Joe Duffy, Founder & CEO, Pulumi
1:17: Background on Joe & Pulumi - the company was started to give infra engineers tools to help them innovate faster.
3:05: Why OS as a core part of Pulumi’s strategy - getting community buy-in was important as it created trust and authenticity. Joe also saw the power of open source when he was at Microsoft and they started exploring an OS strategy.
5:29: Pulumi launched as a company and OS project at the same time. The benefit of this was they could be thoughtful about the business model from the get-go instead of it being an afterthought.
9:53: The SaaS attach rate was insanely high - 2/3 adopted the SaaS. It was the default experience that developers opted out of if they wanted to. The business model was close to what AWS offers - a full managed service.
12:13: Finding product-market fit with the OS project and business at the same time was tricky. It took over a year to release the OS project, however they had their first paying customer before they open-sourced anything.
14:28: They decided to charge for enterprise features only (identity, policy enforcement, webhooks, team management, etc.).
20:25: When measuring the health of the open-source, they focused on engagement over growth metrics. They wanted end users to be really successful. Joe continues to be baffled about investor conversations that center on stars - it’s one light measure of momentum but there are much stronger signals around health from Stack Overflow, Hacker News, Reddit, etc.
25:20: Joe took on the role of the developer advocate himself early on. In fact, he did parts of every function before hiring for them to make sure he understood all parts of the business.
26:53: In raising money for an OS company, focusing on typical SaaS metrics as well as OS metrics is important. Both need to be worked on in tandem.
28:55: Early on, the focus was on building the OS community over the SaaS product. The ratio was 10:1 for people working on the OS vs. SaaS.
31:10: Advice Joe has for OS founders - don’t sweat monetization too soon as once it’s introduced it averts focus away from top of funnel OS growth.
33:05: Joe wishes investors understood developers more. They’re a tough group to make happy and the tech changes rapidly. It’s not just about momentum metrics but solving big problems for the community (TensorFlow is a good example).
35:08: Joe’s final piece of advice is focus on making developers happy and starting in a niche vs. making a solution too broad at the start.
Full episode here