Thursday, July 30, 2009

Judgement Day for Commercial Open Source ... How Did I Miss It?

In a recent Business Week online article, Peter Yared, founder and CEO of San Francisco startup Transpond, describes "The Failure of Commercial Open Source Software". While Yared makes a case that the commercial open source software business hasn't lived up to the hype that it will "change the world," he reaches too far in declaring its failure. A more appropriate conclusion would be to recognize that commercial open source only recently reached broad acceptance as a business strategy and is on a strong growth trajectory. We must measure its success in that context.

Yared's arguments are diverse, but they fail to account for important indicators of open source success:

1. Minimal number of liquidity events for open source businesses. As identified by one of the commenters, several prominent examples of open source liquidation events are missing from Yared's list, including Zimbra, Trolltech and Sleepycat. In addition, the article uses 2003 as a benchmark year, allowing only 6 years to measure exit success. According to a September 2008 MoneyTree Report by PricewaterhouseCoopers and the National Venture Capital Association, the average seed-financing to exit life cycle of a venture-backed company is 8.6 years. As a result, any conclusions about the success or failure of open source businesses, many of which were started within the last 6 years, are premature at best.

2. Continued success of proprietary vendors. Not only is this irrelevant to measurement of open source success, but it fails to acknowledge the growing role of open source in proprietary companies. Companies like IBM, Microsoft, Adobe and Oracle, are milking revenue from their established proprietary business models while they also distribute open source software to generate revenue, influence development communities and drive broader adoption. In addition, measuring open source displacement of proprietary software misses the point, particularly in the short 6 year time frame here. Open source targets adoption opportunities through grass roots growth, which takes longer (but is less expensive) than the hard-hitting direct sales approach of proprietary companies.

3. Success is limited to commodity businesses; Enterprises are not likely to add open source businesses to their lists of approved vendors. These assertions fail to take into account the growing reach of open source throughout the software industry. While it is true that many of the more successful open source companies to date have been in the infrastructure and commodity business, we are beginning to see significant adoption of enterprise and end user open source applications (such as Alfresco, SugarCRM and others). Matt Asay further amplifies the limitations of Yared's assertion by identifying SpringSource as an example of an open source company that innovates and targets enterprise-friendly software development, which is outside the category of traditional commodity software.

4. Cost savings of open source are overstated. This contradicts the conventional wisdom that cost savings is one of the primary reasons companies adopt open source. In an April 2009 Forrester report, 75% of survey respondents said "Reduced IT costs" are critical or very important in their decision to use open source software. In addition, a panel of venture capitalists at OSCON proposed that open source companies can typically save up to 30% in sales and marketing costs as compared to proprietary companies, which leads to quicker profitability, quicker exits and happier investors. Even so, the article is likely correct with respect to mature open source businesses. In two March 2009 Open Sources blog posts, Savio Rodrigues compares the income statement of Red Hat to those of Microsoft and Tibco. In both cases, he concludes it is unlikely that mature open source vendors will be more capital efficient than commercial vendors.

5. SaaS will overtake open source. The general trend away from installed software applications to SaaS and cloud systems is undeniable. As Matt Asay explained on his Open Road blog in May, cloud computing is the natural conclusion of open source, because cloud computing is the ultimate expression of the open source principle that services, rather than the software supported by such services, are the most valuable component of a product offering. But SaaS and cloud business models are having significant growing pains of their own. Customers are concerned about the portability of data, freedom from vendor lock-in, security and standardization and other matters that must be resolved before achieving broader commercial acceptance. (It's a bit ironic that open source software might actually be the best way to address these concerns.) SaaS and cloud businesses appear to be subject to at least the same level of skepticism as open source. As a result, it seems unlikely that SaaS and cloud business models will replace the open source software business in the near future.

6. Open source benefits do not lead to monetary gain. Yared observes that his company uses open source but typically does not pay for it other than contributing code back to projects. This is a common practice and it implies an impending failure of commercial open source. Yared concludes that these factors do "not mean every successful open source project can sustain a commercial company, especially when they are delivering complicated applications rather than simple plumbing." No doubt this is true, but it is also true of proprietary business models and does not lead to the conclusion that commercial open source has failed.

It is fair to question when open source will become as reliable an investment as other technology businesses. In this regard, Yared's article raises some important points for discussion. However, rumors of the failure of open source business models are greatly exaggerated, and any pronouncement on the success or failure of open source is premature.

Wednesday, July 22, 2009

Show Me the Money at OSCON - Venture Capital and Open Source

To my pleasant surprise, Mark Radcliffe (notable open source attorney) announced last week on his blog that his law firm, DLA Piper, would host an open source legal track to run in parallel to the OSCON trade show in San Jose this week. The schedule included a diverse mix of topics that pushed the boundaries of typical legal presentations on the issue of open source. I found the talk on venture capital and open source the most interesting. The panel consisted of 3 seasoned venture capitalists (Josh Stein of Draper Fisher & Jurvetson, Mark Gorenberg of Hummer Winblad, and Vivek Mehra of August Capital) and an experienced open source attorney (Vicky Lee of DLA Piper).

Venture capital is a hot topic of discussion. Several tech publications and blogs are reporting, based on a report issued earlier this week by PricewaterhouseCoopers and the National Venture Capital Association, that venture investment has shrunk to pre-tech-bubble levels, the slow economy is putting a damper on venture capital, and venture funding in the tech industry is down over 50% year-over-year from 2008 to 2009. At the same time, however, the tech sector seems to be attracting more investment in recent months. The panelists at the OSCON session on "Understanding Venture Capital Investments in Open source Projects" seemed to have an enthusiastic view of investment in open source companies.

The panel described several elements that go into their determination of what is a good open source investment. The most fundamental points, however, came during the second session by Larry Augustin on "Choosing a License: Ensuring that Your Intellectual Property Strategy Matches Your Goals". Augustin emphasized that: (1) open source is only a tool and an open source business model alone is not enough to guarantee success or to warrant funding by venture capital firms; and (2) quality is critical to ensuring customers want to purchase a product regardless of whether it is open source or not.

Beyond these fundamental points, the panel discussed the following elements:

1. Successful Business Model

  • Copyright Ownership - This is an immense benefit because it provides maximum flexibility as a company grows.
  • Clear Distribution Rights - Ownership of all the copyrights in a product eliminates uncertainty about distribution rights. When third-party components are involved, however, clarity on rights is important.
  • Disruption, Platform and Infrastructure - Venture investments are most desirable in quality companies that focus on specific areas: (1) market disruptors in an existing industry; (2) technology that can become a platform; and (3) technology that focuses on infrastructure (because it is most likely to help companies reduce costs).
  • Free is a Negative - Serious customers do not find as much value in free products as products that they must pay for (Larry Augustin emphasized this point in the second session).
  • Support and Services - Companies that do not own their code and generate revenue only from services are susceptible to competitors.
  • Open Core - This is the current popular favorite of open source business models and it has a proven record of generating revenue, which is attractive to venture capitalists for obvious reasons.
  • Community - Open source products that spring from existing communities or that are developed in parallel with newly created communities can both be successful, but they present distinct tradeoffs. Existing communities provide an available user base that drives quick adoption, but the likelihood of multiple contributors complicates the intellectual property ownership issues.

2. Benefits to Venture Capital Investors

  • Lower Cost - While the first wave of financing is typically the same for open source and proprietary companies, the second wave often results in significant savings, which the panel estimated to be 30% or more. Specifically, the availability of open source software and related ecosystem allow open source companies to spend less on sales and marketing, which greatly reduces operating expenses.
  • Quicker Adoption - Because end users always have access to open source software, companies are able to avoid the lengthy proof-of-concept and testing cycles that occur when releasing a product for the first time. The panel felt this could save open source companies up to 2 years of development and testing time.
  • Easy to Build a Channel - Open source communities are well suited to perform localization, vertical market and other customization for products, which allows open source companies to enjoy the benefits of channel distribution without devoting sales and other resources.
  • Early Revenue and Exit - The lower cost and quicker adoption enjoyed by open source companies mean that these companies are likely to begin generating revenue and profits more quickly, which means that investors are likely to reach an exit opportunity more quickly too.

3. Net Results for Venture Capital Investors: When open source investment opportunities are carefully screened with the business model considerations above in mind, the benefits inherent in open source companies make these types of investments relatively safe for venture capitalists. At a minimum, investors should not view open source companies as being inherently more risky that other startup venture investments.

In closing, the panel indicated that approximately 15% of their holdings are in open source companies. While this is a significant amount of investment, the percentage is likely to grow over time as new software companies arise and existing ones revisit their business models. At the same time, the 85% of holdings that are not in open source companies is a good reminder that a good idea will attract investors because it's a good idea, not because it's open source.