The strategic trade-offs for beneficial open innovation: the case of "open source" consortia in mobile OS development

Three consortia dedicated to the provision of “open source” mobile Operating System (OS) for smartphones emerged in 2007 and 2008: the Symbian Foundation, the Open Handset Alliance and the LiMo Founda- tion, representing respectively the 1st, 3rd and 6th mobile OS in market shares in 20101. These consortia are particularly interesting when addressing the issue of ongoing architectural innovation (Henderson, Clark, 1990) in the mobile industry: the reconfiguration of the system linking existing digital services to fixed and mobile devices, thanks to the trivialisation of smart- phones. Consequently, we analyse the coherence for these consortia of using an open source framework, in order to perform the necessary reconfiguration at both the technological and organizational level. This question involves three corpora of the existing literature. First, at the technological level, the mobile OS is a complex system: a set of components with strong interdependencies, which “constrains the adapta- tive potential of systems, and, thereby, the possible paths of evolution” (Frenken, 2006, p. 3). In line with Simon (1962), one means of reducing complexity is to hierarchically decompose complex systems into subsystems. This approach is to be found in the transversal definition of platform made by Baldwin and Woodard (2009), which stems from a decomposition in terms of modularity (Baldwin and Clark, 2000), with an emphasis on technological hierarchy. It should be recalled that although platform architecture is still a Three consortia dedicated to the provision of “open source” mobile Operating System (OS) for smartphones emerged in 2007 and 2008: the Symbian Foundation, the Open Handset Alliance and the LiMo Founda- tion, representing respectively the 1st, 3rd and 6th mobile OS in market shares in 20101. These consortia are particularly interesting when addressing the issue of ongoing architectural innovation (Henderson, Clark, 1990) in the mobile industry: the reconfiguration of the system linking existing digital services to fixed and mobile devices, thanks to the trivialisation of smart- phones. Consequently, we analyse the coherence for these consortia of using an open source framework, in order to perform the necessary reconfiguration at both the technological and organizational level. This question involves three corpora of the existing literature. First, at the technological level, the mobile OS is a complex system: a set of components with strong interdependencies, which “constrains the adapta- tive potential of systems, and, thereby, the possible paths of evolution” (Frenken, 2006, p. 3). In line with Simon (1962), one means of reducing complexity is to hierarchically decompose complex systems into subsystems. This approach is to be found in the transversal definition of platform made by Baldwin and Woodard (2009), which stems from a decomposition in terms of modularity (Baldwin and Clark, 2000), with an emphasis on technological hierarchy. It should be recalled that although platform architecture is still a Three consortia dedicated to the provision of “open source” mobile Operating System (OS) for smartphones emerged in 2007 and 2008: the Symbian Foundation, the Open Handset Alliance and the LiMo Founda- tion, representing respectively the 1st, 3rd and 6th mobile OS in market shares in 20101. These consortia are particularly interesting when addressing the issue of ongoing architectural innovation (Henderson, Clark, 1990) in the mobile industry: the reconfiguration of the system linking existing digital services to fixed and mobile devices, thanks to the trivialisation of smart- phones. Consequently, we analyse the coherence for these consortia of using an open source framework, in order to perform the necessary reconfiguration at both the technological and organizational level. This question involves three corpora of the existing literature. First, at the technological level, the mobile OS is a complex system: a set of components with strong interdependencies, which “constrains the adapta- tive potential of systems, and, thereby, the possible paths of evolution” (Frenken, 2006, p. 3). In line with Simon (1962), one means of reducing complexity is to hierarchically decompose complex systems into subsystems. This approach is to be found in the transversal definition of platform made by Baldwin and Woodard (2009), which stems from a decomposition in terms of modularity (Baldwin and Clark, 2000), with an emphasis on technological hierarchy. It should be recalled that although platform architecture is still a modular system, it is itself split into a platform as core, together with its com- plements, interoperability being managed by interface specifications. At the organizational level, collaboration matters for the mobile industry, because such a complex system requires the coordination of innovations from heter- ogeneous technological fields (Kelly, 2006; Maula et al., 2006). According to Funk (2004, p. 202), “the conventional view is that the design of the architec- ture and the alliances that firms make with others are the critical issues in platform management.” So, in line with Le Masson et al. (2009, p. 290), we consider a general framework where “a collaborative process of platform design can actually be itself a specific platform; we shall call it a ‘platform for platform design’”. Hence, there is a real interest in analysing these processes at the systemic level, from the very earliest stage, all the more so as the literature has only just started to address this question (Le Masson et al., 2009; Maula et al., 2006; West, Wood, 2008). In other words, for Henderson and Clark (1990), architectural innovation requires reorganization and the acquisition of knowledge by firms. This approach presupposes, however, that the available elements are readily divisible, which requires a greater formalisation of activities (Gallouj, Wein- sten, 1997, p. 552), the more so as the degree of product decomposability depends on the existing state of knowledge (Buenstorf, 2005). Second, the characteristics of consortia lead us to operate a selection in the literature dealing with the effect of open source communities on indus- tries. The consortia – as sponsor-based, profit-motivated and involving firms – give echo to the idea of open innovation (e.g. Chesbrough, 2006), because of their open source strategy (West, Gallagher, 2006; West, Lakhani, 2008). Equally, we have to take into account the fact that sponsor-based communi- ties differ from autonomous ones (West, O’Mahony, 2008). We need, thus, a clarification of all these concepts: for Pénin (2008), the definition of open innovation does not take into account knowledge accessibility, which is criti- cal here. However, the consortia emerged as a transition from proprietary to open source models. For Bonaccorsi et al. (2006, p. 1094), firms tend to hybridize open and proprietary strategy rather than follow a pure model. This idea is consistent with West (2003, p. 1279): proprietary platform vendors’ strategy evolves from fully proprietary to open standards and, then, to open source, but the strategy is always hybrid (“opening parts”, “partly open”). A certain number of reasons encourage these sponsors to operate this transition: e.g., the market environment (West, 2003), developers’ preference (Sen et al., 2008), users’ involvement in innovation (von Hippel and von Krogh, 2003). Conversely, a transition to open source supposes an effort to redesign the product’s architecture (Baldwin, Clark, 2006; MacCormack et al., 2006). Third, to judge the coherence of this organizational strategy, we analyse the ability of these consortia to produce standards, because they encourage the provision of complements (Gawer, Henderson, 2007) and, more gener- ally, facilitate coordination (Simcoe, 2006). In line with Coris (2006), the consortia could encourage a hybrid form of standardization in which firms work collectively, rather than relying on the market to produce a de facto standardization subsequent to standards wars or, imposing a de jure standard- ization (upstream or downstream of the innovation diffusion) (Steinmuller, 2003) on the political and committee-based authorities. This form of stan- dardization supposes a serious involvement of the stakeholders in the interest of the consortium. But while Bonaccorsi et al. (2006) show that free riding is not rational (even in “hybrid” source projects), Shah (2006, p. 1011) shows that, in these projects, “activities that permit value appropriation by the firm are sometimes detrimental to value creation within the community”. So, it seems that organizational coherence – via the provision of standards – has equally to be adjusted via profit motivation. Clearly, our work is mainly of an exploratory nature, since the organiza- tional and technological platforms have not reached maturity. In the first part, we situate the consortia against a background of mobile OS platform competition. Then, we present an index of this openness, oriented on knowledge accessibility. To complete this conceptual framework, we analyse the standard creation issue as regards profit motivation coherence. In the second part, when we apply this framework to situating the three consortia in terms of an open innovation context, we note a relative difference among the consortia. We then show that the emergence of collective standards within the platform tends to remain uncertain, due to a divergence of con- cerns for knowledge sharing and a potential fragmentation for end-users. We conclude by discussing different speculative roadmaps for the emergence of collective standards among consortia.