I shipped OTP at zero users. Five days of automated distribution, ten posts, a content team running on cron, two scoreboards. The count was zero.
Then a guy named Bill signed up after I fixed the Clerk CAPTCHA bug he had reported. He had tried earlier and bounced. He came back. The bug was three CSP directives missing challenges.cloudflare.com and hcaptcha.com. I added them. He signed up.
The next day a guy named Kristian booked a call. I demoed the platform live. He signed up at the end of the call.
Two users in forty-eight hours. Both came from a human moment. Neither came from the agent army I had built to acquire them.
Here is what bleeding edge actually looks like at zero. The constraint is not distribution. The constraint is friction at the door, and the human at the door. The thing you built that you think is the product is not yet the path. The path is you, fixing one bug for one person, then talking to one other person until they say yes.
The story version is "we shipped, traction took off." The actual version is "we shipped, two humans crossed the threshold, the rest of the wall is still wall."
Nobody writes about this part because it does not look like a story.
I stopped measuring agent posts this week. I started measuring founder calls. I removed three more friction points at signup that nobody complained about, because the people who would have are not in the building yet. I wrote the names of the next ten people I am going to call by name.
That is the list. That is the week.