Why AEC Firms Keep Losing the Knowledge They've Already Earned

Every proposal coordinator in an AEC firm knows this feeling: the deadline is tomorrow, a perfect case study exists somewhere in the system, and no one can find it. Project data lives in one folder. Team bios are in another. The subject matter expert who actually remembers the work is already deep on the next pursuit.

This is not a technology failure. It is a structural one — and it is playing out quietly and expensively at firms of every size.

This article covers why knowledge management is uniquely difficult for AEC firms, what the highest-impact strategies look like in practice, and how AI is changing what is actually possible. If you are a marketing director tired of watching your team burn hours on information retrieval, or an implementation lead trying to make the case for a better system, this is written for you.

The Real Cost of Not Finding What You Already Know

McKinsey research estimates that employees spend an average of 1.8 hours every day searching for information — the equivalent of one full team member generating no output at all. For AEC firms, where every hour is either billable or overhead, that drag is not a minor inefficiency. It is a direct threat to margins and win rates.

IDC research puts the productivity loss from document-related challenges at up to 21.3%. For a 20-person marketing and pursuit team, that translates to more than four people contributing nothing — every single day.

The indirect costs are harder to quantify but just as real: proposals that never surface the strongest past project references, onboarding that takes weeks instead of days, and senior staff spending time explaining context that a well-organized system would have surfaced automatically.

The construction industry faces approximately $65 billion in rework costs annually, stemming in part from inadequate knowledge sharing and the reinvention of solutions already documented somewhere in a firm's archive. For architecture and engineering firms, every proposal is a unique deliverable with a hard deadline — there is no opportunity to recover from poor knowledge retrieval after the submission goes out.

Why AEC Knowledge Is Harder to Manage Than Most Industries Assume

AEC firms do not accumulate knowledge the way product companies do. There is no product roadmap. There is no repeatable sales process. Knowledge accumulates in RFIs, proposals, BIM models, and the heads of the technical staff who delivered the work.

When a firm pursues a new wastewater treatment contract, the team needs to locate past wastewater projects, extract relevant content, and align it with new RFP requirements — all under tight deadline pressure, often with the engineer who worked on the prior project unavailable. That is a content synthesis problem, not just a search problem.

The challenge intensifies with staff turnover. When a senior engineer or BD director leaves, they take tacit knowledge that no folder structure or file naming convention ever fully captured. Standard document management tools organize files, but they were not built to surface firm knowledge for pursuit teams who need content synthesis at speed.

As one implementation lead described it: "it is great to have a proposal library of PDFs that people can go through — they do not, and it is just too overwhelming."

A proposal library is not a knowledge system. It is a storage system. The difference matters enormously when your team is working against a two-week RFP window.

A Framework for Diagnosing Where Your Knowledge Workflow Breaks Down

The 5 C's — Capture, Curate, Connect, Communicate, and Convert — give teams a structured way to identify exactly where their knowledge management process fails.

Capture is getting knowledge out of individual experts and into searchable systems: tagging project data, uploading proposal content, documenting lessons learned. Most AEC firms do this reasonably well. They have content — it is just disorganized.

Curate is keeping that knowledge accurate and current. This is where most firms fall short. Outdated resumes still circulate in proposals. Project narratives reference work completed five years ago without flagging that the team has since moved on.

Connect links related knowledge across the firm. A bridge project in the Northeast should surface alongside similar work in the Southwest when a pursuit team needs it. Most systems do not make that connection automatically.

Communicate makes knowledge accessible through the right channels — an intranet, a knowledge base, or an AI search platform that team members actually use.

Convert is the stage that matters most for pursuit teams: turning stored knowledge into proposal text, team bios, and case studies that win work. This remains the most manual, time-intensive bottleneck in the process.

Christopher Parsons, founder of Knowledge Architecture and producer of KA Connect — the leading conference for AEC knowledge management practice — defines the discipline as integrating people, processes, and technology to maximize a firm's expertise. His framing makes clear that none of the 5 C's works in isolation. Technology without governance fails. Governance without AI-powered retrieval creates bottlenecks that teams eventually abandon.

The Five Dimensions Firms Must Align

Technology is only one part of what makes knowledge management work. The 5 P's — People, Process, Platform, Policy, and Performance — describe the full picture.

People are the origin and destination of all knowledge. Without their active participation, no system succeeds.

Process defines how knowledge gets captured and accessed — from post-project debriefs to proposal kickoff meetings. Firms that embed knowledge capture into active project phases, rather than asking exhausted teams to document after closeout, build richer and more current knowledge bases over time.

Platform is the technology layer — and for most AEC firms, it is currently a patchwork: project management tools, document management systems, digital asset libraries, CRMs, and shared drives that do not talk to each other. A proposal coordinator searching for a past project reference might check four separate systems before writing a single word. Joist AI's FAQ addresses why this fragmentation is so costly and what a unified platform layer looks like in practice.

Policy is governance — and it is the most underinvested dimension. Gartner estimates that poor data quality costs organizations an average of $12.9 million annually. For AEC firms, the risk is more specific: proposals built on outdated content risk misrepresenting firm qualifications, a significant problem on regulated public-sector pursuits where accuracy is scrutinized at every stage. Explicit governance policies — defining who updates resumes, when project narratives get refreshed, how outdated content gets flagged — are what keep the system trustworthy over time.

Performance closes the loop. Win rates, proposal turnaround time, and onboarding speed are the clearest signals of whether a knowledge management investment is working. Without baseline metrics established before implementation, improvements remain anecdotal.

Firms that treat knowledge management as a platform decision alone — without addressing people, process, policy, and performance — consistently underperform on adoption and ROI.

What AI Actually Changes

More than half of architecture and engineering firms now use AI in business development, proposal writing, and project analytics, according to industry research published in 2025. The 46th Annual Deltek Clarity A&E Industry Study found that 62% of firms expect AI to improve their operational efficiency.

The shift is significant because AI does not just make search faster. It changes what is searchable.

A traditional document library requires users to know what they are looking for and where it might live. An AI-powered platform allows a proposal coordinator to ask: "What are our strongest bridge rehabilitation projects under $10 million?" — and receive a synthesized response drawn from the full proposal archive, not a list of files to manually open and evaluate.

One implementation lead described early results clearly: onboarding staff no longer had to go find things and learn things — they could search, ask questions, and contribute from their first week on the job. That compression of the knowledge transfer timeline is one of the most immediate and tangible benefits firms report after deploying AI-powered search.

The efficiency gains are material. Firms integrating AI-powered search into their content workflows report productivity improvements of 25 to 40 percent on proposal-related tasks. In one documented case, an engineer at Benesch — a national engineering firm with more than 1,000 employees — completed a technical approach section in two hours that would typically require six to eight. Read the full Benesch case study to see how a 34-person marketing team scaled that result across nearly 400 technical staff.

What Real Implementation Looks Like

Firms that centralize their knowledge and apply AI-powered search report faster onboarding, higher team confidence in their content systems, and stronger first-draft quality on proposals.

At Robins & Morton, a senior marketing leader described the before state clearly: "I used to get constant messages asking if I remembered an RFP with certain language. Now, people can search for themselves. It's a game changer." The firm also used the platform to turn around a major aviation proposal in days rather than weeks — a pursuit window that would have been unworkable with their previous system. Read the Robins & Morton case study.

At Woodard & Curran, the team estimated time savings of 30 to 40 percent per proposal after centralizing their content — time reinvested directly into pursuit quality rather than overhead. Read the Woodard & Curran case study.

Those outcomes compound over time as the knowledge base grows and the platform deepens its understanding of a firm's portfolio.

Building a Knowledge System That Actually Gets Used

The most durable knowledge management programs in AEC firms share three characteristics.

First, dedicated ownership. A knowledge manager or content coordinator accountable for quality — not a committee, not a shared responsibility — is the single strongest predictor of long-term system health.

Second, defined content standards. Templates, consistent tagging taxonomies, and clear rules for what counts as a complete project record reduce the cognitive overhead of contribution. The lower the friction, the higher the participation.

Third, content review cycles tied to pursuit activity. Firms that review and refresh content as part of active pursuit preparation — rather than on an arbitrary calendar — keep their knowledge base current without creating a separate administrative burden.

The technology layer should support all three, not substitute for them. The best platforms reduce the friction of contribution so that subject matter experts can share knowledge without becoming part-time librarians.

The Culture Question Technology Cannot Solve

Firms that achieve genuine knowledge-sharing cultures do not just deploy a platform and wait for adoption. They make the value tangible and immediate for every team member. They make knowledge sharing visible and rewarded rather than invisible and optional.

Culture change in knowledge management starts with leadership behavior. When principals and BD directors actively use and contribute to the firm's knowledge systems, the signal to the broader organization is clear. Recognition programs that credit staff for contributing high-quality, well-organized content create positive reinforcement loops that sustain participation beyond the initial rollout.

The ultimate test of a knowledge-driven culture is whether it survives the departure of key individuals. Firms that have built genuine knowledge cultures retain their institutional knowledge even through significant staff transitions. Those that rely on heroic individual contributors find themselves rebuilding from scratch each time a key person leaves.

That resilience is not a technology problem. It is a leadership and culture problem that the right technology makes significantly easier to solve.

Where to Start

Before investing in new tools, the highest-value first step is almost always a content audit: identifying where knowledge currently lives, how it is tagged, how outdated it is, and how many steps it takes for a proposal coordinator to find a specific reference.

That baseline reveals where the retrieval gap is widest. For most firms, the answer is not that they lack content — it is that the content they have is not findable. AI-powered search is the fastest available lever to close that gap, because it unlocks the value of material the firm already possesses.

The organizations winning today are not necessarily the largest or most tech-forward. They are the firms that have aligned their people, processes, and platform strategy to put the collective intelligence their teams have built over years of project delivery to work on the next pursuit.

Key takeaways:

• Knowledge retrieval is a direct driver of win rates, not a back-office efficiency issue
• Most AEC firms score well on Capture but fail at Curate, Connect, and Convert
• Governance (Policy) is the most underinvested dimension — and the most expensive to ignore
• AI-powered search delivers the fastest time-to-value by unlocking content firms already have
• Culture and leadership behavior determine whether technology investments sustain over time
• Start with a content audit before evaluating platforms

Ready to see what AI-powered knowledge management looks like at your firm? Explore Joist AI case studies or visit our homepage to learn more.

‍