Interview: Mårten Mickos of Eucalyptus Systems and MySQL

By any measure, Marten Mickos is one of the most successful executives in Open Source. As CEO, he led MySQL AB to an acquisition by Sun Microsystems in 2008. Today, he is CEO of Eucalyptus Systems, developers of the Eucalyptus private cloud platform.

OU: You joined MySQL AB in 2001. Ten years later, how is the world different for Open Source startups? What tools or opportunities are available that weren’t there when you started with MySQL?

MM: When I joined, open source was a new thing in the software industry. Customers had to be very brave to use it. The free and open source software movement was already over a decade old, but it was only around the millennium shift that customers started looking at open source products.

Today open source is fully accepted. And it is more than that: it is a required piece of anybody’s software strategy. Even Microsoft and Oracle today produce open source software.

As for opportunities, today’s software world is much much larger, and more varied than before. You can do software for mobile devices, for cars, for industrial automation, for medical appliances, games, and so on.

There are virtually limitless opportunities, and all those opportunities are open for open source software. Just look at the cloud computing space where products like Linux, KVM, MySQL, Hadoop, Puppet, Eucalyptus and others are leading the innovation. You could say that the open source movement invented scale-out – and today everything in the cloud is about scale-out.

OU: Some Open Source software projects (like Apache, the Linux kernel,  Firefox) have thrived under non-profit stewardship, and others (like WordPress, JBoss or MySQL) have done well as the products of  dedicated companies. Is there a qualitative difference between the software that works well in each environment? Or is it just historical accident?

MM: I have spent endless hours thinking of this. I always believed, and continue to believe, that open source is not a business model. It is a production and distribution model. So I happen to believe that the profit or non-profit character of an open source product is irrelevant when you analyze root causes and paths to success.

What matters is the market opportunity and the governance model. The most successful open source projects address a big opportunity in the industry, and they have a governance model with some sort of benevolent director. The leadership mandate may not be fully formalized, but in practice it largely is. Take Linux with Linus Torvalds. Linus could of course be voted down, but in reality it doesn’t happen. The development of Firefox is lead by Mozilla. The development of MySQL was led by MySQL AB and then Sun and then Oracle. Matt Mullenweg is the supreme steward of WordPress. Android is lead by Google. Projects with a less structured or less singular stewardship have traditionally not done as well.

But there is shining exception here: Apache. That’s a project that has reached enormous success without singling out any specific individual or commanding organization as the chief.

So, coming back to your question, I believe that the business side of open source is not the relevant aspect. The governance model is.

OU: In your mind, how much of the success of an Open Source business is due to the product itself? To a healthy community? To the business  infrastructure that supports it?

MM: Great question. Success is in my experience always a result of many favorable developments. Like in a 5 star hotel, all details must be in great shape. There must be a need in the market. You must have a great product. You must have an architecture of participation (and there are different models you can choose from). You must have strong and passionate people who will drive the project relentlessly, even when facing adversity. And you probably need some luck.

OU: The $1Bn acquisition of MySQL AB by Sun Microsystems during your  tenure as CEO is one of the biggest achievements in the history of  Open Source business. What do you think were the main factors that brought it about? Could a similar acquisition happen again?

MM: One day we will look back and smile at the fact that we thought that one billion was a big number for open source. The software market is just becoming bigger, and open source can play a bigger role than before.

As for MySQL, it had unique value in being the singular M in LAMP. Nobody else could claim that position. MySQL also had the guts to go against the old beliefs. We scaled out when others scaled up. We were open when others were closed. We targeted web customers when others tried to go after the enterprise. We favored simplicity when others strove for complexity.

All those bets turned out to be right. That’s why the installed base grew so rapidly. And the installed base lead indirectly to the valuation of the company.

I am sure we’ll see other MySQL-like successes in the future. The interesting thing is that we don’t know when and where.

OU: With MySQL behind you, why did you decide to go back in the ring with Eucalyptus Systems? What do you like about the company and its product? Why do you think there’s a good opportunity there?

MM: I love building teams that build businesses, so I just had to go back into the industry. Cloud computing might be the biggest shift in IT I will see in my professional life. Eucalyptus has an absolutely unique and outstanding founding team, and the product is brilliant. It was an easy decision to apply for the job. Luckily for me, they hired me.

OU: When you’re with a new company, how much do you experiment with new products or business models?

As much as needed! The winner is not the strongest or the most intelligent, but the one most responsive to change. That’s a quote from Darwin.

OU: Cloud platforms seem to be the biggest growth area for Open Source startups. Why is Open Source attractive to cloud users? Why are Open Source cloud startups attractive to investors?

MM: Open source is simply a superior way to produce software. You get high quality faster with open source. And a shift is happening: whereas open source traditionally very much was a disruptor of the old, now it is the innovator of the new.

OU: How important is the Open Source brand to Open Source companies?When do you think companies should use the term “Open Source” in their marketing materials, on their Web site, or when talking to customers?

MM: Open source is no longer a branding statement. Ten years ago it was important for MySQL to say “open source”. It was a way of differentiation, and it allowed us to benefit from a huge shift in mindset in the world. Today everyone accepts open source. For that reason, it’s not part of the branding as much as it used to. It just is part of nearly every piece of software.

Open source has an important role in the discussion about vendor lock-in. Open source removes it. But in addition to the source being open, you must also keep your APIs open (i.e. well defined, stable, and openly accessible) and your customer data open (i.e. allowing customers to withdraw their data at will). In this lock-in avoidance discussion, open source continues to be an important differentiator and can be used in branding.

OU: How would you advise an entrepreneur working on a new product who was considering making the software Open Source? What factors should they consider in their decision?

You need to love open source and you need to be ready to live with the transparency – the reality of users complaining about your code, and making changes to it! You also need to have a plan or business model that calls for broad distribution and use of your software. If you have a very small target group, open sourcing the software may not bring tangible benefits.

OU: How would you advise the leaders of an existing Open Source project who were thinking about forming a company? What factors should they consider in their decision? What about seeking investment capital?

You should form a company if it is a business (as opposed to a product) that you want to build. Once you have a company, you must have business leadership for it. Perhaps the leader is you, or perhaps you need to hire someone. Those are big decisions.

Investment capital is suitable for businesses that grow very fast and have a high potential. Many businesses, however, do not grow fast or will never be large. They are great businesses, but they don’t lend themselves to VC funding.

OU: Would you work on a company or product that wasn’t Open Source? Why or why not?
MM: Great question. I don’t know. On the one hand, I am not religious. On the other hand, I am a strong believer in transparency, openness, participation and huge market opportunities. If you have a software company that stands for those values, then it probably also will opensource its code.