Commentary -Five years ago, I joined a tiny startup that was promising to democratize integration (Talend). Coming from another data integration vendor – a proprietary one – I liked that I would be able to leverage my market expertise. But, like most people at the time, I knew nothing about open source. Of course, I had heard about Linux, I had a vague idea of what MySQL was, but that was the extent of my open source expertise. The two founders of Talend told me not to worry, that open source was only in its infancy, and that we would be part of building it.
At that time, most open source vendors were trying to replicate what proprietary vendors were doing, or what they had failed to do. The value proposition was simple: vendors would say they were like X, but more open, more extensible, and less expensive. Take a few of the successes of the late 2000s and who they were compared to: MySQL (Oracle), JBoss (WebSphere), Jaspersoft (BusinessObjects), Talend (Informatica), SugarCRM (Siebel).
By and large, these vendors were successful. The first “billion dollar baby” of open source was MySQL, when Sun bought the company for $1 billion. At that time, Techcrunch headline was: “Sun Picks Up MySQL For $1 Billion; Open Source Is A Legitimate Business Model.” And indeed, 2008 marked a turning point for open source: more and more enterprise deployments; acquisitions, like in the “real” corporate world; more and more funding. The 451 Group tracks the history of VC funding in open source – the graph in this post shows that investment in 2008 was at an all-time high, which would only be matched again in 2011.
Fast forward to today. Red Hat just became the second billion dollar open source baby. Of course, they are the first billion-dollar-revenue open source vendor. But I believe the impact is as great as when MySQL was valued a billion dollar. Why? Because it settles the debate: you can build, and grow, a sustainable and scalable commercial open source business.
As open source vendors were becoming mainstream, they also started to lose some of their differentiation. Today, you can’t continue to scale a business by merely being “the open source alternative to X.” People running X are not stupid. They have seen the threat, and reacted to it. And even though they will not admit it publicly, proprietary vendors are keeping close tabs on their open source competitors. We are well beyond the basic, primitive FUD (Fear, Uncertainty Doubt) strategy that was used in the past. Proprietary vendors have learned to compete with open source.
In order to continue to strive, open source vendors must continue to be different. One way open source has been retaining this differentiation is through innovation.
Unlike their proprietary counterparts, open source vendors are younger, more agile companies. They use open, standard-based technology stacks. Innovation is part of the open source DNA.Let’s look at three of the major IT “r-evolutions” of the past few years (I don’t really like the term “revolution”, I think most IT changes are evolutionary in nature, and that this evolution is simply going faster or slower).
- Cloud computing – open source powers the cloud. Most of the cloud infrastructure is based on the open source LAMP (Linux Apache MySQL PHP) stack. It’s not only because open source technology works well in this context, but also because the open source licensing model makes it easier on cloud providers to scale their infrastructure. The main non-open-source cloud is Microsoft’s Azure, but I don’t think they run into licensing conflicts with themselves.
- Big data – even though all vendors are trying to jump on the big data bandwagon, the core big data technology stack, Hadoop, is before all an open source project, managed by the Apache Software Foundation. It’s actually interesting to see how competing Hadoop distribution vendors work on the same project, and build features that also benefit the others. It’s of course not new in the open source world (look at all the ESBs built on the same Apache projects) but might be the first time it gets so strategic.
- Mobile – there is probably less obvious of an open source play in mobile computing. Google’s Android is viewed as open source, despite some controversy in the way Google releases source code. And even though Apple’s iOS is clearly not open source, it shares most of its frameworks and underlying operating system components with OS X, Apple’s desktop software. OS X and iOS are built on top of over 200 open source projects, and their very kernel is based on top of an open source kernel called Darwin.