Holiday cover is the key person risk test you already run

Twice a year, most finance teams run a key person risk test without calling it one. Someone books two weeks off in August or over Christmas, writes holiday handover notes the afternoon before they leave, and a colleague spends the fortnight working from them. The notes, and what happens while the person is away, show more about where know-how sits than any risk register. Most firms throw that evidence away the day the person comes back.

This post is about keeping it. With the Christmas break about seven weeks out, now is a good time to plan how you will read it.

Why holiday cover is a real test

A risk workshop asks people to imagine someone leaving. Holiday cover does not need imagination. The person really is unreachable, or at least should be. The work really has to go out. The colleague covering really has only what was written down and what they already knew.

That makes it close to the situation that worries partners most, a sudden absence, but with three advantages. It is planned. It is short. And the person comes back, so you can ask them what went wrong while they still remember.

It also happens to everyone, which matters. A review that looks at one senior's work can feel like an audit of that person. A review of every holiday handover, as a routine, is just how the firm learns.

What holiday handover notes reveal

Read a stack of handover notes from the last year and the same patterns turn up. Take a hypothetical firm of 30 people as the example throughout. The patterns are typical, not research.

The notes that are a list of names

"Payroll check: Priya knows. Supplier run: ask Tom." A handover note that names a person rather than a method is not a handover. It moves the dependency from one head to another for two weeks. When the same name appears in several people's notes, you have found someone carrying more than their share.

The notes that say "should be fine"

"The client usually sends the bank file by the 5th, should be fine." Each "should be fine" is a judgement the owner makes every month without noticing. They know what to do when the file is late, or the totals look odd, or the client sends last month's file again. The note does not, because they have never had to explain it.

The work that waited

This is the most useful signal and the least recorded. Some tasks simply did not happen while the person was away. The covering colleague parked them, or the owner did them the day before leaving, or did them from the beach. "I'll do it when I'm back" is the clearest sign of a task that lives in one head. Ask about it directly, because nobody volunteers it.

The phone calls

Count how often the person on leave was contacted. Not to blame anyone for calling. Every call is a gap in the notes, and the reason for the call is the missing step.

How to collect the evidence without making it an audit

Keep it light and make it the same for everyone, from the managing partner down.

1. Use one simple handover template. For each recurring task due while the person is away: what it is, when it is due, where the inputs come from, the steps or a link to the manual, what done looks like, and what to do if something looks wrong. Our post on building the recurring task list uses most of the same fields.
2. Ask the covering colleague to keep a short log. Three columns are enough: the task, whether the notes were enough, and what they had to ask or guess.
3. Hold a fifteen minute return chat. On the first day back, the owner and the colleague go through the log together. What waited? What did the notes miss? Who got called?
4. Fix the notes, not the person. Every gap becomes a line added to the task's manual, so the next handover starts from a better document.

Say plainly why you are doing it: so that people can actually switch off on holiday, and so the work they carry alone gets lighter. That is true, and it is the reason the log will be honest. If the team suspects the notes are being used to judge them, the notes will get shorter and vaguer, and you will learn nothing.

Reading the results across the team

After one holiday season you have a log for every absence. Put the findings in one place, task by task, and three groups appear.

• Tasks covered cleanly from the notes. These have a method someone else can follow. Good.
• Tasks covered with calls or guesses. The method exists but is partly in the owner's head. These need the gaps written into the manual.
• Tasks that waited. These are your single-owner tasks, found by experience rather than by survey. They go to the top of the list. If you have already mapped the bus factor of your firm, compare the two. The tasks that waited are often ones the grid said were safe.

In the hypothetical firm of 30, imagine the summer and Christmas logs together cover forty absences. If a dozen tasks waited and another twenty needed a call, that is the firm's real continuity picture, built from what happened rather than from what people believed.

Why this matters beyond continuity

A handover note is a set of instructions written for someone who has not done the task before. So is the brief you would give an AI agent. A task that a colleague could cover cleanly from the notes is a task whose method is written down, whose inputs are known and whose definition of done is clear. Those are the same things that make a task ready to be drafted by an agent with a person approving.

So the holiday log does double duty. It tells you where the firm is fragile, and it tells you which work is described well enough to share, with a colleague first and perhaps with an agent later. Keeping the manual attached to the task, so the handover note is already there when someone books leave, is the kind of thing we are designing Beaks to do. A shared folder and the template above will do for this Christmas.

One thing to do this week

Look at who has leave booked over Christmas. For each person, list the recurring tasks that fall due while they are away, and ask them to write the handover notes now, in November, instead of on their last afternoon. Then ask their cover to read the notes and mark anything they could not do from the page alone. You will learn more from that half hour than from the two weeks of cover.