and-how-smart-firms-are-adapting/
- Monograph, "Architecture Business Benchmarks: Understanding and Improving Utilization Rates" (2025) — https://monograph.com/blog/unlocking-utilization-rates-benchmarks-for-architects-and-architecture-firms
- Leadership Garden, "The Constraint You Won't Find on a Jira Board" (2025) — https://leadership.garden/theory-of-constraints/
- Quire/Openquire, "Recruiting and Retention in the AEC Industry: Navigating the Pain Points" (2024–2025) — https://openquire.com/recruiting-and-retention-in-the-aec-industry-navigating-the-pain-points/
- IBISWorld, "Growth Gaps: Industries with Strong Demand but Workforce Constraints" (2024) — https://www.ibisworld.com/blog/growth-gaps/99/1126/
- PlanRadar, "Cost of Rework in Construction: Causes, Data & Prevention" (2025) — https://www.planradar.com/us/cost-of-rework-construction/
- Dan Cumberland Labs, "How to Run a Pre-QC Pass That Saves Architecture Time" (2025) — https://dancumberlandlabs.com/blog/architecture-time/
- Velocity Scheduling System (citing Dr. Lisa Lang), "AI Shifts The Bottleneck: A Theory Of Constraints AI Perspective" (2025) — https://www.velocityschedulingsystem.com/blog/theory-of-constraints-ai/
- Zweig Group, "How AEC Firms Can Use AI to Expand Capacity" (2025) — https://zweiggroup.com/blogs/the-zweig-letter/how-aec-firms-can-use-ai-to-expand-capacity
- SimScale, "The Engineering AI Ambition-Execution Gap: What Our New Global Survey Reveals" (2025) — https://www.simscale.com/blog/the-engineering-ai-ambition-execution-gap-what-our-new-global-survey-reveals/
