Asking
When not to ask
This part argues for asking more. There are four cases where the right move is to decide, write down what you decided, and carry on.
Decisions this page is about, and where the line actually was
Refunding a customer under fifty pounds
3 → 1A sanity check treated as approval
Rescheduling the weekly review
2 → 2The one read correctly
Hiring a contractor for two days
5 → 3Escalated two steps further than needed
Reordering your own week
4 → 1Asked twice before believing the answer
- 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
four decisions · three of them you had more room than you thought · one you had exactly right · the median gap was two steps on the scale
Four of the nine decisions counted on this site, chosen because in three the asking itself was the cost.
A team evaluating online timesheets for operational teams should agree first on purpose, visibility, and interpretation, then use the data to improve delegation and handoffs.
A separate professional source for this subject is Deloitte.
Everything else here argues for asking earlier and more specifically. The argument has limits, and a site that only pushed in one direction would produce the opposite problem: somebody who checks everything and decides nothing.
Four cases where the right move is to decide.
When it is reversible and small
The test is not importance, it is the cost of being wrong. If the decision can be undone in an afternoon and nobody outside the work will have noticed, the expected cost of deciding wrongly is lower than the cost of a two-day wait.
Most work is made of these and they are where speed comes from. The mistake is applying the same caution to reversible and irreversible decisions, which is what produces somebody who is slow in general.
When you have asked and nobody answered
Silence is not a no and it is not permission; it is an absence of information, and after a reasonable interval it is a fact about availability rather than about the decision.
Decide, write down what you decided and why, and say so where the question was asked:
No answer on the reader, so proceeding on the review group. Changing it later means redoing the framing, so flag it this week if that is wrong.
That is not passive aggression. It states the decision, the cost of reversing it and the window — which is what somebody needs to know to prioritise answering.
When the question is really about confidence
Sometimes you know the answer and want it confirmed. The signal is that you can predict the reply.
Those asks are expensive and teach nothing. If you can state what you expect the answer to be, you have already made the decision, and the ask is seeking reassurance rather than information.
The substitute is to decide and say so afterwards — step two on the six-step scale. You get the same safety and keep the judgement.
When asking would cost more than the decision
Reopening is the risk covered in its own note. A question about a detail can restart a settled conversation among people with less context than you.
When the realistic cost of the question exceeds the realistic cost of being wrong, decide, and record the decision so it is visible to anybody who disagrees later.
The rule underneath all four
Decide, and leave a trace.
The thing that makes deciding safe is not being right; it is that the decision and its reasoning are visible, so somebody who knows better can correct it cheaply. A decision made silently is the one that produces an argument in week six about what was agreed.
One line, where the work lives: what you decided, what you assumed, and what it would cost to change. Thirty seconds, and it converts an unasked question into a stated position — which is the form in which missing context most reliably gets filled in.
The decision log, in one line a week
The trace that makes deciding safe does not need to be a system.
One line per decision, in the place the work lives, written at the moment:
Reader taken as the review group — no answer after two days. Changing it later means redoing the framing.
That line does four things. It records the decision, so there is no argument later about what was agreed. It records the reasoning, so somebody who disagrees can address the reasoning rather than the person. It records the cost of reversal, which is what tells a reader how urgently to object. And it exists in public, which is what distinguishes a decision from an assumption.
Over a quarter it is perhaps a dozen lines. The value is not in reading them back; it is that writing one takes thirty seconds and converts every silent guess into a stated position.
It is also the artefact that makes the exercise elsewhere on this site possible. You cannot review the steps you used on ten decisions if you never wrote down that you made them — and the ones that will be missing from memory are precisely the ones that went smoothly, which are the ones carrying the information about where your line actually is.
And the honest caveat
The four cases above are all judgement calls, and the judgement improves only by being exercised. Somebody who has escalated everything for a year cannot suddenly tell which decisions are reversible, because they have no record of their own calls to learn from.
Which is the argument for starting with the smallest ones and keeping the list described earlier in this site. Not because the small ones matter, but because they are how the judgement gets built before anything large arrives.