Asking for Forgiveness vs Permission
Balancing bias for action and being right.
Most “permission” is just a request for someone else to absorb the risk of your decision.
That reframe is the whole game. When you ask permission, you are asking another person to put their name on an outcome they can’t control and don’t fully understand. Of course they hesitate. Of course they ask for more data, another meeting, a smaller scope. They aren’t being slow. They are being rational. The default answer to “can I?” is “let me think about it,” and “let me think about it” is where good ideas go to get old.
So builders learn a different move. Do the thing. Show the result. Apologize if it broke something. This works more often than the cautious would like to admit, and it fails in ways that can end a career. The skill isn’t picking a side. It’s knowing which situation you’re in.
When forgiveness wins
The bias toward action pays off when three things are true at once:
The downside is reversible. You can roll it back, delete it, undo it, or quietly fix it before it matters. A feature flag, a draft, a test in a sandbox. If the worst case is “we revert the commit,” go.
The cost of asking exceeds the cost of being wrong. Some approvals take two weeks and three stakeholders to bless a change you could ship and measure in an afternoon. The meeting is more expensive than the mistake.
You’d be told no for the wrong reasons. Not because the idea is bad, but because no one wants to own the yes. This is the real reason “ask forgiveness” exists. It routes around people who are optimizing for their own safety, not the project’s success.
When those hold, asking permission is just volunteering for delay. You already know the answer is yes on the merits. You’re only missing a signature.
When it backfires
The same move that makes you look decisive can make you look reckless. Forgiveness is the wrong play when:
The blast radius is bigger than you. Anything touching money, customer data, security, legal exposure, or another team’s roadmap. If your mistake lands on someone else’s desk, you don’t get to skip their input.
It can’t be undone. Sending the email to the whole company. Deleting the production table. Telling the customer something you can’t walk back. Irreversibility removes the entire premise. There is no forgiveness, only consequence.
Trust is the thing you’re spending. The first time you act without asking and it works, you’re bold. The fifth time, especially if one of them went sideways, you’re the person who can’t be trusted in the room. You spend trust faster than you think, and it doesn’t refill on the same schedule.
The pattern people miss: the move scales badly. It’s a tool for the occasional high-conviction call, not an operating system. Use it on everything and you’ve just told your team that process is optional, which is a great way to find out how much process was actually load-bearing.
How to tell the difference
I run a fast version of this before acting:
Can I undo this in under a day? If no, ask.
If I’m wrong, who pays? If it’s not me, ask.
Am I skipping permission because it’s slow, or because I suspect the answer is no? If it’s the second one, that “no” is information. Don’t route around it. Go find out why.
That last one is the trap. “Easier to ask forgiveness” quietly becomes “easier to avoid hearing no.” Those are different.
The first is a judgment about process being too slow for a decision you’re confident in. The second is you overriding a constraint you haven’t bothered to understand. One is leadership. The other is how you blow up a project and call it initiative.
The honest version of this principle isn’t “don’t ask.” It’s “ask when the answer matters, act when the asking is the only thing in the way.” Most of the speed people admire in good operators comes from telling those two apart in about four seconds.
The cautious ask permission for everything and ship nothing. The reckless ask forgiveness for everything and eventually run out of it. The job is to spend your forgiveness deliberately, on the calls where being right and being fast are worth more than being safe.