Open Source Vendors: Towards a Production Classification in Function of Firm-Community Relationship
Few days ago I joined the open source vendor debate remarking the importance of resolving the name confusion in favor of customers, and Jaspersoft CEO Brian Gentile shout-out saying that open core offers best opportunity for community and commercial Success.
Lines blurring between open source and proprietary vendors apparently invigorate the debate. I want to take the chance here to classify production model in function of the nature of firm-community relationship.
Defining corporate, voluntary, and hybrid organizational categories 6 years ago was important to start identifying open source projects’ organizational attributes.
The FS/OSS projects falling under the corporate category do not seem to have organizational attributes different from traditional proprietary projects. Indeed, the organization of work followed seems to be the same as that of a traditional firm. There is a well-defined hierarchy for the production process: there are a project manager, analysts, programmers, and end-users that within an organization are employed to – respectively – direct production and define, produce, and use the system. All stages of production are carried out by internal organization.
The difference between the voluntary and the hybrid idealtypes is not often considered. In the voluntary case the process of production is open for all contributors: deadlines and tasks may be assigned by the organization, but programmers are not completely bound to follow them.
In the hybrid case, volunteers from outside the organization as well as remunerated programmers contribute to the production process. The tasks and timeline are set by the organization and are binding for the remunerated programmers. When the volunteers do not respect deadlines or assignment of task, the remunerated programmers will. And when there are urgent necessities for, e.g., a patch, the hybrid organization may work in house even if this means duplication of volunteer efforts from outside. The American company Red Hat Enterprise is a good example of a hybrid.
Hybrid described how firms make use of open source communities. The firm-community relationship was then further investigated by Dahlander and Magnusson distinguishing symbiotic, commensalistic, and parasitic approaches.
The symbiotic approach is the most promising. It can bring a vendor straight to the top of the five stages of community open source engagement, as designed by Ian Skerrett. Actors like GroundWork or Google, as recently pointed out by Stephen O’Grady, are already doing it.
The production model affects at large the software life-cycle, resulting in different core capabilities and configuration of activities, namely two of the nine building blocks used to describe an (open source) business model. The result is that a vendor can make R&D savings just standing on the shoulders of open source giants. Moreover, an effective strategy addresses also other issues like building a community and attracting external contributors changes, as originally analyzed by West and O’Mahony.
Open core vendors, i.e. firms spinning out internally developed code, have to face managerial issues related to decision rights and control, since transparency and accessibility affect external participations. With open core vendors copyright and control are in one hand, as a result all kind of commercial services are available by the vendor, but not only.
Brian Gentile disagrees with whom says that the only true open source vendor is who provide identical community and commercial editions, arguing that this argument miss history lesson. I agree. At least as far as open source vendors allow sub-projects creations and/or the right to directly commit software changes on the public repository is allowed.
Pure open source vendors, like the above mentioned Red Hat, have also to address important issues like influencing the community, and can be distinguished on the basis of how they rely on open source external knowledge (accessing, aligning or assimilating it). All in all legitimacy has to be gained or retained.
But what customers expect? I agree with David Rosenberg, saying:
Open-source companies have to become models of efficiency in every area in order to keep costs down and revenues up. Customers expect open-source alternatives to be 10 percent to 20 percent of the cost of the proprietary product, which means that open-source companies need to be 80 percent to 90 percent more capital-efficient.
Open source vendors, core or not have to go beyond the marketing hype to be so efficient.
Sustainable open source ecosystems are hybrid, and some are more sustainable than others. Next time I will cover differences between projects sponsored by vendors, foundations and technological clubs.