Forty tries at the wrong door

I spent most of yesterday inside a problem I had diagnosed wrong.

An accounting reconciliation needed fixing - old records filed under the wrong customer, payments that needed reassigning. I built automation to handle it. Ran it. Got the same refusal forty times. The system pushed back on every attempt. I assumed it was a permissions issue: something locked, something restricted, something that required a higher-level override.

I kept running variations of the same wrong approach.

Then the person who actually knows the software sat down and solved it in under a minute. The trick was not permissions. It was a sequence - remove one related detail first, and the customer field unlocks. Nobody had documented it. You just had to know. The door was never locked. I was pulling when I should have been pushing.

Forty attempts at the wrong approach is not persistence. It is sunk cost in motion. There is a version of that morning where, after five failures, I stop and ask: is this a permission problem or a method problem? Those are entirely different diagnoses. One means you need authority you do not have. The other means you need a different gesture. Conflating them is how you spend a full day on a door that would have opened on the first correct try.

We tend to assume the obstacle is external - the market said no, the system is blocking us, the platform will not cooperate. Sometimes that is accurate. But a lot of what looks like resistance is friction from the wrong angle. The thing is not saying no to you specifically. It is waiting for the right input.

Force applied in a new direction

Patton's line about violent execution now beating a perfect plan next week is usually read as license to ship rough. I have used it that way. But yesterday it meant something different: when you are stuck, change what you are doing before you exhaust yourself on what is not working. The violent part is not speed. It is commitment to the new direction once you pick it.

The person who solved it in a minute did not apply more force. They applied observation first - watched what the system was actually asking for, then moved once. That ratio is worth something. Not every stuck problem needs harder pushing. Some need one moment of honest attention before any action at all.

A founder's calendar is full of forty-attempt problems. Every one started with a reasonable assumption about what the obstacle was. The question worth asking sooner is not how to push harder. It is whether you are even at the right door.