Open-source software is a hit with enterprises. It gives their IT organizations the ability to cost-effectively provide for their constituents without ever being locked-in to a particular vendor. Indeed it’s hard to find a company today that doesn’t use some open-source software. And certainly you have companies like Facebook that are completely running on an open-sourced infrastructure.
The vendor side of the equation is more complicated. It has proven difficult to command a premium selling a product that is derived from community owned source code, especially when anyone can re-package the same code and compete. Red Hat was the first company to go public with a business model based on an open-source technology (Linux). Thirteen years later the track record of building big businesses on open source is not a good one. Only three companies have exited or gone public at a valuation of $1B or more (Red Hat, MySQL and Hortonworks). Only five companies (Red Hat, MySQL, Hortonworks, Sourcefire and XenSource) have exited or gone public at a valuation of $500M or more.
Even Red Hat a giant in open-source has a business that pales in comparison to the Microsoft Windows franchise. The value captured by Red Hat is a fraction of the value created by Linux, the technology it helped popularize inside the datacenter.
All this begs the question: If the track record of open-source business exits is so lackluster why is so much venture capital flowing to fund these businesses? Consider the following facts: Cloudera (commercializing open source Hadoop) alone has raised a staggering ~$1B in venture funding. Docker (commercializing open source Linux containers) has raised $162M in venture funding and Hortonworks raised close to $250M before going public.
Is this a case of investors ignoring the lessons of the past or are there different forces at work this time around?
I am going to posit one plausible explanation to why the current crop of open-source companies should be looked at with a new lens. The difference lies in the nature of the technology being commercialized. Until recently the open-source movement could succinctly be summarized as a force driving commoditization of the software stack. As parts of the technology stack matured they became standardized and modular making it easier for customers to substitute one part with another. This coupled with diminishing rates of innovation opened the door for credible fully featured open-sourced alternatives to take market share from their proprietary closed-source counterparts – and in most cases the incumbent companies reacted by shifting their focus up the software stack and capturing value through a combination of selling apps, value added services and management tools (see table below).
However monetizing these successful open-source projects proved elusive because at some level they were merely cheap substitutes to proprietary software. For example, MySQL built a credible open-source database alternative to Oracle but databases are relatively standard commodities these days, making it impossible for the company to differentiate in any meaningful way. Quite simply MySQL was too late to the party – everything needed of a database had already been built in the 40 years since the invention of the relational database.
The new open-source projects are different in a fundamental way; they are not copycat versions of closed-source software playing catch-up to an existing proprietary technology. New open-source, developed by the web-scale pioneers (Google, Facebook etc.) to handle the unique requirements of the cloud, is bleeding edge and innovative. I call these open-source technologies native open-source to underscore the point that they have no closed-source counterparts. Five or ten years ago, most of these projects would have been born closed-sourced but the market realities of today dictate otherwise.
The fact that these native open-source technologies have no closed-source substitutes and are built for a new market where the old stack is largely irrelevant creates unprecedented potential for value capture. Customers bought traditional open-source to reduce costs and switch away from expensive proprietary alternatives. Customers buy native open-source because it solves an unmet need they have. The former speaks to a cost-conscious mindset and the latter to a premium and quality conscious mindset.
More specifically companies based on native open-source have the following advantages relative to their predecessors:
To be clear, I am not suggesting that native open-source is easier to monetize than closed-source. It isn’t. Nor am I suggesting that all native open-source projects can support big companies, the same way no one would suggest that all proprietary technologies lead to big businesses.
We live in a world where enterprise IT buyers will not be sucked into another closed-source platform lock-in. Infrastructure platforms going forward will have to be open-source. Period. That said, companies selling native open-source based products have more avenues for differentiation and monetization than could be said for any of their open-source ancestors. It will take an exceptional piece of core technology at the hands of a great team to pull it off; but it can be done. The winning companies will innovate faster to stay ahead of both the competition and the commoditizing arc of ever expanding open-source functionality.
The good news is we are entering a period where innovation will drive increasing value creation for customers and the open nature of the platform will keep all vendors honest and focused on technology differentiation. Much the same way amazing companies were built on open Internet and Web standards, we can and should expect the same in this new era of cloud infrastructure.