Who
The Newcomer Cannot Tell What Is Unusual
Somebody in their first weeks can see what is strange about a place, which nobody else can, and cannot yet tell which of it matters.
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.
Documentation is easier to defend when the team can connect it to measurable time saved; the feature page provides an operational reference for that conversation.
The broader workflow can also be checked against guidance from SurveyMonkey.
One team's onboarding pages, and the weeks somebody opened each, across most of a year
Written in week nine by a joiner in their sixth week
Opened once per subsequent joiner, which is what it is for
The same page, after a second joiner added to it
Three more openings in the following quarter +2 beyond week 40.
A general introduction written by a founder in week one
Never opened by anybody, including its author
A list of surprises written by a joiner in week five
Opened six times, twice by people who had been there two years
four documents · 11 weeks in which somebody opened one of them · the longest wait between writing and a reading was 29 weeks · one never read at all
Four documents, thirteen weeks in which somebody opened one of them, and one never read at all. The one never read was written by the person who knew the most. The one opened most often was written by somebody five weeks in, who was recording things that surprised them rather than things they judged important. One team's own three quarters.
A person in their first month has exactly one documentation capability that nobody else on the team has, and a corresponding blindness that makes it hard to use.
They can see what is strange. They cannot tell what is important.
What the first weeks let them see
Everything that a long-serving person has stopped noticing is, to a newcomer, still visible as a choice somebody made.
That there are two dashboards. That the deploy pauses during month-end. That the client's week starts on a different day. That one service is named after a project that ended four years ago.
None of this is secret and all of it is unwritten, because the people who could write it no longer see it. The newcomer sees it constantly, and spends their first weeks quietly accumulating a list of things that made no sense until somebody explained.
That list is the single most valuable unwritten document in most teams, and it evaporates on a known schedule.
The window, and that it closes
Roughly six weeks. Some people hold it for two months, almost nobody for three.
What closes it is not forgetting. It is normalisation: the facts stop being facts and become the shape of the world, and at that point the person has the same blindness as everybody else and will never again be able to see what they could see in week three.
A team that does not use this window loses the same asset with every hire, repeatedly, without noticing, because the loss leaves nothing behind.
What they cannot do
They cannot judge importance, and asking them to is the usual way this goes wrong.
Asked to "write up anything that would help the next person", a newcomer produces a document that is partly valuable and partly a careful explanation of things that are standard everywhere, because they have no way of telling which is which.
They also cannot tell whether something is deliberate or accidental, which matters: a page describing a workaround as though it were the design is actively misleading.
The request that works
Ask for surprises, not for documentation.
"Write down everything that surprised you in your first month" is a task they can complete without judgement, in an hour, and the output needs no accuracy about importance because that judgement is supplied afterwards by somebody else.
The instruction is narrow on purpose. Not "what was confusing" — confusion gets resolved and forgotten. Not "what would help the next person" — that requires the judgement they do not have. Surprise is a feeling they can still remember and can list.
What to do with the list
Somebody who has been there two years reads it and does three things.
Marks the items that are true and unwritten. These become short pages or lines added to existing ones, and they are the yield.
Marks the items that are standard practice elsewhere. These need no page and are useful anyway, because they show where the induction conversation was thin.
Marks the items where the newcomer's impression is wrong. These are the most valuable of the three, because a wrong impression that a reasonable person formed in their first month will be formed again by the next person, and that is a sign to write something.
An hour of the newcomer's time and twenty minutes of somebody else's.
Pairing them with the expert
The stronger version combines this page with the previous one: the newcomer writes and the expert corrects.
The expert supplies accuracy, which only they have. The newcomer supplies the questions, which only they can still ask. Neither could produce the page alone — the expert would omit the obvious, the newcomer would get the details wrong.
It is also the cheapest arrangement for the expert, who spends twenty minutes talking and ten minutes correcting rather than an afternoon writing.
The part that makes it easy to ask
Newcomers generally want a task they can do well while they still cannot do much.
Asking somebody in week five to write up their surprises is not an imposition; for most people it is a relief, because it is useful, it is bounded, it requires no prior knowledge, and it is visibly their own contribution rather than a thing they were helped through.
That is a rare combination and it exists only in a window that closes.
The request, word for word
Asking badly produces a general document; asking precisely produces the list. The difference is a single noun.
- Works: "Write down everything that surprised you in your first month."
- Works: "What did you have to ask somebody, that you expected to find written down?"
- Fails: "Write up anything that would help the next person." — requires judgement they do not have.
- Fails: "What was confusing?" — confusion gets resolved and forgotten within days.
Surprise is a feeling people can still recall at five weeks. Importance is a judgement they cannot make until they can no longer see the surprises. Asking for the first and supplying the second yourself is the whole method.
What this does not produce
It does not produce the deep documentation — the billing reconciliation, the reasons behind the architecture. Those need the expert and are the subject of other pages here.
What it produces is the layer of short, obvious, precedence-shaped facts that cost everybody twenty minutes at a time and that nobody who knows them can see. That layer is cheaper to obtain than any other, and it is available once per hire.