There’s a person at your design firm or GC who knows things that few others in the organization know. Which subcontractors to trust for which scopes in your market. Which detail fails in your climate. What the structural engineer really meant by that note on sheet S-4.7. How to handle a certain type of owner when the schedule starts to slip.

When that person leaves (retirement, a competitor, burnout) how much of what they know goes with them?

Most of the time: almost all of it.

The expensive cost of knowledge in people’s heads

AEC firms have a knowledge problem that masquerades as a talent problem. The complaint is “we can’t find experienced people.” The reality is often “we can’t find people who already know what our experienced people know”, and those are very different problems with very different solutions.

Experienced knowledge lives in heads because firms have never built the systems to move it anywhere else. The after-action review that could have captured lessons from a difficult hospital project gets skipped because everyone is already on the next job. The senior superintendent who knows why the concrete sequence was changed on that data center never writes it down. The PM who navigated a particularly difficult owner relationship takes that playbook with her when she goes.

This is institutional amnesia. It compounds over time. Every wave of retirements takes another tranche of hard-won knowledge out of the firm permanently.

The result shows up on your projects. You pay for the learning curve every time. The team that’s new to your building type makes the same mistakes the previous team made on the previous project, because there’s no system connecting those two teams’ experiences.

What knowledge management actually is

It’s not an IT project. It’s not a software platform. It’s the organizational decision to treat what your people know as a strategic asset that needs to be captured, maintained, and made accessible, rather than assuming it’ll transfer naturally through proximity and osmosis.

Firms that do this well have built searchable libraries of past decisions and their reasoning. Not just “we chose this product” but “we chose this product because X, and these were the alternatives we considered, and here’s what we’d do differently.” Not just “the project came in over budget” but “the scope gap appeared here, driven by this, and this is the earlier signal we should have caught.”

That kind of institutional memory is the difference between a firm that learns across projects and one that starts over every time.

What KM 3.0 means in practice

The current generation of knowledge management tools, paired with AI, has made something genuinely new possible: the ability to make institutional knowledge searchable and queryable in natural language.

A junior PM who has a question that a senior PM would answer instinctively in thirty seconds. “how do we typically handle owner-furnished equipment in a healthcare renovation?”, can now get that answer from the firm’s documented experience rather than tracking down the right person and hoping they have time.

That’s not replacing mentorship. It’s extending it. It’s making the collective wisdom of the firm available at the moment someone needs it, rather than dependent on who happens to be down the hall.

The owner perspective

If you’re a building owner running a capital program with multiple projects, your partners’ knowledge management capacity directly affects your project outcomes.

A firm with strong institutional memory delivers more consistent results across projects because they’re not reinventing the wheel on your building type. They’ve seen your specific challenges before, and they’ve captured what worked.

A firm with weak institutional memory delivers results that depend on which specific individuals end up on your team. Get the experienced ones, you’re fine. Get the ones who are still learning the lessons the firm should have institutionalized ten projects ago, you pay for the education.

It’s a fair question to ask during selection: how does your firm capture and share lessons learned across projects? What happens to project knowledge when a key team member moves on? Can you show me an example of a decision your current team made better because of something learned on a previous project?

The honest answer tells you a lot. The vague answer tells you even more.