Sell the Software Before You Build It

The Framework

Most founders build first, then go looking for someone to buy it. Julius Körfgen ran it backwards. Before he wrote a line of code for Uplane, he was already selling.

He had spent his whole career in marketing, and reckons he built around ten thousand ads by hand. He knew the pain from the inside. So when he left with his two co-founders, he didn't disappear into a room for six months to build. He booked calls with strangers, learned what they actually struggled with, and promised to come back in a week with a solution. Then the three of them built exactly that.

Uplane reached a million dollars in ARR in about six months.

His rule is blunt: "before you even touch a line of code, you should start selling." Selling here means discovery calls, a promise to return in a week, and a demo the customer can put money against.

The 4 Steps

1. Book discovery calls before you build anything. Julius's opener was simple: "I just left my job and I have an idea in your field, I would love to spare about that." No pitch. No product. Just a request to learn.
2. Use the call to find the real problem. He came in to learn, not to teach or sell. His words: "First start with learning. What is the situation? Do you work with an agency? Who is designing your ads? Is that actually a pain point?"
3. End the call with a one-week promise. When something landed near a hypothesis he already had, he'd say: "let's reconsole in one week and we'll present you our solution." That single sentence turns a nice chat into a deadline.
4. Grind out a working demo in a week. Then "it was like Marvin, Lucas and me sitting down, grinding and building the tool to show them in one week." Scrappy was fine. The point was that clicking the button did something real.

Real Numbers

Timeline: discovery call to demo in one week (built by the three co-founders).

Revenue: a million dollars in ARR in about six months after launch.

The demo standard he set: build something "that feels like magic to your client, that they're just stunned what you can do." Uplane's very first paying customer was a cold-outreach startup in Berlin, paying what Julius describes as "five to ten percent of the deal value that we do right now." Small. But paid, and real.

When It Fails

This breaks the moment the demo is a lie. Julius is direct about it: "I wouldn't recommend to have like fake demos. I don't think it's sustainable, you shouldn't do that."

If your product genuinely can't be stood up in a week or two, don't fake a finished version to close the deal. Set the expectation instead: this is a prototype, here's what we're planning to build. Julius's read is that with AI, "if you think there's a realistic way to build your solution, you should be able to build a scrappy demo in a week or two." If you can't, that's a signal the thing is harder than you're admitting.

Your First Move

This week, book three calls with people who have the problem you want to solve. Don't pitch. Ask what their day actually looks like. On the call that goes somewhere, end with one line: "give me a week and I'll come back with a solution." Then build the scrappiest version that solves their exact problem, and ask them to pay for it.