Ten customer interviews. Completely useless. That’s what Vitalii, one of Stan’s two founders, told us on Wednesday night.
And we were feeling pretty good about ourselves. Tuesday was the first elimination day, and we’re through. That same day we ran six customer interviews, with recruiters, an operator from a big tech company and associates at VC funds. People looked right past our bugs. On one call Kobe generated intro recommendations for a conference live, and one of the users said something close to: “Fuck the UI library, send me the beta test now.” For an engineer who had been chasing crashes two nights earlier, that was a pretty good moment.
Daymond John was also at camp on Tuesday. His advice for us was simple: deliver the result right away and drop the generic free trial. After that we cleaned up the frontend. Outreach, the LinkedIn post feature, the network view and a few more menu items got cut. Focus.
On Wednesday two conversations turned the week around. John, Stan’s other founder, pushed us clearly towards B2B. Customers don’t want a new workflow, they want Kobe inside the tools they already use, for example through a Claude integration over MCP. And in the evening, Vitalii with the line from the top: interviews don’t pay. You need to sell.
He was right, and that was uncomfortable. I’m an engineer. My reflex is to make it stable first and show it later. That exact reflex kept us from asking the simplest question for ten calls straight. Do you want to be our first paying customer? The fear was always shipping something that isn’t ready. One line from a session stuck with me: perfectionism kills the startup. Early adopters knowingly buy something unfinished.
So we rebuilt the process. We set the product up live on the call. Then we ask whether they want to be our first customer. The payment link goes out while we’re still talking, with a money-back guarantee. On Thursday there was a live sales call as a teaching example, and its last line was “Would you like to become our very first paying customer?” Most of the founders in the room had never asked anyone for their credit card.
Same day, a session about co-founders, probably the most intense working relationship there is. One line from it now hangs over our table, at least in my head: it’s not you versus me, it’s you and me versus the problem. After that first night with René, it fits.
From then on the split was clear. Nils sells. René and I make sure what he shows holds up. On Thursday evening I merged René’s new V3 backend with my frontend and built a progress bar with time left and percentage for the first LinkedIn import, which takes two minutes or longer. Without a bar, that looks like a crash in a demo. Then bug fixes late into the night, pretty repetitive work, but the kind nobody sees on a call once it’s done.
Our first two sales came on Thursday, but way too cheap. An investor looked at our prices and told us we were selling it way too cheap and should go higher. On Friday we raised our prices. And for the first time, someone paid while we were still on the call.
Perfection feels safe. But it doesn’t sell, and it’s never done. Everyone says ship before you’re ready. I only really understood it this week.
Week one is done. We’re still in.