The Institutional Knowledge Crisis in AEC: The Retirement Wave Is Coming

What's Actually at Risk: Understanding Institutional Knowledge in AEC

Institutional knowledge is the accumulated expertise, decision-making patterns, and unwritten know-how that keeps an organization running. Roughly 90% of it is tacit— held in people's minds, not in any document or database⁴.

That number is worth sitting with. It means the vast majority of what makes your firm effective isn't written down anywhere.

In AEC, tacit knowledge shows up in everything your best people do:

• How to read a site before conditions become problems
• Which subcontractors will actually show up and deliver under pressure
• What the client really means versus what's written in the contract
• Why a particular detail was designed the way it was— the story behind the spec
• How to sequence complex work when the textbook approach won't survive first contact with reality

These aren't skills from a certification course. They're patterns built through thousands of decisions across hundreds of projects. And the person who holds them often can't fully explain how they make those calls— the knowledge has become intuitive, embedded in how they see the work.

Explicit knowledge— project specifications, safety procedures, contract templates— can be written down and shared. Tacit knowledge resists that by nature. A senior estimator who's been pricing work for 25 years doesn't consult a formula to know when a bid feels wrong. That judgment was built project by project.

And AEC's project-based structure makes this vulnerability worse. Knowledge scatters across teams and projects. Once a project closes and the team disperses, lessons learned go with the people— unless someone deliberately captured them first.

Research suggests organizations lose approximately 45% of their institutional knowledge when an employee departs⁴. For large businesses, inefficient knowledge sharing already costs an average of $47 million annually⁵. In an industry where 41% of the workforce is heading for the exit, those numbers become existential.

The Retirement Wave: By the Numbers

Across every AEC discipline— construction, architecture, and engineering— the numbers tell the same story: more experienced professionals are leaving than entering, and the gap is accelerating.

The construction pipeline numbers are stark. The industry needs roughly 390,000 new workers every year over the next decade, but fewer than 100,000 are currently enrolled in construction training programs⁸. That's a pipeline running at a quarter of capacity.

And even when new workers do enter the industry, the ramp-up is slow. Apprenticeships in the trades typically take four to five years⁹. Aggressive recruiting started today won't produce fully skilled workers until 2030 at the earliest— by which point a large portion of the experienced workforce will already be gone.

Architecture is facing its own version of the crisis. In 2024, the number of licensed architects in the U.S. dropped by 4%— a net loss of roughly 4,600 professionals, driven entirely by retirements outpacing the 3,600 newly licensed architects entering practice⁷. The median architect in the U.S. is 51 years old⁶. That's a current reality, not a future risk.

On the engineering and infrastructure side, a Brookings Institution analysis estimates that 1.7 million infrastructure workers leave their positions each year, in large part due to retirement¹⁰. These aren't entry-level roles being vacated. They're the positions that require years of accumulated field judgment.

The retirements are only part of it. An estimated $8 trillion in wealth will transfer within the construction industry by 2031 as baby boomer-owned firms change hands¹¹. Only 25% of firms plan to sell to employees— down from 38% in 2020¹¹. That means the majority of transitions will involve people from outside the organization who carry none of its institutional memory.

That's the numbers. What's less obvious is what happens when that experience actually walks out the door.

The Business Impact: What Knowledge Loss Actually Costs

When institutional knowledge leaves, firms pay for it in rework, delays, and safety incidents— costs that show up on the balance sheet long after the retirement party.

Rework is the most direct hit. It accounts for 5-15% of total construction project costs¹³— roughly $15 billion annually in the U.S. alone¹². When experienced crew leaders retire and newer workers lack the judgment to catch problems early, errors make it further down the line before anyone notices. The cost of fixing those errors grows at every stage.

Project delays compound the damage. According to the AGC's 2025 workforce survey, 45% of contractors report experiencing project delays due to labor shortages— their own or their subcontractors'³. Delays mean penalties, strained client relationships, and revenue sitting on the table.

Safety risks increase when experience leaves. Construction workers themselves cite inexperienced colleagues as the largest safety concern on jobsites— 83% rank it as their top worry⁹. When the people who know which situations require extra caution aren't there to flag them, incident rates go up.

Competitive position erodes. Your senior estimator who's been pricing projects for two decades knows which numbers to trust and which to question. Your project executive who built relationships with your top three clients over 15 years isn't replaceable through a job posting. Lose both in the same year and your firm doesn't just get less efficient— it loses the judgment that wins work.

And the effects compound. Knowledge loss doesn't degrade just one project— it degrades organizational capability over time. A firm that loses its most experienced people without capturing what they know doesn't just get worse at one thing. It gets worse at everything, gradually, in ways that don't show up in a single quarterly report.

Why Mentoring Alone Won't Solve This

Mentoring programs are essential for knowledge transfer. But they can't operate fast enough to capture what 41% of a workforce knows before those workers walk out the door.

This isn't a knock on mentoring— it works. Research shows mentorship programs can reduce employee turnover by up to 20%¹⁴. Organizations like the AIA have invested significantly in mentoring initiatives through programs like ALIGN, Nexus, and the College of Fellows¹⁵. The industry recognizes that experienced professionals learning alongside newer ones is how knowledge has always been passed down.

The problem is math, not method.

When 89% of firms report worker shortages¹¹ and 43% don't even have a formal workforce development budget¹¹, the conditions for effective mentoring barely exist. You need experienced people available to mentor— and they're the ones who are overloaded or heading for retirement. You need time for structured learning— and chronic understaffing eliminates that margin. You need enough new hires to pair with veterans— and the apprenticeship pipeline delivers a quarter of what's needed.

Traditional mentoring assumes you have enough experienced people and enough time to pair them up. When 89% of firms report worker shortages and apprenticeships take four to five years, neither assumption holds.

There's a deeper challenge, too. The most valuable institutional knowledge is often the hardest to articulate. A superintendent with three decades of experience doesn't consciously think through why she makes certain calls. The judgment has become instinct— pattern recognition built through thousands of field decisions. Ask her to explain it and she'll give you part of the story. But the full picture lives in a form that words don't easily capture.

This is the tacit knowledge paradox: the knowledge you most need to preserve is the knowledge that most resists being put into words.

Mentoring transfers some of it. Working alongside someone for years transfers more. But at the scale and speed AEC needs right now, mentoring alone can't keep pace. Building an AI culture in your organization means finding tools that extend what mentoring does— not replace it, but make it reach further.

How AI Changes the Equation for Knowledge Capture

Here's where it gets interesting. AI-powered tools can capture, organize, and make searchable the institutional knowledge that traditional mentoring and documentation miss— not by replacing experienced workers, but by extending the reach and durability of what they know.

The core principle matters here. AI doesn't replace the 30-year superintendent. It makes what that superintendent knows accessible to the next generation in ways that weren't possible five years ago. Think of AI like a new intern from an Ivy League school— brilliant processing power, zero institutional knowledge. The value comes entirely from what experienced workers feed it. No matter the question, people are the answer. AI just makes their expertise more durable.

Here's what that looks like in practice for AEC firms:

• Video documentation analysis: Record experienced workers walking jobsites, explaining decisions in real time, and narrating field conditions. AI tools can transcribe, index, and make those recordings searchable by topic— so a junior PM can find "how to handle clay soil conditions on a grade beam" without knowing who to ask.
• Project documentation mining: AI can analyze thousands of project files— RFIs, change orders, lessons-learned logs— to surface decision patterns that would take a human analyst months to compile.
• Digital twins: Platforms like Autodesk Tandem are being used to preserve spatial and procedural knowledge about buildings and infrastructure¹⁶. A digital twin captures not just what was built, but can serve as a foundation for documenting how decisions were made and why— when teams layer decision logs and field notes onto the model.
• Pattern recognition across projects: AI can identify recurring problem types across your project history and connect them to the solutions experienced workers applied— turning individual expertise into organizational capability.

A practical note: the value of AI for knowledge capture isn't in the tool itself. It's in the institutional knowledge you feed it. Better input produces better output. A knowledge base built from 30 years of project history is far more useful than a generic AI tool pointed at your file server.

But there are honest limitations. AI won't capture the trust a project executive has built with a client over 15 years. It won't replicate the rapport your lead carpenter has with the framing crew. Relationship-based knowledge still requires human-to-human transfer. The technology is the easy part. The human change— developing an AI governance strategy your team will actually follow— is where the real work happens.

A Knowledge Preservation Roadmap for AEC Leaders

A practical institutional knowledge strategy starts with identifying what you're about to lose, then building three layers of capture: structured mentoring, systematic documentation, and technology-assisted preservation.

Here's a framework AEC firm leaders can start using now:

1. Conduct a Knowledge Risk Audit

Identify every person within five to ten years of retirement. Map the unique knowledge each one holds. Ask: if this person left tomorrow, what would we struggle to replicate?

Don't limit this to senior leadership. Your most critical knowledge holders might be a field superintendent, a senior estimator, or a project coordinator who's been managing the same client portfolio for 20 years. Rank by criticality and retirement timeline.

2. Build Structured Mentoring (With Teeth)

Pair knowledge holders with junior team members— but with structure that goes beyond "shadow the veteran." Define specific knowledge domains for each pairing. Set documented learning objectives. Schedule regular check-ins. And give mentors protected time to do this work, not as an afterthought on top of a full project load.

3. Systematize Documentation Beyond Project Files

Move past the standard project closeout report. Create decision logs that capture why choices were made, not just what was decided. Build lessons-learned databases organized by problem type. Record "the story behind the detail"— why a particular approach was used, what went wrong the first time, what the client's real concern was underneath the contract language.

4. Add a Technology Layer

Introduce AI tools for video capture, documentation analysis, and searchable knowledge bases. Start small— one retiring expert, one knowledge domain, one capture method. Prove it works before you scale. AI implementation services can help you match the right tools to your firm's specific workflows without building from scratch.

5. Build Ongoing Knowledge Capture Into Your Workflows

And knowledge preservation isn't a one-time project. If you only capture knowledge in a panic before someone retires, you'll miss most of it. Weave capture into existing project rhythms: debrief meetings that get recorded and transcribed, field documentation that gets indexed automatically, and clear metrics for measuring AI success so you know what's actually working.

The key to all of this? Start now, start small, and start with your highest-risk knowledge holders. Pick one person. Pick one domain. Run one capture method. You don't need to solve the whole problem at once.

The firms that act while their most experienced people are still in the building will carry a durable advantage over those that wait.

The Window Is Open, But It Won't Stay Open

The AEC industry has a narrowing window to capture decades of institutional knowledge before it retires out the door. Firms that invest in structured capture now will carry a competitive advantage for years to come.

The data is unambiguous. Forty-one percent of the construction workforce will retire by 2031¹, ninety percent of what they know was never documented, and the replacement pipeline runs at a quarter of capacity. The business costs— rework, delays, safety incidents— are measurable and growing.

The answer isn't any single approach. It's the combination: mentoring that's structured and protected, documentation that goes beyond project files, and technology that makes knowledge accessible at scale. Both the human side and the technology side. All of it matters.

But here's the good news. Most of the people who hold this knowledge are still working. They're still available. Many are willing to share what they know, if someone creates the structure for it.

The question for AEC leaders isn't whether institutional knowledge loss is a problem. It's whether you'll capture what your best people know while they're still around to share it.

If building a knowledge preservation strategy feels complex— mapping the right tools to your workflows, integrating AI into your capture process, figuring out where to start— that's exactly the kind of challenge an AI strategy assessment is designed to solve. A focused look at what you're about to lose, and a practical plan for keeping it.

FAQ: Institutional Knowledge in AEC

What is institutional knowledge?

Institutional knowledge is the accumulated expertise, processes, decision-making patterns, and relationships held by an organization's people. Roughly 90% of it is tacit— meaning it exists in employees' minds, not in documents or databases⁴. In AEC, this includes field judgment, client management instincts, and problem-solving patterns built over decades of project work.

Why is institutional knowledge important in the AEC industry?

AEC firms depend on project-based expertise that takes decades to develop. Knowing how to sequence complex work, manage client expectations, and avoid costly mistakes isn't something you learn from a manual. When experienced professionals retire without transferring this knowledge, firms face rework costs of 5-15% per project¹³ and delays affecting 45% of firms³.

How much institutional knowledge is lost when an employee leaves?

Research suggests organizations lose approximately 45% of their institutional knowledge when an employee departs⁴. The most critical losses occur in tacit knowledge areas— the judgment, intuition, and relationship skills that are hardest to document. In AEC, this includes field decision-making, subcontractor evaluation, and project-specific problem-solving approaches.

How can AI help capture institutional knowledge?

AI tools can analyze video documentation of experienced workers, mine project files for decision patterns, build searchable knowledge bases from unstructured data, and create digital twins that preserve spatial and procedural knowledge¹⁶. These tools augment human mentoring and documentation— they don't replace the need for experienced workers to share what they know.

What is the difference between tacit and explicit knowledge?

Explicit knowledge can be written down and standardized— project specifications, safety procedures, contract templates. Tacit knowledge is the intuitive expertise that experienced professionals often can't fully articulate: why they chose a particular approach, how they read a client's real concerns, or what warning signs they watch for on a jobsite. The challenge for AEC firms is that tacit knowledge makes up roughly 90% of what an organization knows⁴.

References

1. NCCER, "Construction Workforce Age Progression" (2023) — https://www.nccer.org/media/2023/03/construction-workforce-age-progression.pdf
2. U.S. Bureau of Labor Statistics, "Employment by Detailed Industry and Age" (2024) — https://www.bls.gov/cps/cpsaat18b.htm
3. AGC, "2025 Workforce Survey Analysis" (2025) — https://www.agc.org/sites/default/files/users/user21902/2025%20Workforce%20Survey%20Analysis%20(3).pdf.pdf)
4. Mem.ai / Knowledge Management Research, Nonaka & Takeuchi framework (2024) — https://get.mem.ai/blog/how-t0-capture-institutional-knowledge
5. Panopto, "Workplace Knowledge & Productivity Report" (2018) — https://www.prnewswire.com/news-releases/inefficient-knowledge-sharing-costs-large-businesses-47-million-per-year-300681971.html
6. NCARB & AIA, "What Happens When Baby Boomer Architects Retire" (2024) — https://www.ncarb.org/blog/what-happens-when-the-baby-boomer-architects-retire
7. NCARB, "By the Numbers 2025 Edition" (2025) — https://www.ncarb.org/sites/default/files/NBTN-2025.pdf
8. Fieldwire/Hilti, "Aging Construction Workforce" (2023) — https://www.fieldwire.com/blog/aging-construction-workforce/
9. McKinsey, "Delivering on Construction Productivity" (2024) — https://www.mckinsey.com/capabilities/operations/our-insights/delivering-on-construction-productivity-is-no-longer-optional
10. Brookings Institution, cited by ASCE (2024) — https://www.asce.org/career-growth/salary-and-workforce-research
11. FMI Corp, "Construction Industry Research" (2024) — https://fmicorp.com/insights/
12. Construction Industry Institute; MDPI Journal (2023) — https://www.mdpi.com/2071-1050/17/21/9880
13. CMI Global, "How Construction Software Can Help in Reducing Rework" (2023) — https://cmicglobal.com/resources/article/How-Construction-Software-Can-Help-in-Reducing-Rework
14. The Diversity Movement, "Bridge Generation Gap at Work" (2024) — https://thediversitymovement.com/bridge-generation-gap-at-work-mentoring-knowledge-transfer-sharing/
15. American Institute of Architects, "Career Growth: Mentoring" (2024) — https://www.aia.org/career-growth/mentoring
16. Autodesk Tandem, "Digital Twins for Building Preservation and Restoration" (2024) — https://intandem.autodesk.com/resource/realizing-the-potential-of-digital-twins-for-building-preservation-and-restoration-efforts/