This recently showed up in my inbox:
Larry-
I hope you have been well over the last year. Not sure if you remember but ... you were kind enough to participate in my VC panel last year at SAP for Columbia Business School. I am now graduating in a few months and entering the work force. I’d like to get into VC after school and have been talking with quite a few firms. I am also developing a basic investment thesis. Open source software comprises a portion of this but I keep running into the same wall when I work through the potential business model in my head. I was hoping you would spare a few minutes to talk about open source or no-license software at a high level with me.
I keep creating a circle in my head that open source will actually stifle innovation in the long run. The incentive for any company to invest in R&D seems to go out the window when the opportunity for multiples on returns that are better than consulting/service type companies are what seems to be out there. Margins and scalability are certainly lower on support than they would be on software. It seems like open source will breed commodity solutions which by nature are not terribly innovative. What am I missing because I am sure it is a fundamental flaw. I am coming at this space from an enterprise software background so I think part of it is that I am hamstrung by my bias.
Another thought that I had is that middleware and application integration has to be better with open source and that there will be companies that have a chance to change this space in the next few year. Have you seen any of these?
I was really surprised to get this, since I thought the falacies in this argument had been put to bed a long time ago. Consider the recent massive investment in Open Source companies by VCs. They obviously feel that entrepreneurs (and investors) will get rewarded for creating Open Source.
My friend here is assuming that just because the software is Open Source, you can’t get high gross margins. Look at actual gross margins in Open Source companies. Red Hat has about $210M TTM gross profit on $257M of TTM revenue – just over 80% gross margins. And that’s a pure Open Source play. Most Open Source companies use a mixed model. You can get paid well for Open Source software.
Don’t confuse the idea of an Open Source “services” business with services businesses like EDS that are essentially time & materials + 30 points of margin. That’s not what these businesses are about. Although Red Hat does not sell licenses, calling them “services” businesses is probably a misnomer.
Here’s another angle from which to look at it. Traditional Enterprise Software companies sell software with an up-front license fee and then annual maintenance and support. Historically, they always made the argument that the large up-front perpetual license fee was necessary to recoup the cost of R&D. But increasingly in practice that’s not the case. According to Rick Sherlund at Goldman-Sachs, 81% of new license revenues go to cover the cost of sales and marketing. It’s not the R&D that new license revenues are paying for, but the cost to sell and market the software. The real profit in Enterprise Software companies today is coming from the maintenance and support contracts on that software (what you might classify as services revenue). Open Source companies just cut-out the new license fee portion of that business. They correspondingly cut sales and marketing. They continue to get paid the maintenance and support revenue (which is typically called a “subscription” is this model.) The result is a more efficient business model for both the customer and vendor.
In fact, I would argue that the traditional Enterprise Software model has stifled innovation in recent years. Success is more a factor of sales and marketing muscle, and less of software innovation. The result is taht sales and marketing have been rewarded, notinnovation in products. Open Source is changing that, and allowing innovative software to once again compete against sales and marketing muscle.
In the Middleware space, the number one product by market share is Open Source, JBoss, with 34% market share. Likewise, Apache Tomcat is number 1 in its respective market. Hibernate is number 1 in the ORM [object/relational mapping] market. See Marc Fleury's comments on Open Source Middleware market share.
So can we finally put the idea that Open Source will stifle innovation to rest?