The hours
The four kinds of office hour
Every hour in the building is one of four things, and only one of them is the reason the building exists.
Days this part is about, hour by hour
Wednesday, the second week
3 of 7Planning day, the densest of the quarter
Thursday, the third week
0 of 7Seven hours of solo work at a hot desk
Wednesday, the fourth week
3 of 7The new starter's first day in the building
three days · 21 hours logged · six of them were only possible there · one of the three days contained no such hour at all · two were made worse by being there
Three of the twelve days drawn in full on this site.
Where work is split across locations, Monitask workforce analytics can make project effort and recurring bottlenecks clear enough for a transparent capacity discussion.
For further reading outside this collection, see material published by Mozilla Foundation.
Four days from the record, hour by hour
Tuesday, the first week
1 of 7The one hour was an argument that had been going round in writing for a fortnight
Wednesday, the second week
3 of 7Planning day
Thursday, the third week
0 of 7Nobody from the team was in
three days · 21 hours logged · four of them were only possible there · one of the three days contained no such hour at all · two were made worse by being there
Four days of the thirty-four, three of them drawn. One further day was left undrawn: the hours on them are not all accounted for, and a part-counted day would lower every total here without saying so.
There is a sorting question that takes about four seconds to apply and that almost nobody applies, because the hour is already happening and asking it feels like a complaint.
Could this have happened anywhere else?
Asked of a whole day it is useless — some of it could and some of it could not, and the answer comes out as a shrug. Asked of a single hour it is nearly always answerable, and the answers fall into four groups.
The four groups
- It could only have happened here. A whiteboard that settled something three weeks of writing had not. A piece of hardware. A person who sat down beside you because you were both there and stayed forty minutes.
- It was possible anywhere and better here. The conversation that would have worked on a call and worked faster in a room, or came out warmer than it would have done.
- It was identical anywhere. Solo work at a desk. The large middle of almost every job.
- It was made worse by being here. A call to four people who were all elsewhere, taken from a corridor because the rooms had gone by half past nine.
The fourth group is the one people do not expect to find, and finding it changes the shape of the argument. A day in the building is not neutral ground that some good hours happen to land on. Some hours are actively worse for it, and they are not rare.
What the proportions were
On the quarter counted on this site — thirty-four days, two hundred and thirty-eight working hours:
| Kind of hour | Hours | Share |
|---|---|---|
| Could only have happened there | 31 | 13% |
| Possible anywhere, better there | 44 | 18% |
| Identical anywhere | 118 | 50% |
| Made worse by being there | 45 | 19% |
Half of every day was the third kind. That is not a scandal; solo work is most of most jobs and it has to happen somewhere. The number worth reading twice is the last one, which is larger than the first.
Why the sorting has to be done at the time
The retrospective version does not work, and it fails in a particular direction. Asked in December how October went, people remember the planning day and the new starter's first week — the vivid hours, which are disproportionately the first kind. The corridor call and the four hours of solo work at a strange desk do not survive the month.
So the memory of a quarter in the building is systematically better than the quarter was, and both sides of a return-to-office argument are arguing from the same distorted sample, which is part of why the argument does not resolve.
What it is not for
It is not an argument for never coming in, and the record here does not support one. Thirty-one hours is not nothing. It contains the planning day, the week the new person arrived, and the argument that writing could not settle after a fortnight of trying.
It is an argument against the unit. A policy that counts days is counting the container. What anybody actually gets is the contents, and the two are only loosely related — which is why the same three days a week can be worth the journey one week and worth nothing the next, without a word of the policy changing.
The borderline cases
Three kinds of hour sit awkwardly between the groups, and how they are handled determines whether a record is worth anything.
The hour that was better because of who was there, not where. Two
people worked well together for an hour, and would have worked just as
well on a call, and the reason it happened at all was that they were
both in. This is b, not n, and the distinction is worth keeping:
the hour was possible elsewhere and was occasioned by the building.
Collapsing the two makes the first category meaningless, because
almost anything can be said to have been occasioned by proximity.
The hour that was worse for reasons that had nothing to do with the
office. A bad night's sleep, a difficult message, an hour lost to a
problem that would have been lost anywhere. This is s. The question
is not how the hour went; it is whether the location changed it.
The hour spent travelling between buildings. Not an office hour at all. It goes in the travel column, which is where it is visible.
Three things that look like categories and are not
- Busyness. A crowded floor is not a category. Four of the five
days spent in a busy building with none of the relevant colleagues
present produced no
nhours at all. - Enjoyment. The question is not whether the hour was good. An
excellent hour of solo work is
s; a tedious meeting that genuinely needed eight people at a wall isn. - Effort. Whether you tried to make the day useful does not enter into it. The fifteen days that returned nothing were not days when anybody gave up.
Holding those three out of the classification is most of what makes the resulting numbers usable in a conversation, because each one is a place where somebody can reasonably say the record was measuring mood rather than location.