I built the quote funnel in an afternoon. It was for BatteryMe and it was good and I knew it was good.
Then a designer I used to work with looked at it. She looked at it a long time. Then she put her finger on the place.
"Here," she said.
It was not a bug. The numbers had not found it. It was a place where someone would stop and not know why they had stopped. The road went on but the reason for it ended there.
She was right. I rebuilt it. It was still the same day.
That is how it is now. You see the thing and then the thing is there. There is almost nothing in between.
You build it. You test it. You break it. You build it again.
Once it took a squad. The product person carried the problem. The designer watched the people. The engineer kept the machine honest. The tester went looking for the crack. They all looked at the same thing and they all saw it from where they stood, and between them they saw most of it.
Now I sit alone and the machine does their work. It writes the code. It makes the screens. It changes the words. It wires the numbers. And it tells you, very well, why all of it is good.
That is where the danger is.
It is not when the machine makes something bad. You can see a bad thing. The danger is when it makes something fine. Something that looks finished. Something so clean you forget to ask if it is true.
The old way was hard. People stopped you. Someone said no. The designer saw what you did not see. The engineer told you the nice idea would die in the real system, and they were right, and you hated them for a day. The tester found the hole you had walked past twenty times.
It was slow. But there were many eyes.
Now there is one pair. Mine.
So I am learning to make the friction myself. To stop. To push on the thing. To try to break what I have just made, before someone else breaks on it.
Speed changes the work. When a thing takes three weeks you think about it because there is nothing else to do. When it takes three hours you make ten more things instead.
The machine has made building cheap. Judgement is not cheap. It never was.
The best product work I got this week did not come from the machine. It came from her. She looked at what I had made and saw what I could not.
Maybe that is the one-person squad. Not one person doing every job.
One person who knows when they need another pair of eyes.