As Weill et al. point out, an observation shared by many, few concepts in business, especially those related to Internet- or IT-related ones, are as widely discussed as “business models” but as poorly explicated. Their analytical study of 1000 US companies also showed that business models are a better predictor of financial performance than industry classifications and that some business models do, indeed, perform better than others. Specifically, selling the right to use assets is more profitable and more highly valued by the market than selling ownership of assets. Using a similar approach, Zott and Amit also found that the business model matters, with novelty-centered ones important to the success of new firms. The consensus seems to be emerging that business model innovation may be more important than any other source of new venture success.
Unfortunately, what even constitutes a “business model” has little agreement and frameworks that attempt to organize these concepts extend from the global to the minutely absurd. Most bandy the term about imprecisely. However, let it suffice to define a business model as the “why, how, and what means for a business to generate revenues and achieve profit objectives.”   A business plan includes both an explication of the business strategies and the business models underlying the business. A company’s business models, of which there may be a few, combine as a means to achieve profitability goals under the direction of an overall strategy, which itself also concerns itself with positioning and the competition. Joan Magretta argues that a good business model is essential to every successful organization, that business models are in essence stories that explain how enterprises work, and an effective business model needs to be supported by numerical analysis about how the organization will make money.
From this big picture (and sometimes preceding or leading it), many have attempted to define or organize business models associated with the open source phenomenon. John Koenig most recently presented a view of seven open source business strategies to generally positive reviews. Other recent commentators have been Matthew Asay and Steve Walli from Optaros among others. There have been quite a few academic studies, most well presented. Earlier, the Open Source Initiative presented four business models, deriving from the seminal open source paper on the Cathedral and the Bazaar by Eric Raymond; these four were later expanded by another five by Frank Hecker.
The Internet, and now open source, are challenging traditional views as to how businesses make money. These challenges, in turn, are the genesis I suspect for this recent attention into business models.
A Taxonomy of Open Source Business Models
Three new terms are now cropping up surrounding open source business models, and a fourth is suggested in the taxonomy I present below:
- Commercial open source — this is the collective term applied to for-profit ventures that have a strong open source component within their business model(s)
- Professional open source — this term, being pushed most strongly by JBoss and its VCs, refers to a commercial open source business model wherein revenues are obtained from training, consulting and other services based on an open source project
- Dual license — this business model innovation, embraced by MySQL, Trolltech and Sleepycat, refers to providing both open source licenses with commercial limitations and commercial licenses for a fee. There are many nuances to this model, which is emerging as the most popular approach for new commercial open source ventures
- License choice — because there are not simply two options and there are many variants, this is the more expansive term I’ve given to the ‘dual license’ model. Under this business model, users or customers may choose from a variety of licenses offered, one of which is open source, the others commercial.
It should also be noted that new ventures that embrace these models are also now being referred to as “second generation open source” companies or ventures. Recent venture capital money is going most strongly to the dual license model, though subscription for services has been strongest the past couple of months.
Nonetheless, synthesizing the various views and writings on open source business models has caused me to derive my own taxonomy, presented in the chart below.
Open Source Role
|Create||‘Pure’ OSD*||‘Pure’ OSD*|
|Mixed Source||Mixed Source|
|Delayed Release||Sell it, Free it|
|License Choice||Dual License||Hybrid Model|
|Commercial Extensions||Commercial Extensions|
|Service||Support Service||Consulting||Support Sellers|
|Deployment Service||Integration||Professional Open Source|
|Direct Use||Embedded Device|
|Market Leverage||Hardware Benefits||Hardware Sales||Loss Leader|
|Drivers and Interfaces||Widget Frosting|
|Software Revenues||Awareness Revenue|
First, the chart distinguishes the four ways in which open source may be involved in an organization (the no involvement option is not presented): creating open source software, servicing it in some manner, using it to generate revenue or to lower costs, or gaining marketing leverage by associating with it. Within these roles, the taxonomy then presents 10 discrete business models, each described below. Then the chart presents about 30 revenue or cost streams associated with these business models (note using open source to lower internal IT costs, for example, is a legitimate business model). The last column (‘AKAs’) presents some of the labels given by others to these business models.
Thus, this taxonomy suggests 10 business models that may involve open source, which may also be combined or not with one another within an overall strategy. These are, with example organizations using them at present:
- ‘Pure’ OSD — this refers to the standard open source development (OSD) effort staffed solely by volunteers and which is non-profit. If embraced by a commercial venture, one of the other nine business models below must be used to generate revenue (Mozilla, Debian, WordPress)
- Mixed Source — in this model a commercial software company embraces both commercial license and open source software. This can be a mix of different product lines or positions in the software stack (Oracle, Sun, IBM, Novell) or a time differentiator where the latest versions are commercial and earlier ones open (Ghostscript) or differences between client and server (Jabber) or limited in scalability (Sybase), among possible other variants
- License Choice — this model is where a choice of open and commercial licenses is offered for the same software. The open license restricts commercial use in some manner. To bypass these restrictions, be it the “viral” requirements of say the GPL license (MySQL, db4objects), obtain commercial support and warranties (MySQL, ActiveGrid, SugarCRM, OpenOffice), allow use within commercial products (Trolltech, Sleepycat), or even more complicated variants (Six Apart), choice is given to the user. This model is limited to entities that have clear title to the software intellectual property (IP) rights
- Commercial Extensions — this model begins with a core of open source software but extends it in useful and proprietary ways to support commercial licensing. Some examples include companies pursuing large-scale Linux clusters (Levanta, SiCortex), improved interfaces (Xandros), better functionality (Zope, XenSource, Sendmail, MontaVista), etc.
- Support Service — Covalent, SourceLabs), open source certification or version compatibility (SpikeSource, OpenLogic), compliance with licensing terms (Black Duck, Palamida), or publication or documentation (O’Reilly). Actual revenue models can be based on time-and-materials or subscription fees
- Deployment Service — this service model is specific to open source implementation and deployment and may involve integration or custom development (Gluecode, Optaros, IBM, HP) or packaging (many of the Linux vendors) or hosting open source applications (Six Apart, PlanetXoops)
- Service Enabler — exposed directly to customers. This can include transaction vendors (Ebay, Amazon), advertising vendors (Google, Yahoo), or subscription services companies (salesforce.com, Lindows)
- Direct Use — this business model involves the use of open source embedded in a physical device, most often involving Linux (Nokia, TIVO, Cisco) or for internal use to lower IT costs (many)
- Hardware Benefits — in this business model, as in the next one, the business model involves associating with open source intiative(s) in order to increase market visibility, awareness or acceptance. Demonstrating strong ties with open source can help increase server hardware sales (IBM, Sun, Apple) or open source utilities can be included in the hardware as operating systems, etc., to lower costs (Sun, IBM, Dell)
- Software Revenues — there are many adjunct software benefits that can flow from market association with open source initiative(s), subsumed under this business model. Some of the variants include: software revenue from component modularization in other portions of the software stack (IBM, Oracle, Novell); software package sales as an upgrade path from open source baselines (IBM, Red Hat); influencing standards to support internal development capabilities (IBM, many); or building open source-centered partnerships to thwart competition (many against Microsoft).
Note that a separate distinction was not made above for subscription fees. That is perhaps arbitrary, since market acceptance may be promoted or not by a pay-as-you-go model.
 P. Weill, T. W. Malone, V. T. D’Urso, G. Herman, and S. Woerner, “Do Some Business Models Perform Better than Others? A Study of the 1000 Largest US Firms,” MIT draft May 6, 2004, 39 pp. See http://seeit.mit.edu/Publications/BusinessModels6May2004.pdf. This is one of the seminal, large-picture overviews of the business model question, but one with little direct relevants to IT or open source.
 C. Zott and R. Amit, “Business Model Design and the Performance of Enterpreneurial Firms,” Dec. 4, 2004, 38 pp.; see http://www-management.wharton.upenn.edu/amitresearch/documents/New_Folder/Business%20Model%20Design%20120504.pdf
 For two additional starting points, look up ‘business model’ on Wikipedia (http://en.wikipedia.org/wiki/Business_model) or refer to an article such as from H. Chesbrough and R.S. Rosenbloom, “The Role of the Business Model in Capturing Value from Innovation: Evidence from Xerox Corporation’s Technology Spin-off Companies,” Industrial and Corporate Change, Vol. 11(3): 529-555, 2002; also found at: http://www.cvn.columbia.edu/jl/readings/Chesbrough_ICC_2002.pdf
 J. Magretta, “Why Business Models Matter,” Harvard Business Review, May 1, 2002, 6 pp. See http://harvardbusinessonline.hbsp.harvard.edu/ for ordering information.
 J. Koenig, “Seven Open Source Business Strategies for Competitive Advantage,” see http://www.riseforth.com/images/Seven%20Strategies%20-%20Koenig.pdf. David Pool later suggested another 3 models for a total of 10 on this baseline; see D. Pool , “Business Strategies for Free and Open Source Software Companies and Communities,” http://nakedape.cc/info/Adopt_OSS_or_Die.doc
 M. Asay, “Open Source and the Commodity Urge: Distruptive Models for a Distruptive Development Process,” November 8, 2004, 17 pp.; see http://www.open-bar.org/docs/matt_asay_open_source_chapter_11-2004.pdf
 Stephen R. Walli, “Under the Hood:Open Source Business Models in Context,” Optaros Corpany; see http://stephesblog.blogs.com/my_weblog/2005/04/usenix_talk_ope.html
 See, for example, S. Krishnamurthy, “An Analysis of Open Source Business Models,” see http://faculty.washington.edu/sandeep/d/bazaar.pdf
 E. Raymond, The Cathedral and the Bazaar, see http://www.catb.org/~esr/writings/cathedral-bazaar/
 F. Hecker, “Setting Up Shop: the Business of Open-source Software,” last revised June 20, 2000, 23 pp.; see