A few months back I read Turn the Ship Around, and one idea from it has stuck with me ever since: as a manager, your job is to push ownership down to your team, not pull it up to yourself. Every time someone asks you “can I do this?”, you’re quietly taking on a piece of their decision. The book’s fix was the “I intend to…” framework — a sailor states what they’re about to do, and unless the captain objects, they go do it.
I bring this up because a short post called “Ask for No, Don’t Ask for Yes” just hit the front page of Hacker News — 177 points, 85 comments — and it’s basically the staff-side version of the same idea. The author’s claim: when you ask your boss “can I do X?”, you’re asking them to share ownership of that decision. When you say “I’m doing X on Thursday unless I hear otherwise,” you keep ownership, and you don’t sit in their inbox queue waiting for a yes.
That’s a clean idea. The comments section is where it actually got useful.
Two camps, one real disagreement
The top comment nailed the mechanism: asking shares responsibility, announcing keeps it with you. That’s the whole insight. But the most upvoted pushback came from someone arguing that doing this with your own manager — especially with a deadline attached — can read as disrespectful. His example: telling your boss “I’m going to tell HR you approved this raise unless I hear from you by noon” isn’t a courtesy, it’s a trap.
Then a handful of actual managers and VPs jumped in and said the opposite — they want their senior people communicating exactly this way. One described it as the culture he’s building on his team: ownership and autonomy inside clear guardrails.
Both sides are right, depending on the situation. Reading through all 85 comments, four conditions kept showing up as the actual dividing line.
The four-condition test
Before you announce instead of ask, check all four:
- Scope — Is this genuinely your call? If it’s not actually your decision to make, announcing it isn’t a courtesy, it’s an end run around whoever’s job it actually is.
- Trust — Do you have a track record? If you’re new to the team or the relationship is untested, this isn’t your move yet. Earn it first.
- Context — Are you giving the full picture, not just a date? “I’m doing X Thursday” tells them nothing. What you’re doing, why, what else you considered, what happens if it goes wrong — that’s what lets someone make a fast call on whether to weigh in.
- Willingness to stop — If someone raises a real objection before your deadline, do you actually stop? The deadline is a courtesy window for feedback. It’s not a countdown to a default approval you’re hoping nobody notices.
If all four are true, announce. If even one isn’t, ask.
Where this shows up in accounting
Our profession trains people to over-ask by default — documentation culture and audit risk make “ask first” feel like the safe move, and a lot of the time, it is. Keep asking for anything with real liability exposure, client-facing commitments, or technical positions you’re not fully confident defending.
But think about the decisions that are genuinely yours that you ask permission for anyway, out of habit. Say a senior associate finds a $4,000 variance during fieldwork — clearly below materiality, fully documented. The over-ask version is “is it okay if I just note this and move on?” sitting in a manager’s inbox for two days. The announce version is: “Found a $4,000 variance, below our materiality threshold — documenting and closing out Thursday unless you want to discuss first.” Same decision, same care, but one of those creates work for someone else and one doesn’t.
The same test applies to how you direct AI tools, too. If Claude or another assistant keeps stopping to fire back clarifying questions before doing anything, you can flip the instruction: make reasonable assumptions, do the work, flag what you assumed at the end. You’re running the exact same scope-and-trust check — the only thing that changed is who’s on the other end of the message.
Key takeaways
- Ownership, not politeness, is the variable that actually matters. Asking shares it; announcing keeps it with you.
- Run the four-condition test before you announce: scope, trust, context, willingness to stop.
- Missing even one condition? Default to asking.
- Always attach a specific deadline — but treat it as a courtesy window, not a trap.
- The same framework works for directing AI tools, on the calls that are actually yours to make.
Want the CPE credit? Take the full lesson on EverydayCPE and earn 0.2 CPE credits: https://everydaycpe.com/courses/non-technical-communication-marketing/ask-for-no-not-yes/
Leave a Reply