Monday, May 8, 2006
Interview with Winston Damarillo, Simula Labs
My interview this morning is with Winston Damarillo, CEO and Founder of Simula Labs (www.simulalabs.com), a Marina Del Rey-based firm that is developing companies based on open source software. Winston is the founder of Gluecode, which was purchased by IBM last year, and a former venture capitalist at Intel Capital. I spoke with Winston Friday, exactly a year to the day after the Gluecode's acquisition.
Ben Kuo: Tell me a bit about Simula and the idea behind the company?
Winston Damarillo: Simula is an execution of a goal I have had for a long, long time now. A lot of my early career was spent inside Intel, and the last few years inside Intel Capital, and that gave me a great opportunity to have an overview of the tech industry and the parts I'd like to participate in. Software was particularly interesting to me, simply because it was one of the industries I thought was suitable for what I call a business architecture change—and, as an investor at Intel investing in software companies – that the only way to get the software industry to change, was to start with something very disruptive. And, I saw open source as a way to do that. So, in March of 2000 I decided my career at Intel was over, and I wanted to do this as my own venture. Specifically, I wanted to build and open source software company, and, to build it not to specifically focus on just one product, but on how to introduce a new way of doing business of doing software, in a manner that allows for more participation, more entrepreneurs, and more likelihood of success against established vendors, like Oracle, BEA, and IBM. Open source was the way to do that.
Gluecode was the first company coming out of that idea. I actually self funded it for awhile. September of 2001 was the genesis of Gluecode, the open source company, which we then bootstrapped—and, as you can imagine, in year 2001, if you were talking about open source people would look at you funny—but that actually ended up a great time to do it. The reason why was the bubble burst, and people were looking the amount of technology they needed to deliver and the new reality of how much budget they had left—and open source was just starting to become viable. Gluecode capitalized on that. It specifically focused on a couple of good opportunities. We had the opportunity of building an open source software startup around the emerging, and reliable, and growing community like the Apache software foundation. We also had the chance to look at and understand the different kinds of licensing in open source, such as Apache software license vs. GPL; the buying characteristics of people when they buy open source software—are they buying services, subscriptions, or products?—and if you should do a hybrid product, or not a hybrid product; and more importantly, what markets to attack. Specifically, which billion dollar market can you disrupt by having an open source play. The particular market we picked was the J2EE application server market. We decided to start with Apache, continue to propagate community development, and then commercialize it. Open source has always been and will continue to be the best place to build software and to get the best products—but I've always labeled them as the best written, eighty percent complete software. There's always room for improvement, in the aspects of making it easier to use, making it more reliable to use, and also in the aspect of making it work with other stuff. That's what Gluecode really focused on—hence the name Gluecode.
BK: What what the revenue model for that?
WD: The most established business models work; that is, download the product, test it out, but when you put it into production, you're going to have to count on someone to help you with issues, and you're going to want to pay for things that make things easy to use. For example, with Red Hat you download it because it's easier to use than if you download the Linux kernel directly. So, those vendor value adds can easily be monetized. In Red Hat, they have what they call the Red Hat Network, which is where you get support subscriptions for the use of the Linux technology, but that also comes with bells and whistles such as automatic patch updates, a better installer, and integration with other applications—the Red Hat Bundle. It's very similar to that. While it's similar, it also leverages the fact that was already established. We also had the opportunity to innovate—we were not bound by the GPL license, and anything we built on top did not have to go to the community, so we were able to license them as add-ons—still subscriptions, but add-ons. We also had the opportunity to enable a support community which provided better tools to let people to support each other, in addition to us supporting them. We enabled people to put their code into our environment, and so that we could automatically test that against what they buy from us.
BK: So it sounds like you specifically picked the Apache license?
WD: Two very important reasons for selecting Apache: A) the community is very strong and very well respected by the buying community, and B) the license. It allows the vendor a lot more flexibility in what you can do to monetize the technology. It's much more business friendly.
BK: How did the whole IBM acquisition come about—were you looking to be acquired?
WD: It was opportunistic. I always keep thinking about it in terms of how I feel about the idea of acquisition. My goal is to change the software industry. It wasn't specific to changing the J2EE market. If you look at it from IBM's standpoint, IBM could have easily gone to Apache themselves and supported Apache Geronimo and productized it. They chose to acquire Gluecode because they believed in the business model, they understood that you need to build a technological layer on top of open source to make it digestable by a customer, and you want to deliver it in a model that the core founders are part of the company, as opposed to just taking the code. Open source is more than code, it's also people. The founders already worked at Gluecode. They also believed that you need to offer open source differently. It's not just a cheaper product, but you have to also deliver a more complete product. Gluecode had those elements, had proven itself in the market, and sold it to key customers, and in IBM's eyes that was an easy way for them to participate in the market.
From my side, what I really wanted to do was build Simula—to build another six companies, and that acquisition allowed me to do more. I'm extremely pleased with the results. I'm extremely close to my old team at Gluecode, IBM, and Logicblaze, and Simula, and we continue to be really good friends and sell each others products. The Logicblaze Fuse software release we did a month ago can be configured and bundled with IBM's WAS CE release from Geronimo. In a sense, it's fulfilling my larger goal, and provided a very enabling, tactical step in getting there—both in terms of resources and industry validation. It also has enabled me to attract into Simula the next “uber-geeks”. All in all, I was very pleased with the outcome, and my investors were pleased with the outcome. It's really enabling us to take off and build Simula.
We now have two operating companies, Logicblaze and Mergere, and both companies have exceeded the milestones of the time when I sold Gluecode to IBM, so certainly all the learning and experience I have leveraged from the experience of Gluecode at Simula Labs. That larger goal is easier to get to with the IBM acquisition.
BK: How has the response been from the open source developers working on these projects?
WD: This is the reason we selected Apache, versus something like JBoss. Apache Geronimo, while productized by Gluecode and acquired by IBM, is still a fairly open codebase. A lot of people can still make money on Geronimo, including us at Logicblaze, there's a company called Covalent that makes money on Geronimo by participating in the community, BEA is dipping their toes into it, and there are also companies like Virtuas and others leveraging the outcome of collaborative development. You never buy a community, you can only participate in it. From that standpoint, the open source project continued to be open. You can contrast that with JBoss, where no one make money from that except JBoss. The other thing is, from who made a financial return from the IBM/Gluecode acquisition, clearly the guys who started Apache Geronimo participated in it. So, they were clearly rewarded and continue to be rewarded. Plus, we didn't want to end there. We wanted to make open source founders into open source entrepreneurs, and want to be very inclusive from that point. Whether they want to work for a Simula Labs company and becoming a founding part of that team, or to become a part of its ecosystem. We still want to enable them to make money by productizing and adding to the solution. For example, there's a company called Chariot Solutions, not part of Simula, headed by someone named Aaron Mulder, a founder of Geronimo, and we keep him in our ecosystem—we partner with him, sell his product, and sell each others services. It's become a new industry around a community of open source guys. It's a $192 billion market, just in North America, and I think the whole industry should transition to an open source industry.
BK: How many of these companies do you can handle getting involved in?
WD: Our stated goal was 6-8 companies with Simula. What we've learned is we're really maxed out with three companies at a time. But, three companies is what we can handle depending on how you serialize the life cycle. The way you can scale, of course, is through adding additional management team and resources. Our goals is six to eight, if we can do a successful six to eight company startup that would be plenty of impact on the industry.
BK: It seems almost like an incubator for open source?
WD: Well, yes, and no. There's obviously a mixed reaction to the term incubator. But what we've discovered is that there's a common way to develop a business model around open source. And that common way is repeatable across market segments. The way to leverage it a little bit better is to choose market segments with certain characteristics—like the size of the market and maturity of the community. With Simula Labs we're trying to propagate our business model. We're trying to prove it out and execute it ourselves. It's not that far out that we might even open source our business model. What we've found is that we've been able to create a technical framework around executing the business model—through a process we call the Community Oriented Realtime Engineering Network—or the CORE Network. The whole idea is that when you acquire open source software, you're getting better software—better because you can bug fix it, you can make it better, you can make it more adaptive. However, you need an environment that enables all of that—which is what the CORE Network provides. That provides a rapid readiness to commercialization. That's not something a typical incubator provides. As an example, the two startups we have right now have a higher growth ramp than Gluecode ever achieved—simply because we have this CORE Network capability.
BK: You've done something people tell me is very difficult, if not impossible to do in Southern California—which is build a very successful software company—because of the lack of talent. How did you get around that issue?
WD: Because it's open source. The talent isn't in any one place—we have employees in nine countries right now—and actually, it's specifically best suited to do this in Southern California. The last thing I need is to have someone tell us “this is how you build a company”—which is prevalent in Silicon Valley. So, within this movement to open source and transitioning to globally accessible talent for engineering, I don't see any reason, and in fact it's a disadvantage to do this in Northern California.
The one thing that makes this additionally interesting, is that this model—the foundation of Simula and its open source companies—is the most likely model to produce the best possible return for the company and its investors. Because, although your development and R&D is global, operations are centrally delivered, and the marketing is really the best value. So, that presents a really interesting investment opportunity. My investors right now can play in six to eight opportunities, at the price that they would normally only get to play in one of those opportunities. It's transforming investment patterns a little bit.