I knew that if I ever started a company, I would build it on some of the principles that drive open source projects.
While most companies leverage product roadmaps and rigid structure, open source projects operate differently.
Let's look at the Linux kernel, for example. The Linux kernel is among the most fascinating pieces of software ever written. There is no corporate structure of development behind it, yet it has evolved to be one of the most amazing technological advances mankind has created.
The kernel does away with armies of product managers, project managers and office managers. It wastes no time on things that managers love to do. The kernel is very much the evolution of code: the good ideas stick and the bad ideas leave (or elements of them turn into good ideas).
When we started Mailgun, we wanted to minimize the requirement of a centrally-planned structure and used this evolution to our advantage. Here are some open source principles that were applied to our startup.
Kill The Meetings
Open source projects run on mailing lists, IRC channels, Wiki pages and occasional beer parties. This lends itself to a couple of advantages over traditional meetings.
- First, all of the discussions and conversations are naturally documented and searchable. This creates a treasure trove of information that can be used later. This is very different from traditional meetings where people come face-to-face to discuss issues and ideas, but typically leave the room with a lot of undocumented knowledge. Mailing lists and Wiki pages make everyone on the team equally productive regardless of their physical location (more on this later).
- Second, we found that the most interesting conversations often occur after hours. Something happens when people get home, relax, pop a beer and get on the mailing list or IRC to discuss technical things. Thinking about these difficult problems throughout the day often results in inspiring ideas and novel solutions. Having this information recorded for future documentation or blog posts makes it that much better.
No HQ Required; Hire Across The Globe
There is not a single building that has "Offices of the Linux Kernel" plastered on the outside. Just as open source projects know no boundaries, you can employ people from all over the globe to help you with your startup. You don't have to have everyone located in the same place, which can be liberating.
Instead of having to hire from a pool of people in your city, the Internet has connected the entire world and enables teams to hack together from thousands of miles away. Moreover, the quality and caliber of talent you can attract increases significantly if you're not encumbered by a single location.
At Mailgun, we never had a proper HQ.
Contributions Trump Conjectures
Each startup is like a young organism at the start of an evolution. The business is about to go through the grinder in a full survival of the fittest mode against other startups. At this point in time the business is in an extremely fragile state. If your startup consists of two people and it is time to hire a third one, you are about to increase the size of your company by 50 percent. Making the wrong hire could be devastating. Making this decision based on a single interview is like marrying someone after a single blind date!
Instead, at Mailgun we approached new hires in the same way people contribute to open source projects. As developers and engineers were interested in working with us, we did not rely on interviews. Instead, we always maintained an open invitation for anyone to contribute on a part time basis, usually remotely. Folks who "clicked" with us have naturally transitioned to full-time roles, others remain contractors to this day.
Leveraging some of the key principles of open source projects helped us take Mailgun to where it is today. We built a great company on these principles and are confident that following some of tenets of open source can help you build a great startup yourself.