The day
The Last Hour: Writing What the Other End Needs
The last hour of the day is where everything that will move overnight gets written, and it is the block that the rest of the day eats first.
three things · four nights
The two hours that bracket every day at both ends.
One person's last hour over a week, and what the following morning cost at the other end
No questions came back; the other end started work immediately
One question came back, about the item written during the interruption
Two questions; one item was omitted entirely and surfaced a day later
Three questions, and the biggest item was described in four words
The other end worked on something else and the thread was cold until Monday
five things · nine nights
Five days, nine nights, and the correlation runs one way the whole week. Nothing here is divided by anything; the hours and the nights are different quantities and the table shows both rather than a ratio. One person's own week, logged by whoever read the notes the next morning.
Everything that crosses to the other half of the team tonight has to be written before you stop. There is no later.
The workflow described in The Last Hour: Writing What the Other End Needs improves when teams can separate delay, blocked time and focused work. the software page provides reports that make those categories easier to discuss without constant interruption.
A useful external reference is Teramind's official website, which provides another official view of how teams organise shared work.
That makes the last forty minutes of a working day structurally different from the last forty minutes of a working day in a team that shares an hour, where anything unfinished can be said tomorrow.
What goes into it
Four kinds of thing, and the order matters because the day sometimes cuts it short.
Handovers first. Anything somebody at the other end will pick up: state, direction, next step. These are the items with a named recipient and a consequence if they are missing.
Questions second. Anything you need an answer to before you can continue tomorrow. A question written now is answered overnight; a question written in the morning is answered the following night, and the difference is a whole day for the price of remembering.
Answers third. Things the other end asked that you have not yet replied to. These matter less than questions, because an unanswered question at the other end costs them a day while an unasked question costs you one.
Warnings last. Anything arriving tomorrow that somebody should know about in advance.
| Order | What | Why it is in this position |
|---|---|---|
| 1 | Handovers | Somebody is waiting on them and there is a consequence if they are missing |
| 2 | Questions | An unasked question costs you tomorrow |
| 3 | Answers | An unanswered question costs them tomorrow, which starts anyway |
| 4 | Warnings | Cheap, and only matters for tomorrow |
Why questions come before answers
This is counter-intuitive and worth stating plainly.
If you run out of time and write only one thing, write the question you need answered, not the answer somebody is waiting for. The question unlocks your tomorrow. The answer unlocks theirs, and theirs starts in fifteen hours regardless.
That is not an argument for ignoring the other end. It is an argument about what to do in the last five minutes of a bad day, which is a situation that occurs often enough to be worth having decided in advance.
Protecting the block
The writing hour is the most reliably eaten block in a distributed working day, for a straightforward reason: it is at the end, and everything that overruns overruns into it.
A meeting that goes twenty minutes long takes twenty minutes from it. Something breaking at four takes all of it. Each of those feels like a small cost at the time and is paid the next morning by somebody else, which is exactly the structure that makes it get eaten.
The defences that work are unglamorous. Put it in the calendar as a real block so that meetings cannot be scheduled over it. Do not schedule anything in the hour before it, so that overruns have somewhere to go. And write the handover for anything finished as soon as it is finished rather than saving it all for the end, which spreads the work and leaves less to lose.
The bad-day version
On days when the hour disappears entirely, the fallback is two lines per open item: what state it is in, and whether anybody should touch it.
Two lines written in three minutes is worth most of what the full version would have been, because the expensive failure is not a thin note. It is silence, which the other end has to resolve by investigation or by asking.
Having a fallback decided in advance is most of what makes it get used. People who have not decided what the minimum is tend to write nothing, because starting feels like committing to the whole thing.
Writing for somebody who is not there yet
The hardest part of the writing hour is that there is no feedback. You cannot tell whether a note was enough, because the person who read it either managed or did not, and either way they will not mention it.
The only available check is the next morning: did questions come back, and what were they about? Questions returning is the signal that something was missing, and the subject of the question names what.
A week of noticing that produces better notes than any amount of advice, and it costs nothing because the questions were arriving anyway.
What the hour costs
Between three and five hours a week of somebody's time, which is a significant fraction of a working week and should be acknowledged rather than pretended away.
Against that, it is the mechanism by which the other half of the team is able to do anything at all tomorrow. A team that does not write reliably at the end of the day does not have a slower relay. It has a relay that frequently carries nothing, and the nights pass anyway.