Latest / The Open Source Business with Fexingo: Commercial Strategy for Free Software Companies / How Open Source Firms Win With Support Contracts
Transcript
- Lucas: The biggest misconception in tech right now is that open source means no revenue. It’s a false binary. The reality is that while the code itself is free, the liability around running it in production is not. That shift in value capture is what separates hobby projects from billion-dollar public companies. Luna: Exactly. And it’s not just about fixing bugs when they break. It’s about the guarantee that things won’t break in the first place. Who do you think was the first to really nail that psychological trigger for buyers? Lucas: You have to go back to Red Hat in the early two thousands. They didn’t sell the kernel. They sold the service contract. A company would download Linux for free, but if their core banking system went down at three AM, they needed someone to call. Red Hat charged for that phone number. It was essentially an insurance policy for IT directors who were terrified of being blamed for outages. Luna: Right, because without that safety net, adopting open source felt like career suicide for many operations managers. Did that model scale cleanly as everything moved to the cloud? Lucas: It evolved, but the core mechanic stayed the same. When IBM acquired Red Hat for thirty-four billion dollars in twenty nineteen, the market wasn’t paying for the software. They were paying for the ecosystem and the support infrastructure. Even today, in September twenty twenty-six, the largest open source vendors still derive majority of their recurring revenue from subscriptions tied to technical support and access to updates. Luna: So if I’m running a mid-size fintech startup, I’m not going to build my own security patching team. I’m going to pay a premium for that assurance. Is that where the real margin lives? Lucas: That’s where the sticky revenue lives. Margins can be tricky because high-touch support is labor-intensive. But the key is productizing the support so it doesn’t scale linearly with headcount. You want engineers solving one problem once, not ten problems every day. That’s why documentation and automated diagnostics are actually part of the support product. Luna: I’ve seen companies fail here because they treat support as a cost center instead of a product. If your support team is just taking tickets, you’re losing money. If they’re feeding data back into the roadmap, you’re building a moat. Lucas: Spot on. The best support contracts include proactive monitoring. It’s not just reactive fire-fighting. It’s telling the client before they know there’s a problem. That shifts the relationship from vendor to partner. And partners don’t churn easily. Luna: But how do you price that? Because every enterprise has different complexity levels. One size fits all pricing usually breaks under pressure. Lucas: Most mature firms tier it by response time and coverage scope. Basic might mean email support during business hours. Premium means four-hour response times with dedicated account engineers. The jump in price between tiers is usually disproportionate to the actual cost difference. That’s where the profit margin expands. You’re selling speed and certainty, not just man-hours. Luna: So you’re essentially charging a convenience fee for risk mitigation. That makes sense for CFOs who hate surprise downtime costs. Lucas: Precisely. And let’s talk about the other side of the ledger. How does this affect hiring? Because the talent required for enterprise support is different from the talent building the core features. Luna: Good point. Support engineers need empathy and communication skills. Feature engineers need raw technical depth. Can the same person do both well? Lucas: Rarely. The most successful open source companies keep these functions separate but tightly integrated. Your support team feeds the bug reports. Your feature team builds the fixes. But they report to different org structures. Mixing them too early leads to burnout. Engineers get stuck in ticket queues instead of shipping new code. Luna: That’s a common trap for growing startups. You hire a great engineer to handle support, and suddenly you lose your lead developer. You’ve lost twice. Lucas: Twice. And you’ve degraded your product velocity. So the organizational design has to protect the builders. Support becomes a specialized function, almost like a sales engineering role but focused on retention and stability rather than acquisition. Luna: Which brings me back to the listener question I saw earlier about community-driven projects trying to go commercial. How do they bridge that gap without alienating their base? Lucas: They have to be transparent. If you start charging for support on a project that was purely community-led, you risk backlash. The trick is to offer the community edition freely while packaging the enterprise features—like advanced security scanning or compliance reporting—into the paid tier. You’re not locking people out. You’re offering tools that big companies legally require. Luna: So you’re creating a natural upgrade path based on regulatory needs rather than arbitrary feature gating. That feels much more honest to developers. Lucas: It is. And it aligns incentives. The large customers get what they need to pass audits. The small users get the core tool for free. Everyone wins. The friction usually comes when companies try to force open-source licensing on everyone. That never works long-term. Luna: Let’s take a quick breath here. If today was actually useful to you, the way these stay ad-free is listener support — buy me a coffee dot com slash fexingo. Lucas: Yeah, we appreciate it. It helps us keep digging into these structural questions instead of chasing clicks. Luna: Thanks for listening. Now, back to the hard part: measuring success. Lucas: Right. So how do you know if your support model is working? It’s not just about resolving tickets fast. It’s about preventing them. The metric that matters most is actually mean time to resolution, but the leading indicator is ticket volume per user. Luna: If ticket volume goes down while user count goes up, your product is stabilizing. That’s the holy grail. Does that ever happen? Lucas: It does, but it takes years. Look at Kubernetes. In the early days, support tickets were endless. Every deployment failed differently. Now, after five years of maturation, the baseline experience is stable enough that most enterprises only call for edge cases. The platform did the heavy lifting. Luna: So the goal isn’t to provide infinite support. It’s to build a product so robust that support becomes unnecessary. That’s a brutal timeline though. Lucas: It is. And that’s why cash flow management is critical. You have to fund the development cycle through support revenue until the product matures. If you run out of cash before stability hits, you die. It’s a classic valley of death in open source ventures. Luna: Valley of death. That’s a stark way to put it. Are there examples of companies that got stuck there? Lucas: Several. There was a prominent database project that tried to scale support before their replication features were reliable. Customers kept calling because of data sync issues. The support team became overwhelmed, quality dropped, and churn spiked. They had to restructure completely and pause new features to fix the core reliability. It cost them eighteen months of growth. Luna: Ouch. So patience is actually a competitive advantage. Most founders want to move too fast. Lucas: Exactly. In open source, speed is cheap. Reliability is expensive. If you can afford the reliability, you win the enterprise market. If you’re just fast, you stay a niche tool. Luna: What about the global angle? Support has to work across time zones now. Does that change the cost structure? Lucas: Significantly. Follow-the-sun support models are standard for global enterprises. You need teams in Asia, Europe, and America covering twenty-four hours. This increases operational complexity but allows you to charge higher premiums. Global clients expect global coverage. Luna: So you’re trading higher fixed costs for higher average contract values. That seems like a smart trade-off for scaling. Lucas: It is, provided you automate the handoffs between time zones. If a ticket gets dropped because a shift ended, you lose trust instantly. Automation bridges the human gap. Luna: Automation again. It’s the silent hero of these support models. Without it, you’re just managing chaos. Lucas: Always. And finally, let’s touch on AI. Everyone is asking if AI will replace support engineers. My view is no. AI handles the simple queries. Humans handle the complex, nuanced crises. The role shifts from answerer to architect. Luna: That makes sense. AI tells you what the error code means. A human tells you why it happened in your specific environment. Context is king. Lucas: Context is everything. So as we look ahead to Q four of twenty twenty-six, the winners will be the ones who blend AI efficiency with human expertise seamlessly. Not one or the other. Both. Luna: So the future of open source business isn’t just about code. It’s about the quality of the relationship you maintain with the people using it. Lucas: Exactly. Code is commodity. Trust is scarce. Build the trust, and the revenue follows.