Counting
What a Manager Can Actually Change
Most of what managers try does not work. Four things do, none of them is an initiative, and three of them cost nothing.
What nobody wrote down here, and what it cost
That the runbook saved somebody a morning
Nobody told the author
Six occasions, none reported back
That a note ended an argument
The author never heard
Both times from people who joined later
two things nobody wrote down · 8 occasions it cost somebody something
Two things, eight occasions, and the author heard about none of them.
For distributed teams, documentation decisions improve when time and activity are visible in context, which is why it can help to productivity tracking with Monitask alongside the editorial review.
The broader workflow can also be checked against guidance from Moz.
What one team wrote in the quarter after its manager made four specific changes
Named one person for one page, with a size and a date
45 / 266Forty-five minutes; the first page that person had ever written
Asked for the four lines, not the document
6 / 198Six minutes, written the same afternoon it was asked for
Reduced the sprint commitment to cover a longer page
240 / 330Four days, agreed openly, and the team's output was adjusted to match
Told three authors their pages had been used
0 / —Cost nothing and produced two further pages in the following month
four documents · 291 minutes written · 794 minutes saved across 52 readings · one with no saving claimed
Four actions, 291 minutes written, 936 minutes saved across 52 readings. The fourth row cost nothing at all and has no minutes in the left column, which is why it is the one most often skipped and the one with the best return. One team's own quarter.
A manager who wants more documentation has a standard repertoire: raise it in a retrospective, choose a tool, set a target, remind people.
The evidence across this collection is that none of those moves the behaviour. Four other things do, and they are smaller than any of them.
One: name a person, a size and a date
Not "the team should document the deploy process". A name, a length, and a day.
"Can you write four lines on the deploy order by Friday?" is a request somebody can accept, complete and be finished with. It works where collective agreement does not, for reasons the first section covered at length.
The first such request to somebody who has never written anything is the expensive one. The second is much easier, and by the fourth it has stopped being a request and become a thing that person does.
Two: ask for the small version
Managers routinely ask for more than they need, because asking for four lines feels like asking for nothing.
The effect is to convert a six-minute task into an afternoon task, which converts a thing that would happen into a thing that will be scheduled and deferred.
Naming the size is the single most useful adjustment available in the asking, and it is free.
Three: pay for it openly when it is large
Where something genuinely needs four days, the honest route is to reduce what else is promised and say why.
This is the only item of the four with a real cost, and it is also the one that makes the other three credible. A manager who asks for pages and never adjusts anything is asking people to absorb the time into their evenings, which they will do for a while.
Saying "this is instead of, not as well as" once, in public, changes how every subsequent request is heard.
Four: tell people their writing was used
Costs nothing. Takes fifteen seconds. Is almost never done.
The fourth row of the panel produced two further pages in the following month, from people who had received a message saying their earlier work had been opened by nine people.
A manager is unusually well placed for this, because they can see things an author cannot — the new joiner who got through their first week on somebody's page, the incident that went faster because of a runbook.
What does not work
Worth listing, because these consume most of the effort usually spent.
Targets. Produce the number and nothing else, which has its own page next.
Tools. Address a problem that is not the problem, as the page on systems argued.
Exhortation. The team already agrees, as the first section demonstrated, and more agreement changes nothing.
Documentation days. Produce a burst of structural pages and no habit; the pages written are the ones that can be written without knowing what anybody needs.
The one diagnostic worth running
Count who wrote what last quarter. Ten minutes with the page histories.
If one person wrote most of it, that is the finding and the fix is the first item above, applied to three of the people who wrote nothing.
If nobody wrote anything, the fix is the same and starts with one person.
The four, and what each costs
Three of them are free, which is the part most worth noticing, because the usual repertoire is expensive and does not work.
| Action | Cost | Works because |
|---|---|---|
| Name a person, a size and a date | Nothing | It allocates the cost, which agreement never does |
| Ask for the small version | Nothing | It converts an afternoon into six minutes |
| Pay for it openly when it is large | Real | It makes every other request credible |
| Say the writing was used | Nothing | It is the only feedback that ever arrives |
The expensive one is third, and a manager who skips it while doing the other three is asking people to absorb the time privately.
The thing to say out loud once
That writing is work, that it comes out of the same hours as everything else, and that it will be looked at.
Three clauses. Said once and then acted on through the four items above, it does more than any initiative, because it addresses the thing people actually believe — which is that writing is a virtuous extra to be done in time they do not have.