Investing in Open Source Startups with Robby (Amanda Robson)
Robby Robson has invested in open-source infrastructure for eight years and expected to learn a repeatable formula for it. Instead she found the opposite: an open-source project that raises venture money too early is stuck on a growth timeline that's fundamentally at odds with how open-source communities actually build trust.
Robby (Amanda) Robson co-hosts the Open Source Startup Podcast with Tim Chen and has spent eight years investing in developer tools and infrastructure at the earliest stages. The podcast started almost by accident — a blog post interviewing the founders of HashiCorp, LaunchDarkly, and LightStep about their early go-to-market thinking turned into realizing the recorded conversations themselves were the most valuable part.
We get into why she no longer believes there's one successful formula for building an open-source company, how to tell whether an open-source project with impressive GitHub stars actually has venture-scale potential or is mostly hobbyists who will never pay, and why she treats a founder's process for gathering community feedback as a bigger signal than the product itself.
The Q&A below is experimental — extracted from the episode transcript with AI. If you spot a mistake, let us know at jack@scalingdevtools.com.
After 150 episodes interviewing open-source founders, what's the biggest thing you've changed your mind about?
I expected to learn something close to a formula — here's how you successfully build an open-source company. It's turned out to be nothing like that. There are so many different flavors of open-source companies and business models, and they shift depending on the community you're serving. One specific thing I've completely reversed on: early on I assumed the most successful path was always a project's original creator starting a company around it. But there are now plenty of examples, like OpenTelemetry, where multiple companies form around a new open standard with no tie to who originally created it — they've just seen it rise and built on it as a standard other people are adopting.
Why is rushing an open-source project to a venture-backed business model so risky?
Once you put something out in open source, there's an unspoken rule that it's going to stay there and stay free. During 2020 and 2021, when there was a lot of capital available, investors pushed project creators into companies and gave them money very early, which meant those companies had to monetize fast to justify their valuations — but a lot of these projects take a long time to really resonate with a community. Forcing a business model, or trying to grow a community in a way that isn't organic, can genuinely hurt the opportunity. Sentry is a good example of the opposite: David Cramer didn't try to turn it into a company for years, nurturing the project and the community first. It's a fragile way to build a company, and the open-source model can be fundamentally at odds with venture timelines, where each round is roughly expected to 3x your valuation by the next one.
What's the first question you ask a founder who says they want to build an open-source startup?
Why. Founders without much open-source experience often treat it like a go-to-market tactic — basically freemium — but it's not the same thing at all. An open-source community is essentially a standalone product, and depending on how it's structured, it can even end up competing with what you're building. There's a whole community that might be adding functionality, contributing to documentation, and answering each other's questions somewhere like Slack or Discord — it's much bigger and more sensitive than a distribution channel. People tend to think hard about the upside — free distribution, community contributions — and much less about the downside of maintaining and continuing to build for that community while also trying to build a business, especially with venture money and a clock already running.
How do you tell whether an open-source project with big adoption numbers actually has venture-scale potential?
I saw one open-source project — essentially an open-source version of QuickBooks — that was really widely adopted and genuinely useful, more customizable than QuickBooks itself. But when you looked at who was actually using it, it was mostly hobbyists who were never going to pay for anything. Venture-scale, to me, means a real path to something like $100 million ARR in five to ten years, and a project can look like it's taking off — thousands of GitHub stars overnight — while the underlying user base simply isn't one that will ever have budget. We're seeing a lot of this right now in the 'experimental AI economy': usage numbers look incredible, but a lot of it is hobbyists or other AI startups whose own ability to pay doesn't match the funding rounds getting thrown at them. A smaller, less flashy project whose users come from real companies with actual budget in a category that already spends money can have far more potential than one with much bigger raw adoption.
Can you give a concrete example of that distinction working in practice?
Mobile Dev, which built the open-source testing framework Maestro, is a good one — I invested before they were even an open-source company. The founder, Leyland, had been a senior mobile engineer at Uber and understood the space deeply; the mobile testing market was large and full of manual work that engineers didn't want to do. About a year in, he saw the potential for Maestro to become a new standard for mobile testing, and for a quarter that year it was the fastest-growing open-source project, period — but with real users at real companies, not just hobbyists, who are now converting into customers of the paid product built on top of it. That kind of mid-shift in strategy is common at the earliest stages, and it's fine, as long as the founder has a real process for iterating with their community rather than just declaring a pivot from a vacuum.
How do you assess whether a founder actually has that kind of iterative, community-informed process?
A lot of it comes from spending time with founders and understanding how they gather user feedback, and honestly, how coachable they are in conversations with me is a good tell for how well they take input and translate it into output. Every founder I back is a genuine product visionary, but none of them operate in a vacuum — they take in what their community wants and shift positioning, go-to-market, or parts of the product accordingly. One of the biggest red flags for me at pre-seed or seed is a founder who says something like 'this is just the right way to build it because I know it's right, I don't really need feedback.' That can occasionally work, but it's not sustainable — you end up in your own echo chamber without visibility into what's actually happening with your users.
You've mentioned a security company, Drata, as an example of strong upfront research even outside open source. What stood out there?
How much prework they'd done before writing a single line of code. When I first met the team, all they had was a Figma design of what they wanted to build — at one point in the meeting I asked what happens when you click a specific element, and they told me it was literally just a mockup, nothing existed yet. But they'd done sharp, extensive interviews with customers and prospects to figure out exactly what a compliance solution needed to do, which led to a genuinely interesting product insight: instead of automating a once-a-year audit, customers could know at any point during the year whether they were in or out of compliance, plus an auditor portal that removed the need to email auditors separately. That kind of deep, customer-informed insight, translated directly into product decisions, is exactly what I look for regardless of whether the company ends up being open source.
Why do you personally believe in open source as a model, beyond it just being an investment thesis?
It started more from curiosity than a deliberate bet — my co-host Tim came from a founder and engineer background deeply steeped in open source, and I'm more of a self-taught builder who learned a huge amount from open-source communities and loved the community aspect of it. Meeting founders like Armon at HashiCorp helped me understand the real power of building in a community-centric way, especially when you're trying to establish a new standard — you get participation you simply can't get with something closed-source. I think that's especially relevant right now with foundation models: keeping foundational technology open creates a different level of trust, transparency, and community involvement than a closed alternative can. The more I've seen that power, the more skeptical I've also become of founders who assume 'we'll just open source it and that'll make it succeed' — there's a lot more that goes into it than that.