Social media conv. wisdom is to start out really open, but that assumes new space. #buzz is an add to old space. Rules change. #google

And a friend of mine commented on Facebook (I’ve setup my Twitter feed to update my Facebook status):

Doesn’t it seem like Google is copying facebook’s status updates? I don’t see what the big deal is.

I was trying to make a more salient point, which is that Gmail and Facebook are vastly different social spaces. Facebook is the “new space” where the purpose is originally and has always been for social networking. Gmail is traditionally the space for email, which for some people overlaps with “social networking” but for many others do not. As such, Google, by introducing functionality that is more effective when open into a system that is traditionally more closed, is bound to cause conflicts, and users (not the technorati, but normal people) will get confused. In order to address this, they have to abide by the rules of the old territory so to speak, which is to start closed and then open up, which curtails adoption momentum. However, this is offset by broader reach to an already installed base. On the other hand, Facebook started out being a really open space and then only progressively introduced privacy controls. Google by introducing similar functionality into Gmail does not have the luxury of progressively adding privacy controls, but has to have it from day one in order not to confuse and alarm normal users. In a way it’s ironic that Google in this case seems not to be the disruptive innovator but instead the incumbent (reference “The Innovator’s Dilemma” Clayton M. Christensen).

I list below some common arguments against venture capital mixing with open source and the counter-arguments to each of them. You as the reader can make up your own mind.

As for myself, I remain very neutral to the issue – meaning that I believe that at the present moment there is only a minority of venture capital firms that can work effectively with open source companies, and I hope this will change in the future in a way that benefits both open source and venture capital. Now that I’ve stated my position, the arguments…

Argument 1

Venture capital is obsessed with the possibility of exit, which may not be good for open source after all, especially if the exit is a sale to a proprietary software company (like what happened to MySQL inadvertently through Sun?). And frankly, it’s not necessarily good for innovation (which I believe is the real goal of venture capital, not profits ).  See Fork Well: It Could Be The Last, Best Hope for Community for a “free software centric” but very understandable reaction early on in the Oracle-Sun merger news cycle.

Counter-argument 1

There are many possible good investment exits for open source companies that are good for open source in general, such as a merger with or an acquisition by a larger and more influential open source player (e.g. Red Hat), or an IPO. If companies like Novell who are increasing their open source portfolio become more oriented to open source and successful at the same time, it can also be beneficial. And in the Oracle case, there are others that have argued that they have supported many other open source projects after acquisition, such as JBoss and BerkeleyDB (but admittedly it is difficult to analyze what would have been if these stayed independent or were acquired by other organizations instead).

Argument 2

The whole venture capital industry is based on valuations through traditional corporate measures, which is ill-equipped to value open source communities. In fact, it can be argued that the open source communities transcend corporate boundaries and therefore company-based investment simply is somewhat antithetical to the structure. It’s like fitting a square peg into a round hole.

Counter-argument 2

The venture capital industry is in a state of turmoil (CNET News: Big changes needed in the venture capital market), and coping with the nature of open source is part of the transformation that it needs to have in order to reinvent itself. Part of that is finding new models for valuation, adjusting to different investment timeframes, and part of that might involve some drastic changes (which are coming).

Argument 3

The perceived legal risk around commercializing on top of open source licensed technology scares many venture capitalists.

Counter-argument 3

Venture capitalists and their lawyers are gradually understanding that not all open source licenses are the same, and that the legal text together with an understanding of the community reflects a clear enough line (and therefore measurable risk) that is drawn between what should be “community property” and what falls within the prerogative of the companies operating in the community. With more information, entrepreneurs that are cognizant of these constraints (which are in fact strengths if the business plan is sound) should be able to convince potential investors that risks are mitigated.

Argument 4

Venture capital investment could crowd out and discourage community involvement. There could be many reasons for this but in general it might have the potential of disrupting the well-functioning status quo within an existing community which may reduce the contribution of other members who feel that they benefit less from the investment that the company directly being invested in. There may also be a perception of conflict of interest affecting the judgement of community leaders who are founders, employees or significant shareholders of the company receiving venture capital.

Counter-argument 4

Founders, employees or significant shareholders of the company receiving venture capital can “walk-the-talk” and give back generously to the community. In fact, they must, if their business is to be successful. Venture capital has been shown in many cases to have resulted in successful open source companies that are good for open source. After all, venture capital was invested in Red Hat, Zimbra, Xensource, and many others.

On the other hand, many venture capital companies have traditionally prided themselves of being actively involved in adding value (not just financing) to the companies that they invest in. How many venture capital companies invested in open source companies can say the same in terms of their role in helping the community? But I suspect that this again is more of a “lack of understanding” problem and not necessarily the fundamental conflict of interest that many perceive. Things might be better as more and more successful open source entrepreneurs/investors gain experience. Some organizations that clearly benefit a lot from open source clearly have more role to play, say (possibly controversial), e.g. Google Ventures?

For more information:

1. Remarks by Matt Asay, one of the influential people in open source today in an interview about how open source and commercialization mix: http://www.infoworld.com/d/developer-world/open-source-roundtable-matt-asay-176?r=54
2. Open-source VC investments: Time for payback (post on The Open Road blog at CNET) http://news.cnet.com/8301-13505_3-10217824-16.html
3. Report of venture capital investment in open source related firms from 1997 to 2008: http://blogs.the451group.com/opensource/2009/04/08/the-past-present-and-future-of-vc-investment-in-open-source/

And as a postword, perhaps one day open source may not need venture capital as we know it today, but then maybe venture capital may not be what it is today tomorrow anyway, at least for open source.

Should government intervene? Incumbent companies in Canada charge $30 to $60 a month for broadband services. There are sound arguments that prices could fall with more competition as there are two or at most three firms providing broadband access to most locations in Canada, a market characterized by very high concentration.

Government should do something, but should avoid trying to get into providing services themselves. There are a lot of inefficient Wi-Fi business models that exist, one of the worst of which is to aim for full coverage, which a municipality is likely to pursue due to political necessity. Providing across-the-board municipal Wi-Fi is a broad brush that will create wasteful overlapping infrastructure. In fact, customers tend to demand Wi-Fi to be provided at hubs (social environments?) such as cafes.

Government supported broadband Internet access is already widely available in schools, libraries, community centers, and other specialized locations such as employment service agencies. Any proposed investment to bridge the digital divide should be focused on strengthening the services provided through these channels rather than to build and operate a city-wide service.

Moreover, sociopreneurs from the private sector are better equipped to provide low-cost Wi-Fi than the government – the Ile Sans Fil Wi-Fi service which is supported by volunteer resources and revenue from hotspot locations such as cafes and restaurants in Montreal is extremely successful. Municipalities can support such organizations through existing channels such as volunteer and community support agencies.

The general experience of regulation in the telecommunications sector is that it has stifled innovation. However, this is not to say that all regulation curbs innovation (Anthony, Roth & Christensen, 2002, link to pdf). In a situation where there are few companies providing broadband services to the market, theory tells us that this lack of competition leads to higher prices, and less rapid innovation by incumbents. We note that there is no shortage of motivation by would-be private-sector entrepreneurs to start-up Wi-Fi services. However, their ability to achieve this is hindered by their ultimate dependency on the telecommunication companies to carry their back-haul traffic.

One problem they face is incumbent terms of service that potentially abuse dominant market power. For example, some terms for DSL or cable Internet service prohibit commercial use, forcing Wi-Fi entrepreneurs to purchase less cost effective services based on older leased line technologies. Another example bans residential end-users from sharing DSL or cable connections with neighbors using a Wi-Fi router. This policy biases against small apartment units, and creates an unequal burden on lower-income households. Many Wi-Fi startups are small companies and may not be able to co-ordinate themselves to bring forward an anti-competitive legal case against any of the incumbents without help. The federal government can assist in providing a platform to alleviate these concerns, releasing true points of modularity by taking down barriers.

The federal government could also support advancement in Wi-Fi based technologies by releasing more unlicensed spectrum, or by providing at low-cost spectrum demarcated for Wi-Fi type services only. Is spectrum allocation currently biased against Wi-Fi? Government should have a hard look at their spectrum allocation policies (and internal motivations). Theory suggests that there might be a innovation gains from encouraging the development of Wi-Fi as it could be disruptive to incumbent broadband providers’ markets characterized by customers who are over-served from a technical perspective, but under-served in terms of quality of customer service.

In summary, government should focus on enhancing the ability of entrants, and avoid programs that require detailed selection of private sector partners to implement municipality funded projects. These programs will only distort entrant motivations, undermining entrepreneurial intentions that already exist in the industry.