Teambox’s post about why you should build an open source startup is doing the rounds right now. It’s a great piece, but I thought I’d reply with a few reasons why you might want to think long and hard about it:
- Free (as in speech) software has an accounting value of $0.
- Professional services typically don’t substantially capture value in a company. Don’t sell one-to-one services; sell something you can replicate and grow.
- Open source ecosystem companies – i.e., firms that use your software but don’t contribute to its development – have an obvious competitive advantage over you. OpenSource.com regularly declares that open source business works by pointing to the success of this sort of company.
- Providing managed, pay for the best unlimited hosting is a common strategy, but, again, this can be undercut very easily by non-developing competitors.
- Even though it’s open source, your team is probably going to be doing most of the development.
None of which is to say that open source can’t be a viable business model – but hold something back. Unless you’re building a lifestyle business (something that’ll keep you living well but doesn’t necessarily need to grow), you need to capture value. At its best, open source can be great for marketing, and great for the product. But don’t think it’s enough to be open and awesome; your business, ultimately, has to be worth something to survive.
One response to “Why you should not build an open source startup”
Open-source is not for everyone, and it doesn’t always make sense to do it. We’ve seen it’s a great idea on wide projects where many are facing similar needs: server software, backend applications, databases. But you are right that if your expertise in your system is replaceable but others taking it and just reselling without adding to the development, your business model might be in risk.
At Teambox we haven’t had any problems. Sure, others have tried to replicate our subscription model – as WordPress has seen over the years – but there is no value on going with others over us, because there is much more to the product than it’s features. Support to final users, documentation and the network effect are good enough reasons to choose our service over any upcoming competitor’s one. And whoever wants to run their installation is free to do that for privacy reasons, but ultimately it will end up being more costly than going fully managed under our servers.
Just my 2 cents 🙂