What you can't learn until you ship

This week I shipped the same feature nine times.

Not because it kept breaking. Because the problem kept revealing itself.

I was building a decision-making tool for someone else - a catalogue that had grown faster than anyone could manage, tens of thousands of items that needed to be evaluated one by one. Each round I shipped, something new showed up that the previous round hadn't caught.

Round six: two sections quietly displayed items they were supposed to hide. Round seven: a consistency audit found six real discrepancies nobody had named. Round eight: the whole approval model turned out to be too shallow. What looked like one decision per item was actually sixteen decisions, because the same item had been described sixteen different ways over the years.

Round nine, I stepped back and asked a different question: why couldn't the person who was supposed to use this just walk up to it and use it?

I'd built a tool that worked. What I hadn't built was a door.

So I built the door. A cover page. One plain paragraph explaining what to do. Twenty minutes. It should have been round one.

Shipping is the diagnostic

The Steve Jobs line "real artists ship" is usually read as a permission slip - stop waiting, launch the thing. I read it differently now. Shipping is not the finish line. Shipping is how you find out what you're actually building.

You cannot plan your way to that knowledge. You can diagram it and whiteboard it and reason it out in advance, and you will still be missing the thing you couldn't see until someone tried to use it. The gap between "the system works" and "a person can walk up to this and know what to do" is not a technical gap. It is a human gap - and you cannot measure it until something is live.

Nine rounds is not failure. Nine rounds is what thorough looks like when you're shipping instead of planning.

The entrepreneurs who worry me are the ones still on round zero.