Tim O'Reilly: Open-Source Considerations

Let me point out a couple of things here. First, I do think that the biggest open source success stories were not people like Red Hat. They were people who actually built and sold services that used open source. And obviously a bunch of obvious players, Google, Amazon, Facebook, Twitter, they're all open source companies that took open source, they used it to deliver a service. But this even more direct example, the entire web hosting business, very thin skin of selling DNS as a service, selling WordPress as a service, selling Apache as a service for a subscription fee. And so there's a pretty clear case where there was an open source business model that was sitting in plain sight that didn't look like the business model of a software company. And I think it's really important to look a little sideways. There was a wonderful talk at our open source convention once by a guy named Robert Lefkowitz, named Rommel, and he basically started in with this long conversation about Sharia compliant mortgages, which is under Islam you can't loan money. So he said, basically the Islamic banks, he went through the two or three different ways that they get around this. They basically rent you your house for 30 years and at the end of 30 years, they make a gift of it to you. So he kind of explains this and then he says, okay, so now let me show you the P and L for Borland versus the P and L for Red Hat. And he says, look, the percentages are identical all the way down the line. There's only one difference. This line says licenses and this line says subscriptions. Sharia compliant mortgage, she basically, Red Hat, it was the same business as Borland. So when you think about open source, I think the most important thing to think about is where do you get your business advantage? Now there are open source business models where you get your business advantage because, hey, at least in the early days, this is a great way to get my software in people's hands. But in the era of cloud, that's actually not really much of an advantage anymore because anybody can get their software in other people's hands. So that's really gone. That was the idea for MySQL and the whole idea of dual licensing. We'll build the market by letting anybody use it and then we'll upsell from there. But today it's probably easier to do that with a cloud app than it is with open source. So you have to think, I think, a little bit harder about where you get your business advantage and why you want to open source it. You might want to get contribution from users. I don't know anything about your business. A huge way to do it is that there's some other network effect that comes into play as a result of of your software. Either you get a network effect because lots and lots of people use it, or because it produces data that you are actually going to monetize is another. So there are a lot of answers, but I would just urge you to think sideways and don't get caught in open source being a simple variation on a business model where I would have charged and I would have had a proprietary fee.