Q. I just don't understand how one can develop a business model around open source--after all the resource is freely available.

A. Open source business models continue to be a source of great puzzlement, yet the situation is not so different from monetizing Natural Resources. With Natural Resources such as minerals, oil, wildlife, and water, the primary asset is some Natural Resource, which, historically, has and continues to be viewed as free. Similarly, open source software assets are generally and freely available.

To help in understanding open source business models, we appeal to a technology commercialization theory by David Teece. Teece's theory has several dimensions, and the dimension we wish to apply here is called Complementary Assets. In the development of his theory, Teece observed that it is very rare that a technological innovation (the primary asset) can be commercialized without the support of other assets known as complementary assets. Examples of complementary assets include marketing, sales, human resource management, office space, information technology, transportation, manufacturing, and sales channels. Teece further observed that the role and importance of these complementary assets may be of considerable importance in the commercialization of the primary asset. This is particularly true when the provider of some necessary complementary asset has a monopoly.

To return to our parallel with Natural Resources, with the fish off the Grand Banks of Newfoundland and Labrador viewed as a free asset, competition intensified around such necessary complementary assets as large factory ships that could spend months at sea harvesting and packaging fish.

And this is generally the case: when the primary asset is a commodity, competition intensifies around the necessary complementary assets.

Open source software turns a previously privileged primary asset into a commodity and instantly creates a commodity out of a new-to-the-world primary asset. The important insight surrounding turning a primary asset into a commodity is that the firms purchasing these primary assets do not typically experience savings. Rather, competitive pressures are such that the expenditures of these firms simply shift towards other competitive distinguishers. And commonly, some of these new expenditures are on complementary assets such as training, books and customizations that become affordable as a consequence of the primary asset being priced as a commodity. Also noteworthy is that such complementary assets may, or may not, be a competitive advantage of the purveyor of the primary asset.

In sum, the key to comprehending most open source business models is to distinguish between (i) the primary asset and (ii) the complementary assets that are necessary to support the commercialization and use of the primary asset. And to recognize that when the primary asset is a commodity, competitive pressures will likely increase spending on the complementary assets.

Recommended Reading

1. Teece, D. J., 1986. Profiting from Technological Innovation: Implications for Integration, Collaborations, Licensing and Public Policy. Research Policy, Vol.15, Issue 6, pp. 285-305.

Q. How do the motivations behind open source translate into business models?

A. Assorted explanations have been put forward that contribute to our understanding of the motivations behind publishing source code. These include:

• Richard Stallman circa 1983: The motivation is philosophically grounded in fundamental rights such as freedom and liberty. In particular, source code should be published because, like any other scientific theory, it should not belong exclusively to any one individual or group.
• Linus Torvalds circa 1991: The motivation is a means for achieving superior software programs. More specifically, by publishing the source code, the number of individual examining the code is potentially increased which increases the likelihood of the code being improved.
• Eric Raymond circa 1997: The motivation resides in the superior software development methodology which is meritocracy based, highly visible, and superior to the traditional in-house, closed source, development methodologies.

By 1999, increasing attention was directed towards better understanding business models and motivations. Such interest was partly fueled by the success of firms, notably Redhat, that were perceived as selling something that was generally and freely available. By 2001, interests had expanded to business strategies in search of explanations; for example why IBM was investing a whopping $1Billion to advance Linux. It is arguable that today, Open Source is fueled primarily by business motivations.

Recently, individuals have taken a keen interest in inventorying the business models of firms that use open source to compete. A re-occurring theme is providing value added services such as educational material and courses, professional services, and subscriptions. Individuals have also taken a keen interest in the utility of open source for developing competitive business strategies. For example:

• Firms providing services and customizations around given technologies may be incented to see those technologies open sourced as open source technologies are typically more affordable. More affordable technologies may lead to an increase in their use, and this will likely result in an increase in the demand for the firms' services and customizations.
• Firms competing in a market where there is an effective monopoly on some technology may be incented to see that technology open sourced. Such commoditization of the technology would likely remove the monopoly, change the competition dynamics in that marketplace, and possibly better position the firms' to appropriate a greater market share.
• Firms that have an interest in sustaining the development of technologies that have little marketplace differentiation and/or margins, may elect to collaborate to advance these technologies as open source. By collaborating with an open source project, the firms' may save money, generate good will, avoid anti-trust issues, and possibly create an industry standard that the firms' effectively control, all while retaining each firm's independent interest in sustaining the technology.

Examples of selling models include:

1. Sell subscriptions: Used by RedHat

The software is free but additional features or services associated with commercial software are only available via a monthly, quarterly or annual subscription. Other examples include the use of a dual license, paid updates, indemnification, stack maintenance, maintenance and support, and content through knowledge portals.

2. Sell services: Used by Google, Yahoo and Mozilla

The software or an interface is made available to companies. This ultimately drives traffic to a host website where it is used to generate advertising dollars and business intelligence. Other examples include testing, hosting, training, and consulting.

3. Sell hardware products: Used by Digium

The software is provided for free but the customer pays for the hardware. This embedded approach uses the open source projest as a platform. Other examples include hardware with embedded open source software (OSS) which is delivered on a cell phone, a system product that includes an open source asset and integrates with complementary products such as the Apple computer, and hardware with OSS running on top which is delivered as an out of the box solution such as TiVo.

4. Sell software products: Used by Oracle on Linux, Oracle9i Real Application Cluster (RAC)

The core software is free and additional features such as optimization and consulting are paid for by the customer. Other examples include a software fork for a closed software stack, certified stacks, and OEM products.

The above examples should suggest to the reader that the motivations behind publishing source code is varied, complex and evolving and that an authoritative and comprehensive understanding has multiple dimensions.