The most critical knowledge of your company is not in any system. It is generally in three people, have you wondered what happens when one of them leaves?
Almost no company has it documented in a manual, and yet, that knowledge supports daily decisions: why that customer is handled differently, what to do when the machine fails in that particular way, what exception applies in that process and why. No one wrote it down. It was transmitted, learned over the years, it became intuition.
Why this knowledge never reaches a system
It’s not due to negligence, it’s because documenting that knowledge is not urgent until the day it’s no longer there. While the person is still in the company, the knowledge works perfectly, it is available, you can ask, it resolves. The problem is that this availability creates a false sense of security, since the knowledge exists and works, no one measures how concentrated it is.
There is an important difference between information and institutional knowledge. Information can be stored in a system: prices, inventories, transactions, institutional knowledge is the criterion behind the information, why a decision is made one way and not another, what signals anticipate a problem before it appears in any report. That criterion is exactly what does not remain in any ERP, CRM, or shared folder, because it was not designed to live there.
Why this is also a technology problem, not just a people problem
When a company digitizes or automates a process, it almost always digitizes what it can see, such as the steps, the fields, the flow. What is rarely captured is the reasoning behind those steps, the exception that only that person knew how to recognize, the warning sign that anticipated a problem before it became a problem.
The result is a technically correct system that replicates the procedure, but not the judgment that supported it. It works well as long as the person who designed it is there to correct it outside the system, so, the day that person is not there, the system continues to function but no longer knows what she knew.
That’s why any technology project built without first understanding where that reasoning lives runs the same risk as the company that never documented it, meaning it automates the surface and leaves the most valuable part of knowledge exactly where it was, depending on a person.
The risk that almost no company measures
Most companies can name who is indispensable in the role—a manager, a team leader. Very few can name who, in practice, holds the reasoning that supports a critical process, and these are different questions.That risk is not seen in any turnover indicator, because it’s not about how many people leave, but about how much reasoning leaves with each person who exits, regardless of whether their departure was planned or a surprise.
How to start capturing that knowledge
It’s not about writing a two-hundred-page manual that no one will read, but about a deliberate process, with method:
• Identify the three or four processes where, if the key person were to be absent tomorrow, the business would stop or make costly mistakes. Not every process needs this level of documentation, only those that concentrate critical criteria.
• Document the criteria, not just the procedure. A procedure says what steps to follow. The criteria say why, in what cases it is done differently, and what signals indicate that something is going to go wrong. This last part is what is usually lost.
• Have the person teach the process to another, live, while someone else documents. Tacit knowledge rarely comes out well when someone is asked to “write it down.” It comes out when asked, “explain it to a colleague as if it were their first day.”
• Make that documentation part of the system, not in a separate file. A document in a folder becomes obsolete in months. A criterion incorporated into the workflow, like a validation, an alert, or a rule within the process, stays alive because it is used every day.
No company can completely eliminate dependence on people and should not try to. But it can decide, intentionally, which part of that knowledge will continue to live only in someone's head, and which part is now time to also live in the system.