Handing work over so that it survives you

What actually transfers, and what quietly doesn’t

A handover moves two different kinds of knowledge and is usually built to move only one of them. The explicit kind — the steps, the logins, the calendar, the names — is easy to list and is what most handover templates ask for. The other kind is the reasoning: why the step exists, which parts are load-bearing, what was tried before and abandoned, which deadline is real and which is decorative.

The gap does not show immediately, which is why so many handovers are judged successful and then aren’t. For the first fortnight the new person runs the documented path and everything works. The trouble starts at the first exception, because exceptions are where judgement was doing the work, and judgement is the part that never got written down.

Write for the questions they will actually have

Imagine the person three weeks in, at a moment when something has not gone to plan and you are unreachable. What do they need? Almost always the same short list: where the thing lives, what it depends on upstream, what breaks most often and what that looks like, who to contact when it does, and which of the current practices are deliberate rather than accidental.

That last one repays more effort than the rest. Every inherited process contains steps that made sense once and steps that are simply residue, and someone new cannot tell them apart. If you mark which is which, you save them a year of either preserving pointless work out of caution or removing something load-bearing because it looked pointless. Nobody else can supply that distinction after you have gone.

Overlap is expensive and hard to replace

Written material handles the routine. The tacit part — how a conversation with a particular team tends to go, what the tone of an early warning sounds like, when to escalate — transfers mainly by watching. That is why a week of overlap does more than a thirty-page document, and why organisations that give none pay for it later in ways nobody attributes to the handover.

When overlap is impossible, which it often is, the partial substitute is to record the reasoning rather than the procedure. A short note explaining why a thing is done the way it is beats a longer note explaining how, because the how can be reconstructed from the system and the why cannot be reconstructed from anything.

Nobody is incentivised to do this well

Bad handovers persist because of who bears the cost. The person leaving has already mentally gone and gains nothing from the quality of the document. The person arriving is not there yet and cannot say what they will need. The manager in the middle is covering a gap and is busy. The result is a form filled in on the last afternoon, and everybody involved behaved rationally.

Knowing that changes what you do about it. If you want a decent handover from someone leaving, ask specific questions early rather than requesting documentation late, because specific questions are easy to answer and a blank template is not. And when you are the one leaving, understand that the quality of what you leave is one of the few things that reaches people who will be asked about you years afterwards.

A handover is also a record of your own work

There is a self-interested reason to do it properly that is rarely mentioned. Writing down what you were responsible for forces you to state the scope of the job as it actually was, which is frequently larger and stranger than the job description implied. That document is a better memory aid than anything you will reconstruct months later for an interview.

It also disposes of the awkward category of things you were doing that nobody knew about. Every role accumulates a few, and they are invisible until they stop happening. Listing them is not self-promotion; it is the only way anyone finds out that the reason a monthly report was accurate is that someone was checking it by hand.

Where the advice runs out

How much of this you can do depends on how you are leaving and on what the place is like. If you are going in acrimonious circumstances, or the notice period was cut short, or you are being moved off something with no successor named, then the ideal handover is not available to you and it is not your failure. Do the compressed version: one page, honest, with the exceptions and the contacts.

It is worth saying that some organisations do not value this at all and will not notice either way. Others treat it as a fair proxy for how someone works, and the people who received a clean handover from you tend to remember it, since the alternative is a specific and memorable kind of misery. Which sort of place you are in is worth knowing before deciding how much of your last week to spend on it.