2634949567_1343a57243I was attending TechCrunch Berlin last week. Of all the startups pitching their ideas, doctr.com was was the only one stressing an open source strategy. I have not given doctr.com a closer look yet but from their pitch I take it that the team is using quite a lot of open source frameworks and standards and is pretty keen to contribute back to the community. It also sounded as if some of the key aspects of their software were open sourced. Having just finished Jeff Jarvis’ book “What would Google do”, I wanted to comment a bit on this strategy of opening up all or parts of of a software system to a community.

While Jeff’s views might be purposefully radical and controversial, I also found them quite compelling–something I am not able to say about many books covering Google’s history, strategy, and implication on our acting and thinking. However, opening up the source code before you get serious traction is something that bears great risk and which I do not recommend doing.

Admittedly, there are many great examples when companies opened up their system by releasing their code to the community. MySQL did so from the start in 1994. Mozilla Firefox has always been an open source project–even before its first release in 2004. Even Twitter released parts of its code in January 2008.

But Twitter is attracting a different crowd and does not face the threat of a powerful incumbent. MySQL and Firefox faced strong competitors that already had a strong footprint and were becoming more and more complex. Both of them needed the open source community to drive a very technological product to compete against the goliaths. Their success is partly due to the fast adoption of open source developers who wanted to change the world they were confronted with every day.

Twitter started its open source strategy differently. Instead of releasing all of its code to the public, only Starling, Twitter’s queue server made it out into the wild. This little bit of code might be crucial to Twitter but it is not the whole system. And with users like Ashton Kutcher, Barack Obama or Britney Spears, Twitter is not driven by technical users like software developers. As people got more and more attached to Twitter, Twitter might have become more popular among software developers but it didnät need them to gain serious traction. Let’s not forget: it would probably be easy for many software developers to build a service like Twitter. But any Twitter copy cats failed because they were either too technical or just didn’t attract the right users.

I agree that the benefits from open source are manyfold and in most cases supportive to a better system and can create serious support from a very elite group of software developers. Like it or not, many times software developers are the early adopters of software technology and thus can be of great help gaining serious penetration–in a market they directly influence. As of today, doctr.com’s website on open source remains empty but I will follow its progress with great curiosity. But my advise is: decide wisely which parts of the system you want to open up to the public and know if they really care. Do something to get the right people on board rather than attracting a group that does not influence your target users.