{"type":"rich","version":"1.0","provider_name":"Transistor","provider_url":"https://transistor.fm","author_name":"The Business of Open Source","title":"From Closed to Open: Shortening Sales Cycles with Stephen Goldberg","html":"","width":"100%","height":180,"duration":2026,"description":"This week on The Business of Open Source, I chatted with Stephen Goldberg, co-founder of Harper, about the process of taking Harper from a closed-source software company to a open source company. I already talked to Ethan Arrowood about the transition to open source, but Stephen was able to give a founders perspective about why Harper wanted to move to open source, why it wasn’t launched as an open source company from the very beginning and why they choose the moment in question to go open source. \nHe says that 80 to 90% of the effort around taking a big piece of Harper open source was around communication, both external and internal. It’s not that difficult to change an license or to push something onto GitHub, but a successful open source launch is about so much more than that. \nHere’s what we talked about: \nThe board pushed back a lot, mostly because they felt like going open source was going to complicate things for the company and they were worried that the risks outweighed the benefitsStephen was convinced that even as a closed-source company, their customers weren’t paying for their code, but rather for their services. Harper also sells to large enterprises, and he was convinced this type of customer doesn’t want to cowboy their tech stack. There has been a massive positive impact on the company’s sales cycles. The sales cycles are shorter and require less effort from Harper’s team; it’s also psychologically easier when someone asks him to prove that the product does what he promises — he just has to ask tell them to verify for themselves. They haven’t gotten many outside contributions and that was never the goal, but they do get bug reports. This has significantly reduced their costs of support, because the bug reports are often very detailed and identify where in the code there is a problem. Not only that: customers are able to figure out when the “bug” is actually a problem on their side, and that reduces support volume considerably. AI coding agents...","thumbnail_url":"https://img.transistorcdn.com/8IkNaFmPeP0cacpN72wvWvmuTJkRWMTzK4anhRhTwJg/rs:fill:0:0:1/w:400/h:400/q:60/mb:500000/aHR0cHM6Ly9pbWct/dXBsb2FkLXByb2R1/Y3Rpb24udHJhbnNp/c3Rvci5mbS9zaG93/LzEwNTI5LzE2NTY1/MTIwMDAtYXJ0d29y/ay5qcGc.webp","thumbnail_width":300,"thumbnail_height":300}