In the modern world of open source, software vendors are becoming increasingly irrelevant. At least, at the infrastructure layer.
In fact, in a world awash in open source software built and contributed by companies that have no interest in directly monetizing that software, it is the packager of other people’s innovations that thrives. Small wonder, then, that the one “open source company” making serious money is Red Hat, a company whose real innovation is in how it packages and streamlines the innovations of others.
Giving it all away
In my inbox this morning was a digest email with the following headlines:
- Nike releases open source software to play with the techies
- Facebook is creating an open-source cellular system
- Uber Drives Gains With Open-Source Development
- LinkedIn open-sources URL-Detector Java library, LokiJS 1.4
While open source in its middle age (mid-90s to mid-00s) was all about vendors releasing this or that open source project, today open source seems to have left behind the vendors. The common thread running through the announcements above (and countless others you can see by running a quick Google News search) is a focus on developers, not sales.
SEE Why AWS Lambda could be the worst thing to happen to open source (TechRepublic)
LinkedIn doesn’t sell software, open source or otherwise. Ditto Nike, Uber, Facebook, and a range of other companies (like Capital One). These are companies that hope to enhance their code through developer contributions, or want to better attract and retain talent. Not a single one of these companies is trying to make a mint by peddling open source software.
That’s fortunate, because the surest road to failure runs straight through an open source business model. This isn’t to suggest it’s impossible to make money with open source. It is, and I’ve worked for companies that have made over $100 million a year doing just that, albeit unprofitably.
No, it’s simply an acknowledgment that, in a world intent on giving its best software away, selling software is a fool’s business.
The Red Hat counterexample
Unless, of course, you’re Red Hat.
As Steve Nellis correctly argues, “There’s only one firm that’s proven it can consistently make money off open-source software, and that’s Red Hat.” But then, Red Hat doesn’t sell open source software. Instead, it sells peace-of-mind around popular open source code that companies would like to use, but preferably with training wheels, as it were.
Red Hat’s model has remained remarkably consistent over the years. Way back in 2006 I wrote about the genius of its model, suggesting, “Red Hat’s model is a product of necessity. The company had to figure out how to survive its lack of code ownership, and found a brilliant way to turn this apparent deficiency into a strength.”
SEE Red Hat’s open source success story built on killing complexity in IT (TechRepublic)
Red Hat’s model won’t work for all areas of software. It’s irrelevant to the application layer of the software stack, but a direct hit at the infrastructure layer. Some things–like Hadoop–seem like they’d be a fit (complex, popular, community-driven software), but Red Hat CEO Jim Whitehurst told Nellis that he’s “on the fence” as to whether it’s an area that would work for Red Hat’s model. And so, Red Hat focuses on the operating system, virtualization, storage, and other areas that enterprises are willing to pay for assurance that Red Hat’s certified open source software will work.
The open source leader keeps chugging away with this model, with the only real threat being cloud-based systems like Amazon Web Services. The more AWS, for example, makes running open source infrastructure simple and cheap, it potentially cuts into Red Hat’s ability to charge for that service. Red Hat is hedging this cloud bet with OpenShift and OpenStack, but time will tell if its appeal to datacenter-minded enterprises will hold in the face of a swelling tide for public cloud infrastructure.
In the meantime, however, there has never been a worse time to sell software, just as there has never been a better time to be Red Hat. The more code the Facebooks and Googles of the world release, the more opportunity for Red Hat to package it for mainstream consumption. The only risk, again, is that AWS will make it even easier to digest than Red Hat does. The race is on.