Brokered the first deal by hand before writing a line of platform code

Verdict

Case study TL;DR

The founder

This founder wanted to build a marketplace connecting freelance video editors with SaaS companies that needed short-form content for social media. Instead of building the platform first, he decided to broker the first deal manually.

The problem

Marketplaces have the hardest chicken-and-egg problem: no buyers without suppliers, no suppliers without buyers. Most marketplace founders build the platform first, then struggle to get either side onboard. He flipped the order.

What changed

1. 1He found 3 SaaS companies on Twitter that were posting about needing more short-form video content. He DM'd each: "I know a freelance video editor who specializes in SaaS demos. Want me to introduce you? No fee, no commitment." Two said yes.
2. 2He found 2 video editors on Contra and Upwork who had SaaS demo work in their portfolio. He DM'd each: "I have a SaaS company looking for a 60-second demo edit. Interested if I make the introduction?" Both said yes.
3. 3He introduced one SaaS founder to one editor via email. The SaaS founder asked 'can you handle 4 videos a month?' The editor said yes. The SaaS founder asked 'how do I pay you?' — and the founder had his first transaction.
4. 4Only after the first deal closed did he build the platform. The code reflected exactly what that first transaction looked like: a brief, a timeline, a price, a payment link. No marketplace features, no reviews, no discovery — just the transaction loop that had already worked.

The result

Key lessons

• Build the transaction, not the marketplace. A marketplace is a software problem. A transaction is a people problem. Solve the people problem first (make one deal happen) and the software problem becomes a spec, not a guess.
• Brokering is faster than building. Two weeks of manual introductions proved the deal structure, pricing, and customer willingness to pay. Two weeks of building would have produced a landing page with no users.
• The first transaction defines the product. Every feature you need is whatever that first deal required. Everything else is scope creep.

Related terms