What came back
The number that cuts against the title
This site is named after its warning. The figure most likely to undermine that warning is on the front page, and the reason is a rule rather than a gesture.
The daily stand-upBack twice a week instead of five, standing and timed
week 0 · crept back after 3 weeks
Chasing unpaid invoices every weekBack in full
week 3 · crept back after 2 weeks
The spare key logReplaced by something else doing the same job
week 10 · crept back after 14 weeks
Two people on every delivery
week 4 · crept back after 9 weeks
Crept back 4
Four removals were written down with the week each of them stopped. Every one of them produced an event inside the 78 weeks watched. Of these, four crept back. Four of the eleven that crept back without anyone deciding.
This site is called Nobody Missed It Yet. The name carries the warning: fourteen removals have produced no complaint, and the word yet says that the verdict is not in.
The single figure most likely to undermine that warning — the one a reader could use to dismiss the whole argument — is printed on the front page in the third panel.
That is not modesty. It is a rule, and the rule is worth stating, because sites built out of one organisation's record are unusually easy to write dishonestly.
The figure
Twenty-four of the thirty-eight came back, at a median of forty-seven and a half per cent of their old load. Combined with the fourteen that did not come back at all, the shorter week reduced this organisation's procedural load by substantially more than the count of permanent removals suggests.
In other words: the thing works better than the headline. A reader who wanted to argue that a four-day week is straightforwardly a good deal would reach for exactly this number, and they would be reading it correctly.
Why it is there
Because the record contains it. That is the whole reason, and it has to be the whole reason, or the record is not a record.
A site built on thirty-eight rows of a spreadsheet has one claim to be worth reading, which is that the rows were written down before anybody knew what they would say. The moment the rows get selected for what they support, the site becomes an argument illustrated with data, and there are a great many of those about shorter weeks already — in both directions.
The test is simple and it is not comfortable: if a figure in the record cuts against the thing the site is for, it goes on the front page rather than in a footnote.
What this costs
It costs the clean version of the argument.
The clean version says: shorter weeks are evaluated too early, the quiet removals are the dangerous ones, and the published accounts are measuring the wrong interval. All of that is in the record and all of it is, on this evidence, true.
The complicated version says all of that and then says: also, this transition recovered more hours than its own count of removals implies, most of what came back came back better sized than it went, and two procedures ended up with rules written down that had been missing for a decade.
Both are the same record. A reader who takes only the first half has been given a warning without the thing it is a warning about.
The symmetrical obligation
The rule cuts both ways and the other direction is the one that is easier to forget.
An organisation that wants to report a shorter week as a success has the same obligation with the five problems: on the front page, with their lags, including the one at sixty-four weeks and the one where three people left. Not in an appendix about lessons learned.
On this record the five are on the front page. So is the sentence saying that two of them would have been invisible at six months. So is the forty-seven and a half per cent.
What a reader should do with all of it
Not average it into a verdict. The record does not produce one, and a site that produced one from thirty-eight rows in one organisation would be claiming far more than it has.
What it produces is a shape: subtraction works better than the counts suggest, finds out about itself unevenly, and reports on its own success long before it has finished happening. An organisation deciding whether to attempt this should expect all three, and should expect the good news first.
The order in which the news arrives is not the order of its importance. Everything favourable about this transition was visible by week fourteen. Three of the five problems arrived after week thirty.
That sentence is the site in one line, and it contains the good news and the warning in the order they actually happened.
The other figures that cut the same way
The forty-seven and a half per cent is the largest of them and it is not alone. Four further things in this record favour the shorter week more than the site's framing suggests, and all four belong here rather than in a footnote.
Fourteen removals, permanently gone. More than a third of the list, producing nothing in seventy-eight weeks. That is a substantial quantity of work an organisation was doing for no reason anybody can now identify.
Two procedures acquired written rules — a threshold on second reviews and a size rule on two-person deliveries — that had been missing since each was introduced. Neither would have been written without the removal.
Six substitutions, at a median of thirty-five per cent, several of which appear to be better than the thing they replaced rather than merely cheaper.
Nobody on the record wanted to go back. Including the four people who absorbed the fixed coverage on a longer rota, and including the people whose objections were overruled.
Set against five problems across eighteen months, that is not a close call on this record, and a site that only printed the warning would be describing a different transition from the one that happened.