What to change
Counting your own fortnight
The log behind this site took about twenty minutes a day. The shortest version that produces the same finding takes three days and one column.
Living roomThe room most often rearranged, and the one that resisted it
19 m²
Sitting
SittingPaid workCallsTelevisionCompany
11
Kitchen and dinerWhere the one change that worked was made
16 m²
Cooking and eating
CookingEatingPaid workLaundry and adminCallsCompany
14.5
Paid workDomestic workRest and sleepContact and care
Two rooms covering 35 square metres, built for two things. Between them they now carry eleven uses. Two of the two carry more than one, and kitchen and diner carries six. The median room is in use 12.75 of the 16 waking hours. The two rooms in which every attempted rearrangement happened.
Everything on this site comes from a log kept for fourteen days by a household that was not trying to prove anything.
It is worth describing what it actually took, because the useful version is shorter than the one that was kept.
What was kept
A note, made at the time or within the hour, whenever one of two things happened: Where flexible work changes both schedules and space, decision-makers can review this approach on the product site before setting the baseline, access rules, and review date.
- an activity left a room because another activity wanted it
- something happened that would not have happened from an office
Each note recorded the day, the rough time, the room, the two activities, how long, and which of the two people moved. Readers testing the same question can compare the record with guidance from Steelcase.
That is six fields and perhaps fifteen seconds a note. Across the fortnight there were a hundred and thirty-five notes, which comes to about twenty minutes a day including the evening tidy-up.
The shortest version that works
Three days and one question.
When something interrupts you, write down which room you were in, what interrupted you, and whether you moved.
Three days produces roughly ten entries in a household like this one. Ten entries is not enough for a median, a distribution or any of the numbers on this site.
It is enough for the two things that actually change behaviour: which room most of them happen in, and whether they happen to one person more than the other.
Both of those were visible in this record by the end of day four.
What the extra eleven days bought
Precision, and one finding.
The precision is not worth much. Forty-seven against an estimate of "about three a day" is a better number and leads to the same conclusions.
The finding is the chains — the seven collisions that happened because of an earlier collision — and those required enough entries to notice that a displaced activity kept being displaced again in the room it had gone to. Four days would not have shown it, and it produced the folding table, which was the best change made.
So the honest accounting is: three days gives you the shape, a fortnight gives you one more idea.
What not to do
Do not log everything. An attempt was made in the first two days to record all room use, which produced several hundred entries, took an hour a day and was abandoned. The log works because it only records exceptions.
Do not log how it felt. Tried, abandoned on day three. The two people's accounts of the same event differed often enough that a feeling column would have made every number a matter of opinion.
Do not keep it secretly. Both people logged, both saw the sheet, and that is the only reason the gains column exists — one person was recording interruptions and the other pointed out that nothing was recording the deliveries.
The sheet
Six columns, one line per event, no categories decided in advance.
The categories in this site — four kinds of use, six kinds of gain, four causes of collision — were all assigned afterwards, from the notes. That ordering matters: categories invented in advance determine what gets noticed, and the two most useful findings here came from patterns nobody had a column for.
Why a household would bother
Not to improve anything. On this record, a fortnight of logging produced about a quarter fewer collisions and no change to the distribution.
The stated reason both people gave, asked afterwards, was different and is the one this site would offer: they stopped attributing the difficulty to each other. A household that knows its distribution is set by two calendars and one quiet room argues about different things from a household that does not.
What the sheet looked like
Six columns: day, rough time, room, what arrived, what left, who moved. A seventh for the gains: day, time, room, what happened, which of the two.
Kept in a shared document on a phone, added to in the moment where possible and in the evening where not.
The household's assessment of the format, asked afterwards, is that the only column they would change is "rough time" — which was recorded to the nearest half hour and would have been more useful to the nearest hour-with-a-label: morning block, lunch, afternoon block, evening. Half-hour precision implied an accuracy the entries did not have.
The thing that nearly stopped it
On day four both people independently considered giving up, for the same reason: nothing interesting was happening.
Four days in, the sheet held perhaps a dozen entries, none of them surprising, and the exercise felt like bookkeeping for its own sake.
What kept it going was that neither mentioned it to the other until afterwards, so neither gave the other permission to stop.
That is worth passing on to anybody attempting this. The shape does not appear until somewhere around day six or seven, and the first week feels pointless from inside it. The two useful findings in this record — the chains, and the distribution — were not visible at all on day four and were obvious by day ten.
Also on what the rooms are being asked to do