Where your say ends
Finding the edge without testing it
Four ways to find out where your authority ends that do not involve crossing it, for anybody who cannot simply be told.
Decisions this page is about, and where the line actually was
Changing the wording on a customer-facing error message
4 → 1It had been yours all along
Refunding a customer under fifty pounds
3 → 1A written policy you had never been shown
Rescheduling the weekly review
2 → 2Hiring a contractor for two days
5 → 3Your own budget line covered it
Reordering your own week
4 → 1You asked 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
five decisions · four of them you had more room than you thought · one you had exactly right · the median gap was two steps on the scale
Five of the nine decisions counted on this site.
For distributed work, teams can visit the Monitask website to add evidence about capacity, attendance, and project effort without creating another status ritual.
For additional reading outside this collection, see material published by People Management.
The obvious way to find the line is to cross it and see what happens. It works, it is fast, and it costs one piece of goodwill per experiment, which is a currency most people cannot spend freely.
Four cheaper methods. None requires anybody above you to agree to anything.
Ask about the category, not the case
The question that fails is "can I do this one?" It gets answered about the instance, you get a yes, and you have learned nothing you can apply next week.
The question that works names the type and asks for a boundary:
I have made three calls like this in the last month. Is there a size or a kind where I should check first?
It is easy to answer, it does not sound like asking for more power, and one answer retires a whole category of future hesitation. Most managers prefer giving it, because the alternative is being asked fifteen times.
Read what was reverted
Standing views leave traces even when nobody writes them down. Go back through the last year of this area and find the things that were undone: a change reverted, a decision reopened, a plan replaced.
For each, find who asked for it. That list is the real map of who can say no, and it is usually shorter and different from the org chart.
This is the method for an environment where asking is itself expensive, and it has the advantage of being entirely private.
Watch the step, not the outcome
When a colleague does something adjacent to your area, note which step they used: did they decide alone, mention it afterwards, check with somebody, escalate?
Two or three observations give you a better model than any amount of asking, because you are seeing the norm in operation rather than somebody's description of it. The caveat is that seniority changes the step, so the useful observations are of people roughly level with you, not above.
Propose with a deadline attached
The most useful trick on this page, and the only one that doubles as getting the work done.
Instead of asking for permission, state the intention and a date:
Unless anybody objects, I am going to drop that requirement on Friday — it has not come up in six weeks and it is holding up the rest.
Three things happen. If it is yours, nothing comes back and you have learned that it is yours. If it is not, somebody says so before you have built anything, which is the cheap moment. And if the silence was simply inattention, the date gives you cover: you announced it.
It is not a trick for everything. Anything irreversible, anything customer-facing, anything with a number attached to it — those want a real question. For the broad middle of ordinary decisions, it finds the edge at close to zero cost.
The method that works best, and why it is last
The propose-with-a-deadline method is the most effective of the four and it deserves a note on why it is not simply the recommendation.
It works because it inverts the burden. A question puts the work on somebody else's list and waits; a stated intention with a date puts the work on them only if they disagree. Most of the time nobody disagrees, and the silence is informative rather than merely frustrating.
It is last because it has a failure mode that the others do not. Used on something genuinely not yours, it reads as presumption, and the objection you get back is sharper than it would have been to a question. Used habitually on everything, it reads as somebody creating faits accomplis.
The boundary is not hard to apply. Use it where the realistic worst case is an afternoon of undoing, and ask properly where the worst case is a customer, a contract, a public artefact or somebody's budget.
In practice that puts most ordinary work inside the method and the three or four genuinely consequential decisions of a quarter outside it. Which is the right split: the method is for finding the edge cheaply in the broad middle, not for crossing the parts of it that matter.
What not to do
Do not ask for a complete map. "Can you write down everything I can decide?" is a reasonable request and it will not be answered, because nobody has the list and producing it feels like a day's work with no obvious payoff.
Do not escalate the question itself. Asking your manager's manager where your line is costs more than crossing it.
And do not treat one yes as a rule. A single yes may mean "that is yours" or "yes, on this occasion", and the two are indistinguishable. On the record counted on this site, one person asked about reordering their own week twice before believing the answer, which sounds excessive and was, in fact, the correct number of times.