What tacit knowledge is
The philosopher Michael Polanyi put it in a sentence that has become the standard definition: we know more than we can tell. A technician can hear that a gearbox sounds wrong. Ask her to explain how she knows, and she may struggle. The knowledge is real, but it lives in experience, not in words.
Explicit knowledge is the opposite: procedures, drawings, manuals and checklists. It can be stored, copied and searched. In their well-known work on how companies create knowledge, Ikujiro Nonaka and Hirotaka Takeuchi described how organizations create knowledge through a cycle of converting tacit knowledge into explicit knowledge and back (the SECI model). The point for any employer is simple: if that conversion never happens, the knowledge stays with one person.
Why it matters for institutional knowledge
Institutional knowledge is everything an organization knows that is not in its documents: how things really work here, why a decision was made, what has already been tried. A large part of it is tacit. It is also unevenly spread. In plants, workshops, IT operations and purchasing, a handful of people often hold the answers that everyone else depends on.
When one of them retires or changes jobs, the loss is rarely visible on the day. It shows up months later, when a line stops, a rare fault appears or a contract needs renegotiating, and nobody knows how it was handled last time.
Why it is so hard to capture
- Experts do not know what others do not know. What feels obvious after twenty years never gets mentioned.
- Nobody has the time. Documentation competes with the day's real work, and the real work wins.
- A blank page is a poor prompt. “Write down everything you know” does not produce useful text.
- It is too late when you start. Knowledge transfer started in the last month can move only a fraction of what is there.
Methods that work
Apprenticeship and job shadowing
A newcomer works next to the expert and asks questions in the moment. It is how most tacit knowledge has always been passed on, and it works well. But it needs two people's time at once, and it transfers the knowledge to one person only.
After-action reviews
After a fault or a near miss, the team talks through what happened, what gave it away and what helped. This captures real cases, but only the ones that happen to come up.
Story-based interviews
Instead of asking “how do you do your job?”, ask for specific events: “When did this last go wrong, and what did you do?” Concrete stories bring out what general questions cannot. The weakness is capacity. A good interviewer is scarce, and an expert's career takes weeks to cover.
Documentation, wikis and knowledge bases
Essential for what is already known to be important, but they record what someone thinks to write down. They go out of date unless somebody keeps them up.
A practical approach
- Find where the risk is. List the areas of expertise where only one person knows the answer and where a departure is approaching. Do not start with everything.
- Start early. A year before the departure is far better than a month.
- Keep sessions short and repeated. Ten minutes a week fits a working day. A two-day workshop usually does not.
- Ask specific questions. Use a checklist of topics for the field, such as system structure, deviations from the standard, changes, backups, alarms and suppliers in industrial automation.
- Let the expert approve. The expert should see and sign off what is written in their name. It keeps errors out, and it keeps their trust.
- Make it findable. Knowledge that nobody can look up when they need it has not been transferred.
An example
A fictional example. A process automation engineer has looked after a distributed control system for twenty years and retires in eighteen months. The documentation shows how the system was designed. She knows where it was built differently, which temporary fixes became permanent and what must never be loaded while the process is running.
A sensible plan splits her field into topics and asks a few concrete questions on each, in short sessions, once a week. After six months the company holds a set of approved, searchable answers in her own words, instead of a hope that someone will ask in time.
How Enarra Memo helps
Enarra Memo is an online service built around this approach. A manager records the key people, their areas of expertise and their departure dates, and a Risk radar shows whose knowledge is at risk and when. An AI interviewer asks each expert for about ten minutes at a time, using topics and questions from a field-specific knowledge pack, and follows up on general answers. The answers become knowledge cards, which the expert approves, edits or rejects. Only approved cards can be searched. Any employee can ask a question and get an answer directly from the expert's approved cards, with the expert and date as the source. If there is no answer, Enarra Memo says so instead of guessing.
Enarra Memo measures the risk of losing knowledge, not the performance of people. It does not score or compare employees. Data is stored in the EU, and published knowledge cards can be exported as Markdown and JSON at any time.
Where to start
- List the five areas of expertise your company can least afford to lose.
- For each, find out who knows them and when that person is expected to leave.
- Begin short, regular interviews with the person who is leaving first.