The journey
When the travel is worth it
Four conditions, checkable the evening before, that between them predicted every good day in the quarter and none of the bad ones.
Days this part is about, hour by hour
Tuesday, the first week
1 of 7One hour that had been going round in writing for a fortnight
Tuesday, the fifth week
0 of 7Wednesday, the fifth week
2 of 7three days · 21 hours logged · three of them were only possible there · one of the three days contained no such hour at all · three were made worse by being there
Three days, two of which repaid the journey and one of which did not.
Where work is split across locations, transparent activity monitoring can make project effort and recurring bottlenecks clear enough for a transparent capacity discussion.
For further reading outside this collection, see material published by Berkeley AI Research.
Four trips at the two ends of the range
The quarterly planning day
2h 10m / 3h 00mEverybody in, a room booked, a wall and a disagreement
The day the new starter sat next to you
2h 10m / 1h 00mNothing scheduled; it happened because you were both there
A Thursday booked for focus work
2h 10m / 0mNobody from the team was in
A Monday you came in to be seen
2h 10m / 0mNo purpose beyond the count
four trips · 8 hours and 40 minutes door to door in all · 4 hours of it could not have happened anywhere else · 2.2 minutes of travel for each minute that required the room · two of the trips were made for nothing that needed the building
Four of the thirty-four trips, two from each end.
Across thirty-four days the range was wide: one day returned three hours that could only have happened in the building, fifteen days returned none at all, and the journey was the same.
Looking at the record afterwards, the good days are not mysterious. They share four properties, and the properties are knowable in advance.
The four conditions
Someone specific will be there. Not "it will be busy" — a named
person or two whose work touches yours. On this record every day that
produced more than one n hour had at least two such people present.
There is one thing that needs a room and a wall. A disagreement with a shape, a plan to lay out, a decision with four people in it. One is enough. A day with none of these will produce at most an accidental hour.
Somewhere to sit that is not a lottery. A booked room, or a desk near people you know. On the three days when a room was attempted and not obtained, the hours that were supposed to need the building became hours in a corridor.
The journey is the normal one. Not the strike, not the works. Not because a long journey cannot be worth it, but because the day has to be unusually good to carry an extra seventy minutes, and you can usually tell whether it will be.
How they combine
| Conditions met | Days in the record | Hours that needed the building |
|---|---|---|
| Three or four | 4 | 11 |
| Two | 8 | 13 |
| One | 11 | 7 |
| None | 11 | 0 |
The pattern is steep and it is not subtle. Four days of thirty-four met three or more conditions and produced a third of the quarter's valuable hours between them.
Using it without turning it into a rule
The temptation is to turn this into a personal policy: only go in when two conditions are met. That does not survive contact with a workplace, and it should not — half of the unscheduled hours exist because somebody was there without a specific reason.
The better use is as a sorting tool for days you are going in anyway. If the week requires three days and the conditions say Wednesday is strong and Thursday is empty, that is a reason to move Thursday to Tuesday, not a reason to stay home.
And if a whole month passes with no day meeting more than one condition, that is information about the arrangement rather than about you, and it is the kind of information worth raising — with the four conditions attached, so that the conversation is about what the days contain rather than about how many there are.
The evening-before version
Five minutes, the night before, with a calendar open:
- Does anything tomorrow involve somebody who will be in the building? Not a call — somebody physically there.
- Is there anything that would go better at a wall than in writing?
- Is a room booked, or is one obtainable?
- Is the journey the normal one?
Two yeses is the threshold on this record. It is not a sophisticated instrument and it does not need to be: the relationship between the score and the outcome was steep enough that precision would add nothing.
What to do with a score of zero
Not necessarily stay away — the day may be required, and the unscheduled hour depends on people being in without a reason.
What a zero is good for is choosing what to put in the day. A day with no overlap and no wall is a day to schedule the long run of solo work, not the thing that would have benefited from two people. On this record the opposite happened repeatedly: the demanding solo work landed on the crowded converged days and the collaborative work landed on the quiet ones, because nothing was being chosen at all.
Reversing that costs nothing and improves both halves.
The version for a whole team
The same four questions asked once, on a Friday, about the week ahead, by whoever runs the week. The answers determine one thing: which day is the day worth everybody being in.
Next week: Wednesday. The schema argument needs a wall and all four of you, and I have a room from ten.
One message. It answers all four questions for nine people at once, and it is the difference between a quarter with four good days in it and a quarter with thirteen.