How to Transfer Knowledge From a Departing Employee
When a resignation letter lands you usually have four weeks. Most companies spend them filling in a handover form, and three months after the person has gone they run into questions nobody can answer. This is a plan for using those four weeks properly.
Why the standard handover does not work
The classic handover rests on three assumptions, and all three are wrong.
First: the person leaving knows what they know. They do not. To someone who has done the same job for ten years, that knowledge is not knowledge — it is air. You will hear “there is nothing special to explain, it is all in the system.” Three months later you find out it was not.
Second: what matters is the processes. The processes are already written down. What is not written down is when to depart from them — which customer gets an exception, which supplier’s promised date cannot be trusted, which approval is a formality. That is exactly the valuable part.
Third: if a handover happened, knowledge was transferred. A document having been written does not mean it was read, and having been read does not mean it was understood. Whether a handover worked can only be established by testing it, and almost no company tests it.
Do not tell the person leaving to “write down what you know.” Ask questions. People cannot fill a blank page, but they will answer a specific question put to them in surprising detail.
The 30-day plan
Ask the team: what did you go to them about in the last six months?
Do not make them write. Draw out the reasoning and the exceptions.
Every point they get stuck is a gap in the handover.
Ask the successor ten questions with the leaver out of the room.
Days 0–3: Find out what only this person knows
Do not ask the person who is leaving — ask the team. Put a single question to the three or four people who work with them:
“What did you go to them about in the last six months?”
The list that comes back is far more realistic than the headings on a handover form. It usually runs to fifteen or twenty items, and half of them are things nobody would have thought of.
Ask the same question of the person themselves, but differently: “What do people ask you about most?” Compare the two lists. The items that do not overlap are the riskiest.
Days 4–10: Ask the right questions
You have the list. Now you sit down and talk through each item. Record it — do not make them write, make them talk. People censor themselves when writing and explain themselves when speaking.
The question patterns that work:
- “Why do you do it this way?” — The reasoning is worth more than the rule. Rules change; reasoning adapts to the new situation.
- “When was the last exception to this?” — Exceptions are never written down, and that is where the real difficulty of the job lives.
- “If someone new took this over, what mistake would they make first?” — This question extracts, in one go, the traps the person learned over years.
- “On this subject, who do you trust and who do you approach carefully?” — The relationship map exists in no document and is the first thing to disappear.
- “Which decision you made three years ago still has effects today?” — It explains why the current setup is the way it is.
Half an hour per topic is enough. Two topics a day is twelve topics in six days.
Days 11–20: Let the successor do the work while the leaver watches
This is the reverse of the usual method. Normally the person leaving does the work and the successor watches — which teaches nothing, and merely gives the watcher a feeling of having understood.
Do the opposite: the successor does the work and the person leaving stands behind them. Every point where they get stuck is a gap in the handover document, and you fill it on the spot.
These ten days surface more than the previous six, because you are not trying to guess what is missing. What is missing shows itself.
Days 21–30: Test the handover
This is the step not to skip. Pick ten questions from the list in step one and put them to the successor with the person leaving out of the room.
What they can answer has been transferred. What they cannot is still at risk, and that is exactly what the remaining days are for.
If you do not run this test, you find out whether the handover worked six months later, through an unpleasant surprise.
If the employee has already left
You do not always get four weeks. Sometimes you notice only after they have gone.
In that case, reverse the order. Ask the team: “What question have you hit in the last month that you could not find the answer to?” That list is a map of the knowledge you lost. Then go back through the remaining documents, old correspondence and past projects with those questions in hand — most of the time, pieces of the answer are sitting there, scattered.
If your relationship with the person who left is good, a few hours of paid consulting costs less than months of archaeology. Most companies make it a matter of pride and lose.
Preparing for next time
The real point is this: not having to build the process from scratch every time a resignation letter arrives.
Knowledge transfer can be set up as a standing habit rather than a crisis procedure. Put the questions above to key people regularly, collect the answers in one place, and when the team has a question, look there first. Then when the resignation letter does arrive, what is left is not a four-week race but a few days of finishing off.
Checklist
Handover health
Frequently asked questions
What if the departing employee does not want to cooperate?
Common, especially when the departure was tense. Two things help: framing the handover as a contribution rather than a performance obligation, and where possible making it a written part of the exit process. Force does not work. One hour asked for in good faith yields more than ten hours imposed.
How much time should a handover take?
For a key role, fifteen to twenty hours of focused work is a reasonable target. Spread over four weeks that is an hour a day and disrupts nobody. Compressing it into a two-day marathon is the most common mistake.
Written documents or video recordings?
Record, but produce documents. A recording captures the knowledge; a document makes it usable. Nobody will sit through a two-hour video six months later to find an answer. Searchable text works.
Does a small company need this much process?
The risk is greater in a small company, because the knowledge is spread across fewer people. In a team of ten, one person leaving takes far more with them than one person leaving a company of a hundred. Shorten the process, but do not skip it.
Does it make sense to move this knowledge into an AI assistant?
Here is where it makes sense: as long as the collected knowledge sits in a folder, nobody looks at it. Faced with a question, a team would rather ask a person than run a search. When knowledge is reachable in question-and-answer form, the chance of it being used rises noticeably. What matters is that the assistant answers only from verified knowledge and, on something it does not know, points to the right person instead of inventing an answer.
Do not build this process from scratch every time
AIStore draws out your key employees’ knowledge through questions and turns it into an AI assistant your team can ask at any time. The assistant answers only from verified knowledge, and says so when it does not know.
See the enterprise solution → Book a call