Your Best Employee Is Retiring. Their Knowledge Isn't.
Institutional knowledge is the undocumented judgment your long-tenured people carry — the exceptions, the reasons, the “why we do it this way.” It’s lost when they leave because the standard playbook (a documentation sprint on the way out) captures the what and misses the why. The fix is to capture knowledge in the flow of work, make it answerable, and see key-person risk before the resignation, not after.
Key takeaways
- Institutional knowledge is the tacit judgment people carry — exceptions, reasons, relationships — not the documented process.
- The standard fix (a documentation sprint before someone leaves) captures the what and misses the why, and it’s always too late.
- The real loss is tacit: the exceptions and reasoning that were never written down because they lived in one person’s head.
- Capture knowledge as it’s used, not in a sprint — turn everyday work into an answerable, cited record.
- See it coming: score key-person risk early so you’re acting years ahead of the retirement, not days behind it.
Marta ran exceptions for nineteen years. Which clients get the old pricing and why. Why the Hamburg contract renews in March, not January. Which approvals can be skipped when the quarter’s closing and which absolutely cannot. None of it was written down, because none of it needed to be — Marta was there.
On her last Friday, the team threw a party. On Monday, the questions started. And nobody had the answers, because the answers had never lived anywhere but in Marta’s head. This is the most predictable crisis in any long-lived organization, and the strange part is how rarely anyone acts on it until the party’s over.
What is institutional knowledge?
Institutional knowledge is the undocumented understanding that keeps a company running: the exceptions, the reasoning, the relationships, the “why we do it this way” that never made it into a process document. Some of it is written down. The valuable part usually isn’t — because the valuable part is judgment, and judgment is hard to put into a wiki.
Why the standard playbook fails
The usual response is a documentation sprint. Someone’s leaving in six weeks, so we ask them to “write everything down” and shadow them for a while. It feels responsible. It doesn’t work, for two reasons.
First, it’s too late — six weeks can’t hold nineteen years, and the person is already half out the door. Second, and more important, it captures the wrong thing. Asked to document their job, people write the process: the steps, the what. What they can’t easily write is the tacit part — the exceptions, the judgment calls, the reasons — precisely because it’s second nature to them. The most valuable knowledge is the hardest to summon on demand, and a sprint asks for it at the worst possible moment.
The real problem is tacit knowledge
There’s a reason “just document it” never solves this. Explicit knowledge — the steps — is easy to write and not usually what you lose. Tacit knowledge — knowing that this invoice is the exception, that this client’s “urgent” means something specific, that this system breaks if you touch it on a Friday — is what actually walks out the door, and it only surfaces when a real situation calls for it. You can’t sprint your way to it. You can only capture it in the moments it shows up.
A practical plan to capture it before they leave
The approach that works inverts the sprint. Instead of a big extraction at the end, capture knowledge continuously, in the flow of work, while the person is still doing the job.
- Find your Martas first. Map which processes and systems depend on a single person, so you know where the risk is concentrated before anyone gives notice.
- Capture in the flow, not in a sprint. Turn the everyday — questions answered, decisions made, exceptions handled — into a record, as it happens, rather than a one-time interview.
- Make it answerable. Captured knowledge that isn’t findable is just a bigger archive. It has to answer a real question, with its source, in seconds.
- Keep it current. Because it’s tied to live work, it updates itself as the “why” changes, instead of freezing into another stale document.
Capture it as it’s used, not in a sprint
This is exactly what a Company Brain is for. As people work — answering questions, making decisions, resolving the weird cases — the knowledge accumulates into a governed, cited, searchable layer. By the time Marta retires, the answers to “why does Hamburg renew in March?” already live somewhere the team can reach, because they were captured the twenty times she answered it while she was still there. The cost of not doing this has a name and a number: the Context Tax you pay every time the knowledge isn’t there.
See it coming
The one thing worse than losing institutional knowledge is being surprised by it. Key-person risk is measurable — an enterprise twin scores exactly which processes hinge on a single person — which means you can act years ahead of a retirement instead of days behind it. Marta gave nineteen years of warning. The tools to have noticed were the only thing missing.
Frequently asked questions
What is institutional knowledge?
How do you retain knowledge when an employee leaves?
Why do documentation projects fail to capture institutional knowledge?
What is key-person risk and how do you measure it?
Capture what your experts know — before they go.
A Company Brain turns everyday work into an answerable, cited record of how your organization actually runs, so knowledge accumulates instead of walking out the door.