Who
Who Does It When Nobody Has Been Asked
In most teams one person writes most of what gets written, nobody asked them to, and the arrangement is fragile in a way that is invisible until they go.
What nobody wrote down here, and what it cost
What the deploy order is
Known by one person who is in meetings
Asked rather than written, nine times
Why the import runs twice
Known by somebody who left in March
Nobody now knows, and it is asked at every incident
two things nobody wrote down · 14 occasions it cost somebody something
Two things, fourteen occasions, and one of them is now unknowable.
When a documentation effort needs evidence about where working time actually goes, teams can details from Monitask before deciding what should be written or automated.
Teams can compare this approach with the established material available from Capterra.
Everything one team wrote in a quarter, grouped by who wrote it
One engineer, four years in the team
210 / 532Four of the seven pages written that quarter, none of them requested
A joiner in their second month
90 / 242Two pages, both written after being asked by name
The team's manager
35 / 60One page, written after an incident review
Everybody else on the team, combined
0 / —Four people, no pages, and none of them had been asked
four documents · 335 minutes written · 834 minutes saved across 55 readings · one with no saving claimed
Four rows, 335 minutes written, 724 minutes saved across 55 readings. Over half the writing came from one person who was never asked and whose output is invisible in any plan. The last row is four people who were also never asked, which is the only variable the team controls. One team's own quarter, counted from the page histories.
Look at who actually wrote the pages a team has, and the distribution is almost never even. One person writes most of it. Frequently one person writes nearly all of it.
Nobody decided this. It is not in anybody's role. It usually goes unremarked until that person leaves, at which point the writing stops and nobody can say exactly when it stopped.
Who it turns out to be
Not the most senior person and not the person who knows the most.
It is usually somebody with a particular combination: they find writing easy rather than effortful, they are irritated by being asked the same thing twice, and they have enough standing to spend an hour on something nobody requested without anyone asking what they were doing.
That third condition does a lot of work. People two years in write; people in their first six months mostly do not, because an unrequested hour is harder to justify when you are still being assessed.
Why it is fragile
Four problems, and all of them are invisible while the arrangement is working.
The output depends on one person's disposition rather than on anything the team arranged, so it cannot be relied on and cannot be planned around.
It stops entirely when that person leaves, changes team, or has a busy quarter, and the stopping is gradual enough that nobody notices for months.
The subjects covered are whatever irritated that one person, which correlates imperfectly with what matters.
And it is invisible in every record, so when it stops there is no baseline against which anybody can observe the change.
What it does to that person
Two things, both slow and both unhelpful.
They absorb an uncounted load. Three or four hours a quarter is not much; three or four hours a quarter for four years, appearing nowhere, is a quiet grievance in waiting.
And they become the person it is assumed will do it. Requests start to drift towards them — "could you write that up?" — because they are known to write, which means the one person already doing it voluntarily acquires a backlog of other people's documentation.
Several of them eventually stop, and when asked why, say something mild that is not the real reason.
Making it visible first
Before trying to change the distribution, count it.
Page histories, commit logs, wiki edit lists — most tools record authorship and nobody looks. Ten minutes produces a table like the one in the panel above, and the table is usually a surprise to everybody including the person at the top of it.
That table is the argument. It is very hard to claim that documentation is a team value while looking at a quarter in which one person wrote eighty per cent of it and four people wrote none.
Spreading it without a rota
A documentation rota is the obvious fix and is a poor one: it produces pages written by people who had nothing to say that week, on subjects chosen by the calendar.
What works better is attaching the writing to events that already distribute themselves.
Whoever answers a question that has been asked before writes it down. Whoever handles an incident writes the four lines. Whoever onboards somebody writes what they had to explain twice.
Each of those is already distributed across the team by chance, so the writing distributes with it, and each is bounded and attached to a moment where the material is in somebody's head.
Asking the four people who wrote nothing
The last row of the panel is the controllable one, and the intervention is not a policy.
It is asking one of them, by name, for one specific page, with a size and a date. Most people who have never written anything have never been asked, and a surprising proportion of them turn out to be willing and to have assumed it was somebody else's job.
The first request is the expensive one. After somebody has written one page and had it used, the second is much easier to ask for and much easier to agree to.
Attaching writing to events that distribute themselves
A rota produces pages nobody needed. Events that already rotate produce pages somebody did, and they spread the load without anybody administering it.
- Whoever answers a question that has been asked before writes it down.
- Whoever runs an incident writes the four lines afterwards.
- Whoever onboards somebody writes what they had to explain twice.
- Whoever is on call writes what they had to work out at night.
Each of those falls on a different person each time, by chance rather than by schedule. Each is bounded. And each happens at a moment when the material is already in somebody's head, which is the only reliable condition for writing getting done.
The person who was doing it all
Worth saying directly to them, once: that you have looked at the numbers, that most of it was them, that nobody asked them to, and that you noticed.
It costs nothing. In most cases it is the first time anybody has mentioned it in four years, which is itself the finding.