What autonomy needs
Why it is wanted, and who wants it
The purpose governs every trade-off the receiver makes alone. The stated reason and the real one are often different, and only the real one predicts anything.
Decisions this page is about, and where the line actually was
Picking the library for the new parser
1 → 4The architect had a standing view
Dropping a requirement nobody had asked about in six weeks
2 → 5A commitment made before you joined
Changing a database column type
1 → 1Small, reversible, tested once before
- 1Decide it alone and say nothing
- 2Decide it alone and mention it afterwards
- 3Check with one colleague first
- 4Check with the person who assigned it
- 5Needs a decision one level up
- 6Cannot be decided here at all
three decisions · not one of them gave you more room than you thought · two gave you less · one you had exactly right · the median gap was three steps on the scale
Three of the nine decisions counted on this site.
To keep measurement proportionate, teams can find the latest product details and agree what is collected, who can see it, and how it will support decisions.
For additional reading outside this collection, see material published by Wharton Knowledge.
Every piece of work contains a hundred small decisions that nobody will ever review: what to do when two requirements conflict, where to stop, what to leave rough, which edge case to ignore. The person doing the work resolves all of them, alone, in minutes, using one thing — their understanding of why the work is wanted.
Give them the wrong understanding and every one of those hundred decisions tilts the wrong way. The work arrives looking competent and being wrong, and nobody can point at the mistake, because the mistake was distributed.
Stated and real
The stated purpose is the one in the plan: "improve the import reliability". It is true, it is uncontroversial, and it predicts nothing, because every possible approach improves import reliability to some degree.
The real purpose is specific and usually embarrassing in some small way: a named customer complained twice in one month and the renewal is in March. Or: the board asked a question nobody could answer and the chief executive does not want to be asked again.
The difference matters because the real purpose decides the trade-offs. If it is the renewal, you fix the thing that customer sees and you ship before March, and the architecture can wait. If it is the board question, you need the numbers and a chart, and reliability is beside the point.
The quarterly numbers
On the record counted here, one handover arrived with the purpose and not the audience. The work was to produce the quarterly numbers. The person did it thoroughly, at the level of detail an analyst would want, and the audience turned out to be the board.
The work went back to the brief and was rebuilt from scratch for a different reader. Four stages back — the longest return on the record. Nothing about the first version was incompetent. It answered a question nobody had asked.
I asked what the numbers were for and was told, accurately, that they were for the quarterly review. It did not occur to me to ask who reads it, and it did not occur to anybody to tell me.
Why the real purpose goes unsaid
Three reasons, all of them ordinary.
- It feels like politics. Naming a customer, a renewal or an executive's discomfort sounds less dignified than naming an engineering goal, so the dignified version gets written down.
- It feels obvious. The person handing over has been in the meetings where it was discussed and cannot reconstruct what it is like not to have been.
- It is genuinely unknown. Sometimes the work has been passed down through two people and the purpose was lost at the first handover. This is worth saying out loud rather than inventing a plausible one.
The question that gets it
"Who asked for this, and what will they do with it?"
Two clauses, and the second does the work. "Why are we doing this" invites the stated purpose, because it sounds like a challenge and people answer challenges with the defensible version. "What will they do with it" is a practical question about use, and it produces the audience, the deadline behind the deadline, and usually the real reason, all in one answer.
The follow-up, when the answer is vague: "if we could only do half of this, which half matters?" Nobody can answer that without revealing the purpose.
The hundred decisions, named
It is worth making the abstraction concrete, because "the purpose governs the trade-offs" sounds like a slogan until you list what it actually governs.
On a document: how much background to include, whether to define terms, how long it can be, whether to show the method or just the result, which caveats to state and which to leave out, what to do when two sources disagree.
On a piece of software: where to stop, which edge cases to handle properly and which to log and move on, how much to invest in the parts nobody will see, whether to make it configurable, what to do when the clean approach is slower.
On anything: what "finished" means.
Every one of those is resolved in seconds by somebody who knows the purpose and agonised over by somebody who does not. And none of them is reviewable afterwards, because none of them looks like a decision in the finished work — they look like how it came out.
That is why a missing purpose is the only one of the four that cannot be caught by review. The other three leave a visible mark when they are wrong. This one leaves a hundred small marks that are individually defensible and collectively aimed at the wrong target.
For the person handing over
One sentence, named and specific:
This is for the board pack in March, because they asked a question in January that nobody could answer. It needs to be readable by somebody who does not know the system.
That sentence does more for the quality of the result than any amount of review afterwards, because it arrives before the hundred small decisions rather than after them. And if it feels too political to write down, that is worth noticing: the person doing the work is going to be operating inside that politics either way, with or without a map.