Latest / The Open Source Business with Fexingo: Commercial Strategy for Free Software Companies / How Open Source Funds Its Own Infrastructure
Transcript
- Lucas: We talk a lot about revenue here, but there is a much uglier line item that rarely makes it onto pitch decks: the actual cost of keeping the lights on for open source software. I was looking at some recent data from GitHub's annual survey earlier this week, and it highlighted something stark. The average maintainer of a critical project spends over thirty percent of their personal time just trying to pay for hosting, CI pipelines, and domain renewals. Luna: That number feels surprisingly low given what I see in my inbox from smaller maintainers who are essentially burning cash out of pocket. It is not just about salaries anymore, Lucas, it is about the sheer electricity and bandwidth required to run these modern monoliths. Lucas: Exactly. And when you scale that up to enterprise-grade tools like Kubernetes or Postgres, the burn rate isn't measured in dollars, it is measured in tens of thousands per month just for compute. So the question becomes, how do you build a business model that covers those infrastructural debts without alienating the community that built the product in the first place? Because if you get this wrong, you don't just lose money, you lose trust. Luna: Right, and the irony is that the more successful the project gets, the harder it becomes to fund it because everyone expects the core to remain free forever. Lucas: Which brings us to the two distinct camps we are seeing emerge in late twenty twenty six. On one side, you have the pure foundation model, where entities like the PostgreSQL Global Development Group rely on donations and corporate sponsorships to cover their operational costs. On the other side, you have the dual-licensing giants like MongoDB and Elastic, who shifted their licenses to restrict cloud providers from reselling their software. But there is a third, quieter approach that is gaining traction among mid-tier tools, and it involves selling the infrastructure, not the code. Luna: I want to pause on that cloud provider restriction shift. Wasn't that controversial when it started? A lot of people felt that violated the spirit of open source. Lucas: It was incredibly controversial, and many communities fractured over it. Companies like Elastic and Redis had to rebrand themselves as open-core or even source-available, moving away from traditional MIT or Apache licenses. The argument was simple: AWS and Azure were taking free software, wrapping it in a managed service, and selling it back to customers at a massive markup, while the creators saw zero revenue. So they changed the rules to protect their bottom line. But let me ask you, does that actually solve the infrastructure problem, or does it just create a new kind of vendor lock-in fear? Luna: It solves the cash flow problem for the company, sure, but it creates a huge adoption barrier for enterprises that are terrified of being locked into a proprietary fork. They hesitate to adopt the base version knowing the license might change again. Lucas: Precisely. And that hesitation is exactly why I think the managed service tier is becoming the more stable middle ground. Look at companies like Datadog or New Relic in their early days, or now, tools like Grafana Labs. They keep the core engine fully open and permissive, often under Apache two point zero, which allows anyone to use it. But they make money by offering a hosted version that handles the heavy lifting—scaling, backups, security patches, and yes, the CI/CD pipelines that eat up so much volunteer time. They are selling convenience and reliability, not restricting access to the code itself. Luna: So you are essentially arguing that the moat is no longer the code, but the operational complexity of running that code at scale. Lucas: That is the thesis. Code is commoditized. Operational excellence is scarce. When I look at the balance sheets of successful open source adjacent companies today, the revenue comes from the enterprise support contracts and the cloud marketplaces. The community edition is just the top of the funnel. It is a marketing expense, really. But here is the catch: if you push too hard on the paid tier, you risk cannibalizing your own adoption. You need a clear demarcation line between what is free and what is paid, and that line has to be defensible. Luna: And that demarcation line is usually where the biggest legal and technical debates happen. Who decides what features go behind the paywall? Lucas: Usually the founders, initially. But as the project scales, it becomes a governance issue. If you put authentication and logging behind a paywall, developers will grumble. If you put basic error handling behind a paywall, they will leave. The trick is to identify features that are valuable to enterprises but irrelevant to individual hobbyists. Things like SSO integration, audit logs, multi-region failover, and automated compliance reporting. These are expensive to build and maintain, so they justify the premium price. But they are not interesting enough to individual contributors to motivate them to build their own solutions. Luna: That sounds like a very rational way to segment the market, but it requires deep understanding of both the developer psyche and the CTO's pain points. Lucas: It does. And this is where the talent war we discussed last month comes back into play. You cannot hire the engineers to build those enterprise features unless you can pay them competitive salaries. And you can only pay them competitive salaries if you have captured enough value from the enterprise tier. So it is a self-reinforcing cycle. Better engineering leads to better enterprise features, which leads to higher revenue, which allows you to attract better engineering. But it all starts with that initial decision to treat infrastructure as a product. Luna: Speaking of products and supporting the show, if these conversations are useful for what you're building or running, consider supporting us directly. A couple of dollars a month is genuinely what keeps these going — buy me a coffee dot com slash fexingo, that's F. E. X. I. N. G. O. Lucas: Thanks for the support, Luna. It really helps us dig deeper into these structural issues rather than just skimming the surface. Now, back to the infrastructure model, what happens when the cloud providers themselves start competing on price? Does that squeeze the margins for managed services? Luna: It definitely puts pressure on the low-end tier. If AWS offers a cheaper database instance, why would anyone pay for a managed service wrapper? Lucas: Right, and this is where the smart players are pivoting toward abstraction layers. Instead of just wrapping a database, they are wrapping the entire developer experience. Think about platforms like Vercel or Supabase. They are not just selling hosting; they are selling a workflow. The code lives in your Git repo, but the deployment, preview environments, edge functions, and database connections are all handled by their platform. The value proposition shifts from 'we host your software' to 'we accelerate your development cycle.' That is much harder for a commodity cloud provider to replicate because it requires deep integration with the developer toolchain. Luna: So the moat becomes the integration depth rather than the raw compute capacity. Lucas: Exactly. Integration depth is sticky. Once a team has configured their CI/CD pipelines, their preview environments, and their database schemas within a specific ecosystem, the switching cost becomes astronomical. It is not just about the monthly bill; it is about the hours lost in migration. This is why companies like HashiCorp, despite their licensing controversy, still command such high valuations. Their Terraform and Vault tools are deeply embedded in the infrastructure of thousands of companies. Switching away from them is not just an economic decision; it is a logistical nightmare. Luna: But doesn't that reliance on deep embedding create vulnerability? What if the company mismanages its direction or faces a PR crisis? Lucas: It absolutely does. And that is why the foundation model remains so important for critical public goods. Look at the Linux Foundation or the Eclipse Foundation. They provide a neutral ground where multiple competitors can contribute to a shared standard without any single entity controlling the direction. This mitigates the risk of vendor lock-in because the standard is owned by the community, not a corporation. However, foundations struggle to raise capital quickly. They are slow, deliberate, and often underfunded relative to the scale of the projects they support. It is a trade-off between agility and stability. Luna: Agility versus stability. It seems like every business model in tech is just a variation on that tension. Lucas: It is. And for open source specifically, the tension is heightened because the community demands transparency and freedom, while the business demands predictability and profit. The most successful companies find a way to align these incentives. For example, they might offer a generous free tier that satisfies the community, while charging for the advanced features that satisfy the enterprise. Or they might operate as a foundation-funded entity while spinning off a separate commercial subsidiary to handle sales and marketing. This separation of concerns allows them to maintain goodwill in the community while pursuing aggressive growth strategies in the market. Luna: Is that separation easy to manage internally? Having two different cultures within the same organization? Lucas: It is notoriously difficult. You have engineers who are motivated by impact and recognition, and salespeople who are motivated by quota and commission. If the communication breaks down, the engineers feel like the company is selling out, and the salespeople feel like the engineers are holding back the product. The best leaders in this space are bilingual—they speak both the language of code and the language of commerce. They can explain to a contributor why a feature needs to be prioritized for revenue, and they can explain to a customer why a certain limitation exists in the free tier. It requires empathy on both sides. Luna: That level of empathy is rare in most tech companies, especially the ones that have grown too fast. Lucas: It is. And that is why I believe the future of open source business belongs to the patient operators. Not the unicorns that burn cash to acquire users, but the compounders who build sustainable revenue streams around essential tools. As we move further into twenty twenty six, with interest rates stabilizing and cloud spending coming under scrutiny, companies will look for ways to reduce their IT overhead. Open source managed services offer a compelling path to do that, provided they can prove they are not just extracting value, but adding it. Luna: Adding value through operational excellence rather than restricting code access. That seems like a healthier long-term strategy. Lucas: I agree. It aligns the interests of the creator with the interests of the user. When you succeed because your users succeed, you build a relationship that lasts longer than any license agreement. The question for founders today is not just how to monetize their code, but how to monetize their competence. Can you charge for the knowledge it takes to run complex systems efficiently? Because that is where the real margin lies. Luna: Monetizing competence over code. That reframes the entire conversation about open source sustainability. Lucas: It does. And it suggests that the next wave of open source successes won't come from the biggest names, but from the most reliable ones. The ones that become invisible because they work so well. Do you think enterprises are ready to pay for that invisibility, or are they still chasing shiny new features? Luna: I think they are tired of chasing. After years of vendor churn and failed migrations, reliability is becoming the most premium feature available. Lucas: Reliability as a premium feature. That is a powerful signal. It means the market is maturing. We are moving past the hype cycle into the utility phase. And in the utility phase, the businesses that survive are the ones that treat their infrastructure with the same seriousness as their product. Thank you for tuning in to The Open Source Business. Let us know what infrastructure challenges you are facing in your own workflows.