Kaya.gs
Kaya.gs was a modern, responsive online server for the board game Go. It was built to disrupt existing servers that relied on outdated technology. After raising $20,000 via a custom crowdfunding campaign and reaching thousands of registered users, the project failed due to technical instability, a lack of focus on core features, and a breakdown in founder morale.
The Autopsy
| Section | Details |
|---|---|
| Startup Profile | Founders: Gabriel Benmergui Funding: $20,000 (Crowdfunding via personal website/PayPal) |
| Cause of Death | Market Fit: Yes Other: Yes |
| The Critical Mistake | Neglecting the Core Metric: The founder failed to identify and track a single "North Star" metric (like # of games played per day). Instead, they focused on releasing two new features every week, which led to a bloated, unstable product that didn't improve the actual core activity. |
| Key Lessons |
|
Deep Dive
In his interview with Failory, Gabriel Benmergui shared a creative solution to an early financial hurdle. When they launched their initial campaign on an Argentinian platform, credit card companies blocked international donations. Overnight, the founders built their own crowdfunding system on their website using PayPal. They named the funding section "Candy" with pink icons—a humorous touch that resonated with the community and helped them successfully raise their initial $20,000. By using crowdfunding, the early users became the project's biggest "evangelists." This community involvement solved the "chicken-and-egg" problem of a game server, as these donors were highly motivated to invite others and keep the server populated. Kaya.gs is a case study in "Over-Innovation and Technical Hubris." It serves as a reminder for your project that technical passion must be balanced with operational discipline. After the shutdown, Gabriel open-sourced the code and moved into senior engineering roles at YC-backed startups, applying his lessons on "building for impact" to the medical tech space.
Key Lessons
Build for Impact, Not Flash: Innovative features are useless if the core service is unreliable. Make the "boring" parts iron-proof first.
Execution Over Vision: Selling a vision is great for fundraising, but surviving requires technical execution. If you lack expertise, find the right people to help before the technical debt becomes fatal.
Data-Driven Decisions: Without a key metric to follow, your team's morale is at the mercy of emotions. Using data provides a clear indicator of whether you are actually moving forward.