Recently, at the OSBC, I spoke on how to align your intellectual property strategy to your open source business strategy. This issue can be very simple if you are joining or contributing to an existing project, because you will be bound to use the license of the project. However, if you have more flexibility, you need to consider a number of elements: (1) the sources of revenue (2) the type of product (3) business model (4) type of project (5) channels (6) type of community and (7) competitors. Once you have answered these questions, you then need to review your intellectual property options, such as such as patent, trademark, trade secret, copyright, licenses and domain names to implement your open source business strategy.
For example, a new web infrastructure software company might decide to adopt a dual licensing model and to adopt a license which is relatively compatible in order to interact effectively with other open source software used on the web. The company has decided that its most important intellectual property will be patents and trademarks. The license options include MPL, CDDL, CPAL or if integration is less important, GPLv2, GPLv3 and AGPL.
The materials also describe some of the mistakes that open source companies have made. If you are interested in the presentation it is posted on the OSBC website. http://akamai.infoworld.com/event/osbc/08/docs/GC-Radcliffe.pdf
Harvard Business School recently published a case on whether a software game company, KMS, which makes a device which permits amateurs to sound like professional musicians should adopt an open source business model. http://harvardbusinessonline.hbsp.harvard.edu/b02/en/common/item_detail.jhtml?id=R0804XThe case demonstrates the increased recognition of the strategic importance of decisions about the adoption of the open source software business model. Unfortunately, the case does not reflect the developments in business models for commercial open source software. The case focuses on an open source business model based primarily on providing technical services. Yet most commercial open source companies have adopted a dual distribution model. Moreover, as Marten Mickos noted in his 2007 keynote at OSBC, commercial open source companies have thirteen ways to make money, with four of them which he identifies as “scalable”. In addition, the analysis in the case if confused because KMS’ product includes hardware as well as software. Such hardware could give KMS a substantial advantage against competitors trying to provide an open source version of the product. In my experience, virtually all decisions about the adoption of open source business model deal solely with software products. Consequently, I think that the case would have been more powerful (and more realistic) to focus on case in which the product was solely software.
The Case Commentaries are very interesting. Jonathan Schwartz of Sun Microsystems, Inc. makes the critical point that KMS needs to determine its business goals before the company can make a meaningful decision about adopting an open source business model. He draws a contrast between Apple and Nokia in the handset market: Apple is trying to define what a handset should be and they sold 4 million iPhone handsets last year. On the other hand, Nokia is trying to be the largest handset maker in the world, has adopted an open platform and sold 400 million handsets last year.
Gary Pisano of Harvard Business School was also very insightful about the necessary elements for success in converting to an open source business model: ensuring that your software architecture is “modular” and creating a developer community. The creation of a developer community is a significant challenge for a new product and quite different from the skills required for developing and distributing proprietary software. He also notes that natural advantages conferred on KMS by its role as the creator of the “platform”. Finally, he focuses on the new reality for all “proprietary” software vendors: they need to be prepared for competitors who adopt an open source model.
Eric Levin makes good points about the importance of being able to control the brand and the strategic life cycle, but concludes that KMS has alternatives to adopting an open source business model such as adding personalization. However, I think that this alternative is an illusion and it seems to contradict his prior points.
The final Case Commentary by Michael Bevilacqua focuses on legal issues and, from his view, the significant additional risk of intellectual property infringement in an open source business model. I don’t agree with his conclusions. First, most “proprietary software” includes significant amounts of open source code which would carry risks similar to a pure open source business model. Second, he notes the increased risk of patent infringement in open source software. I disagree that the risk of patent infringement is greater in open source companies than in proprietary software companies. Most proprietary software companies do not undertake patent searches prior to writing software, so both types of companies are equally at risk of infringing a third party’s patents. However, the open source business model does entail legal risks: the scope of many important open source licenses (such as the GPL) are unclear because they use terms, such as derivative works, which are poorly defined in copyright law when applied to software and the licenses have never been interpreted by courts. In addition, the remedies available under open source licenses, whether injunctive relief or only monetary damages, are not clear. Consequently, many companies limit the use of open source software based on the open source license under which it is provided.
It is great that Harvard Business School has acknowledged the strategic importance of decisions about the open source business model, but we hope that their next case is more focused.
The recent report by Gartner, the State of Open Source 2008 (http://www.gartner.com/; report G00156659), as summarized on their site provides some very interesting conclusions:
1. By 2013, a majority of Linux deployments will have no real software TCO advantage over other operating systems.
2. By 2012, 90% of enterprises will use open source either direct or embedded.
3. By 2011, open source will dominate software infrastructure for cloud-based providers.
4. By 2012, software as a service (SaaS) will eclipse open source as the preferred enterprise IT cost cutting method.
I agree with Gartner that open source will continue to penetrate more companies, but I think that it will occur much more rapidly than suggested by Gartner. And they are absolutely correct that use of open source is “elusive”. We find that virtually all of our clients use open source even if they are not aware of it. Gartner captures the reality of open source use in their statement that: “Users who reject open source for technical, legal or business reasons might find themselves unintentionally using open source despite their opposition.”
I don’t agree with their conclusion about Linux and SaaS. I agree with the skepticism expressed by Mark Taylor http://news.zdnet.co.uk/software/0,1000000121,39379900,00.htm. My experience is that the use of Linux continues to grow rapidly and it is likely to take an even more important role in mobile devices. The statement about SaaS confuses a business model with a method of developing software. Many open source companies use SaaS as a distribution model and it does not make them less “open source.” http://lawandlifesiliconvalley.blogspot.com/2008/03/open-source-overview-from-osbc.html
The report once again emphasizes how open source is becoming part of the mainstream. A decade can make a big difference: “Microsoft: Resistance is Futile” http://www.news.com/2009-1023-229218.html.
A recent 451 Group report notes that venture capital investments in open source companies are at an all time high this quarter. http://blogs.the451group.com/opensource/2008/04/01/vc-funding-for-open-source-hits-an-all-time-high/. They raised $203.75m, up from $100.40m in the same quarter of 2007. He expresses caution that few of the deals were seed or Series A and that much of the funding was raised by some mature companies, such as SugarCRM.
This increase in funding for open source experience is consistant with what I am seeing in Silicon Valley where I work with about 40 startups (not all open source). Most software venture deals have an open source component to them and venture capitalists are very interested in new open source projects. I know of at least four new open source companies that are seeking funding, several are based on existing projects. So I disagree with Matt that the relatively smaller number of seed and Series A deals are a cause for concern. Seed deals in particular are difficult to find (several of the companies that I mentioned above have bootstrapped or used friends and family money, so they are basically invisible). In addition, I know of four foreign open source companies that are coming to the US because of the size of the market and the depth of the venture capital market. I think that 2008 will be another record year for open source funding.
The most recent report from Palamida indicates that open source companies are continuing to adopt GPLv3 at a rapid pace: over 2000 projects have adopted GPLv3. http://gpl3.blogspot.com/2008/03/gpl-project-watch-list-for-week-of-0328.html.
Palamida notes that: At this rate the GPL v3 is being adopted by 1000 projects every 4-5 months, and if the trend continues, the license will be used by 5000 projects by the end of the year. 5k will be a very substantial amount of projects under the GPL v3, which may influence larger projects to move over the the GPL v3 as well.