Q. What business models are currently used with open source software (OSS)?
A. Over the past 15 years, I have seen OSS move from a technological novelty or curiosity to a key foundational element of our information economy. As a technology lawyer, I have found it fascinating to witness the parallel evolution of business models in this space. To answer this question, I will give a broad overview of some of the established and emerging OSS business models that companies, organizations, and individuals are currently using.
Established OSS Business Models
1. Service model: this well-established OSS business model is premised on charging for services to support the use of OSS software. While there are many companies of various sizes that have implemented this as their core business model, the best example is Red Hat, which generated $750 million in revenue last year and has a $6 billion market cap.
At a high level, this business model is designed to benefit from large-scale adoption of certain types of OSS software. The OSS company target those user companies and organizations that prefer or require commercial support for the OSS. For example, the Red Hat business model focuses primarily on the popular Linux open source operating systems. In targeting user companies, it may offer its support service to a bank that is using Linux in its enterprise servers. In most cases, the OSS company positions itself as a supplement or alternative to relying on support from either the applicable open source community or an internal support team within the user company. In addition, this class of OSS company typically focuses on the addition of complementary services and products to drive revenue opportunities.
A recent post by Glyn Moody entitled 'Why No Billion-Dollar Open Source Companies?' discusses the challenge that the fundamental economics of OSS are bringing to the service-based OSS business model. Glyn's post has triggered a very interesting and healthy debate on OSS business models in general.
2. Dual licensing: this established, yet declining, OSS business model is premised on a dual (or multiple) license approach. Typically, this has involved the combination of a reciprocal (copyleft) license with a more conventional commercial software license. One of the most cited examples is MySQL, which licensed its database product under both a GPL and a largely conventional commercial software license.
At a high level, this business model is designed to use the OSS license to facilitate large-scale adoption of the licensed product and then follow such adoption with the offer of the commercial license. The commercial license is positioned to address certain actual or perceived deficiencies of the OSS license for certain types of adoption and use of the licensed product. In many cases, the marketing of the commercial offering has focused on creating "Fear, Uncertainty, and Doubt" (FUD) in the mind of actual or potential end users about the legal implications of licensing software under the so-called "viral" GPL, and now Affero GPL.
According to a study by the 451 Group (reported on by GlobalThoughtz Research), the proportional use of a dual-licensing approach among open source software vendors has declined from 20% of vendors two years ago down to just 5% of vendors using this approach today. I believe that the decline in dual licensing is being driven by both the inherent challenges of creating and maintaining a code base that is capable of being dual licensed as well as the increasing education and sophistication of end users with respect to OSS licensing.
Emerging OSS Business Models
3. Open core: this somewhat established, but still emerging OSS business model is premised on the open source licensing of the core software offering and the proprietary licensing of certain add-ons to that core product. Good examples of this approach include SugarCRM and Word Press.
At a high level, this business model is designed to use the OSS license (and OSS economics) to drive the large scale adoption of the licensed product and then follow on or supplement that offering with proprietary add-ons. In the case of SugarCRM, this may take the form of new CRM modules or increased scalability. In the case of Word Press, which hosts my personal blog, this may take the form of premium services that it makes available as part of its hosted-service offering.
While the open core model has generated a lot of interesting debate and polarized positions, such as the scathing critique by Brian Prentice of Gartner in his 'Open-Core: The Emperor’s New Clothes' blog post, it does, by most accounts, represent an important evolution in OSS business models.
4. Hybrid: this novel OSS business model is premised on the combination of the commercial licensing of a software product with the availability of both the source code for the product and certain rights to modify and use the source code under the umbrella of the commercial license. Since QNX, which was recently acquired by Research In Motion Limited, is the pioneer in this space, I have quoted its description of this business model type from its About Page below: "The company has pioneered an innovative hybrid software model with three main components: 1) open access to product source code; 2) a commercial-friendly licensing model that lets customers modify source code and retain ownership of their modifications; and 3) a transparent development process that allows customers and community members to participate in product development – a benefit normally restricted to open source projects. Put simply, the new approach combines advantages of both commercial and open-source software models."
While we are still in the very early days, the hybrid model represents an intriguing twist on existing OSS business models and offers an interesting blend of openness, transparency, and commercial certainly.
5. "Entersource:" earlier this year, I came across the term "entersource" in Eric Knorr’s post on Infoworld’s Modernizing IT blog entitled 'Open source: Less profit, more fun', which leads off with the provocative statement that "Open source ain't what it used to be. It's both more and less." According to Eric, "a diminished percentage [of developers] work for healthy open source software vendors, where the old-fashioned business model -- give away the code and make money on support -- isn't doing so hot." Eric quotes Black Duck CEO Tim Yeaton, who sees the area of enterprise application development as the "real open source explosion", and Eric reports that Michael Skok of North Bridge Venture Partners dubs this co-mingling of open source and enterprise software “entersource”.
Since the four other models discussed above are grounded in the fundamental business equation of how the company makes money, it may seem, at first blush, that the customer-centric approach of entersource is a poor fit as a business model. In Sesame Street lingo, “one of these things is not like the others”! In fact, I see entersource and its variants as the major emerging OSS business model.
According to Eric’s blog, Michael Skok sees entersource "chiefly as a means for collaborative development". As I noted in my related blog post (and discussed in greater detail in my October 2008 OSBR article 'Treasury of the iCommons: Reflections of a Commons Sourcing Lawyer'), this exploding trend of "commons sourcing" is increasing. As Eric points out, this is reflected in the fact that "enterprise developers are collaborating across company boundaries to develop components that can be shared under open licenses". This phenomena is driving unprecedented cost savings within companies that are frequently order-of-magnitude improvements over the status quo. When interviewed by Makesh Sharma in The Australian article entitled 'Open source enables innovation without lawyers or fees', Roger Burkhardt, Ingress CEO, makes the point that "the open source model allows us to bring together the best minds in the world to work on a problem" and allows engineers from different companies to "collaborate without months of legal work."
As Michael Skok notes, there is an increased focus around OSS on "real ROI and payback, which has had the effect of making open source a 'mainstream, reliable, de facto part of the landscape'". In addition, Michael observes that "very few of these [open source] projects will reach the critical mass required to create a company," adding that, "a good product doesn't make a good open source project." In fact, he says, it is the reverse: you need a community first, and then a project to serve that community." In this context, we can expect to see entersource and its variants as a critical addition to the OSS business model landscape.
I'll end with Matt Aslett's quote from the GlobalThoughtz Research post noted above, "it is now more difficult to actually isolate and identify all of the vendors that have an open source business strategy. More vendors than ever are now using open source at different points in their process and applications and usage is not limited to pure-play open source vendors." As such, it is tough to wrap up a piece on open source business models since there is always more to write!