How to Succeed in Business…with Open Source SoftwareJennifer Kuan, Stanford University, USA
These questions concern not just open source entrepreneurs, but also software companies that compete with open source, hardware companies that see open source as a way to save money, engineering students who wonder they’ll make a living, and open source practitioners who have ranged from alarm to enthusiasm toward money-making ventures. It may also be important for policy makers to understand what consequences open source procurement or subsidy policies have on open and closed source software.
In my research, I have analyzed the competitive implications of open source software when open source development is driven by own-use. I have argued that individuals who contribute effort voluntarily to open source projects do so because they are users themselves; i.e., people fix bugs that they themselves need fixed. I argue that empirical evidence supports this “lead-user” or “consumer vertical integration” analysis. Specifically, the pattern of open source founding largely follows that which an own-use motivated development would predict; i.e. mostly highly technical software used by people with programming skills. Also, the quality of open source software is shown empirically to be superior to similar closed source programs, as predicted when users organize production around private information.
However, at least two extensions to this analysis come to mind in trying to understand the interaction between open and closed source software and the resulting new equilibrium. First, the variety of entrepreneurial efforts to exploit open source programs suggests that more research could be done to identify profit opportunities beyond the market segmentation of my analysis.and (I argue that because users produce, only software that is both neglected by closed source firms of interest to programmers will be produced open source. Thus, examples of open source programs: software that performs technical functions needed by programmers, such as programming tools, scripting languages, etc. Examples of non-open source: programs that everyone uses, like word processors or tax programs, and programs that do not interest programmers, like 3D spreadsheets).
Second, while I focus on selfish, own-use motives for contributing to open source, even my own data show that ideological motives can be important. I discuss a program called Gnome, which was founded in order “to level the playing field.” Gnome is a graphical user interface program designed to make Unix work more like Windows, and thus help Unix systems compete with Windows. I find that, unlike with own-use driven FreeBSD and Apache, Gnome’s bug response rate is slower than its closed source competitor. This isn’t surprising because the theoretical prediction of open source superiority over closed source derives from the own-use motive.
Nevertheless, it may be important to understand ideological drivers better. Sun Microsystems recently abandoned its own closed source graphical user interface program in favor of lower-quality, lower-cost Gnome. What will the resulting equilibrium quality of Unix-based user interface programs be? That depends on the objective of ideologically-driven open source developers. Is it to produce just enough quality to drive out closed source competitors and choice? If the product is free, that quality need not necessarily be very high to achieve this end.
This possibility that good intentions might inadvertently reduce quality and choice may merit some consideration. On the other hand, own-use driven programs most likely have a positive effect on equilibrium quality. Microsoft makes improvements to its software security in response to high quality open source operating systems. Also, open source programs now enjoy “voluntary contributions” from corporate users, like IBM, users who also have money to spend.
In short, money and profits may have an important role to play in an open source world. The interaction between profits and “nonprofit” open source software remain an interesting and understudied issue.