How Red Hat is Monetizing JBoss

A little more than a year ago Red Hat (RHT) acquired JBoss to enter the open source middleware market and the financial results thus far have been mixed. But now analysts are becoming more upbeat about JBoss and what it can contribute to Red Hat.

In an analyst note, Credit Suisse analyst Jason Maynard upgraded shares of Red Hat to an outperform from a neutral and outlined many of the items we already knew: Oracle’s unbreakable Linux effort hasn’t been a threat; Red Hat Enterprise Linux 5 has been well received and business overall is solid.

But what caught my eye was Maynard’s optimism around JBoss. JBoss had its sales issues in prior quarters, but Maynard said that Red Hat may have figured out a licensing model that works.

Maynard argues that JBoss is likely to show improvement on sales in fiscal 2007. Here’s why:

In particular, we think the new licensing model is a promising move by Red Hat to monetize the JBoss installed base. JBoss as a stand-alone company was very successful at getting customers to adopt and use the technology, however getting them to pay for subscriptions was the challenge. A couple months after the Red Hat acquisition, JBoss founder Marc Fleury disclosed that, “on average we only monetize 3% of our user base for JBoss”. Red Hat is attempting to remedy that issue by employing a similar RHEL-Fedora licensing strategy that it successfully used to build its Linux business. In short, Red Hat will make cutting-edge but untested versions of the code available for download at no cost (jboss.org), while requiring support contracts for customers to use “certified”, tested versions. We have already heard some good initial field sales feedback about this approach even though the company is just a month into the process. JBoss accounts for less than 10% of revenue but could finally prove to be a more meaningful contributor if this new model gets the traction we expect.

The JBoss licensing makes sense on many levels. For starters, Red Hat customers are already used to that setup. And given that Red Hat’s business is already doing well Red Hat’s model for JBoss could provide a lot of extra gravy. If successful, Red Hat could use the same licensing scheme for future acquisitions.

The big question is whether JBoss’ existing customers–the 97 percent that aren’t used to paying for anything–will go along with Red Hat’s monetization efforts. Any JBoss customers want to comment?

Comments

[Brian Croner]

From the heart: To become an open source developer, you must have the mind to earn knowledge on your own. It implies incessant research and practice. JBoss is free. Java is free. You can't monetize a producer in the same manner a consumer. An open source software developer such as myself learns how to learn upon his first learning of an open source tool. Chances are, whether or not JBoss is the first open source technology a user learns, it's not the only one. These users are the innovators themselves, and are smarter than the corporate businessmen. The corporations have nothing worth trading for money to offer to them. Capitalism thrives by anticompetitive trust practices. Open source, free software dominates by its strictly collaborative nature. To be a successful open source developer, you must first love it enough to do it for free. That weeds out lazy people, and restricts the audience to a community of people who write source code in a cooperative manner, readily adopting each others' code and returning a larger, more functional piece to the community, which is completely immune to the anticompetitive nature of proprietary monetization.

To the masses: In other words, if I'm already an open source programmer, you can't charge me to receive your software information or support services. You'll never make a dime. I already have whatever you can try to offer me. If it's written in English somewhere, I don't need you and in fact am better off without your interference.

Theory: To improve monetization, they need to capture the audience who would spend money to make their own lives easier with this software. People want what money can buy, but if they can already program anyway, you can't charge them for the service of helping them program. Microsoft makes money off of enabling lesser-gifted engineers program. So extend the audience of JBoss users to those who are dependent on GUI-driven development, like with the Microsoft developers. The visual developers without much abstraction would readily buy the ability to program middleware-enabled apps like those of JBoss. They are reliant on spending money for software development tools that reduce coding to configuration of GUIs in order to achieve productivity. However, this increases the value of the left-brained engineers, who already are being targeted to get bucks through sourceforge.net as tech consultants over the web.

Nutshell: Maintain the text-editor programmers, but connect them with the visual click-and-configure programmers that don't really know Java as just another language. Make money from the developers who can't develop without the help of a GUI, and give up charging the left-brained open source coders. And, following this, you're probably going to have to attract the open source coders to your site, and achieve a competitive advantage in doing so. So in the long run you're going to eventually find yourself paying some expert people to help others use your stuff to increase your audience. Knowledge is the new power.