Two dilemmas and one trap with open innovation

In closed innovation (CI), each firm does R&D on its own; however, on the other hand, in open innovation (OI), companies use knowledge from external sources when necessary or let other firms use their proprietary technologies. Chesbrough (2003) pointed out the effectiveness of OI. OI is accompanied by (1) outsourcing dilemma that prevents a firm from getting a competitive advantage when rival firms can also gain the knowledge from same external sources, (2) integrator’s dilemma, which occur when firms sell parts to assemblers that produce and sell the same product, thus becoming formidable rivals, and (3) modularity trap, whereby when it is necessary to redesign the total product structure due to radical product innovations, the firm which selected open modular architecture and OI cannot respond because knowledge has dispersed among firms. OI is not necessarily the most efficient choice when these dilemmas or traps exist.


Introduction
points out the effectiveness of open innovation (OI). Comparison with closed innovation (CI), in which each firm conducts R&D on its own, OI can add to the firm's performance by combining internal and external knowledge, for example, using knowledge from external sources when necessary or promoting other firms to use their proprietary technologies.
Although Chesbrough (2003) showed the effectiveness of OI in the high-tech industry, it was also pointed out that OI is not always effective in the high-tech industry or in industries with similar circumstances. This paper discusses the outsourcing dilemma of Takeishi (2003), the integrator's dilemma of Sakakibara (2005), and the modularity trap of Kusunoki and Chesbrough (2001), and further proposes some issues that will have to be resolved for OI to function effectively.

Outsourcing Dilemma Associated with OI
Chapter 5 of Chesbrough (2003)  increase the product's value. However, we need to bear in mind that these studies have indicated that the following problems can occur.
(a) Using parts or technologies that are easy to purchase in market is linked to a decline in a product's competitiveness, as the product becomes less rare and its imitability goes up (Barney, 1997).
(b) If a part or technology is essential, the bargaining power of suppliers will increase and the product will be deprived of most of its added value (Porter, 1980).
As Takeishi (2003) proposes with respect to outsourcing dilemma, there are both positive and negative aspects to utilizing external knowledge through outsourcing, as in OI. Neither case studies nor theories can determine whether CI or OI with outsourcing is more efficient.
Overcoming the outsourcing dilemma requires that (i) the firm produces in-house parts that have the most impact on the quality of the final product or that (ii) in the case of a product assembled with parts procured from external sources, the product should be enhanced with internal resources such as design capabilities and/or production know-how that increases its value (Takeishi, 2003).

Integrator's Dilemma Associated with OI
In discussing the case of IBM, Chesbrough (2003) mentions that IBM decided to supply other firms with its hard disks for laptop PC, and this is regarded as an OI success story of providing technology to external firms. Although selling parts to external firms makes it more difficult to differentiate a firm's products from the products of other firms that are using its parts, Chesbrough points out that economies of scale in parts manufacturing outweigh the demerit.
To counter this assumption, we present the case of Japanese watch makers selling parts to external firms as evidence to the contrary. Seiko Epson and Citizen Watch among others in the watch industry sell watch movements 1 to other firms. By using these movements, Chinese assemblers can make cheap watches that have the exact same degree of precision as products made by Seiko Epson and Citizen Watch, so that they have become formidable rivals to those two Japanese companies. Moreover, the movement business has become commoditized and profitability has declined as a result of mass production, mass marketing, and price competition between Seiko Epson and Citizen Watch (Sakakibara, 2005). In response, Seiko Epson has moved from OI to CI by developing a movement that uses proprietary technology known as a spring drive, which is being used exclusively in its high-end watch models. Sakakibara (2005) argues that for assemblers involved in producing parts and assembling finished goods, selling parts externally has both merits and demerits, which presents an integrator's dilemma. Like OI that employs outsourcing, OI that sells parts to external firms does not necessarily improve a firm's competitive advantage or its performance.
For OI to increase profits by selling parts to outside firms, there needs to be additional conditions, such as manufacturing competitive parts that cannot be imitated by other companies, or acquiring the de facto standard. Chesbrough (2003) holds that the shift to OI will continue as integral architecture changes over to modular architecture and as external outsourcing and sales of modules become possible.

Modularity Trap with OI
However, Kusunoki and Chesbrough (2001) points out that a virtual organization that uses modular architecture to divide work among multiple firms will fall into a modularity trap as it will be difficult for the organization to come up with innovations that redesign the product's overall structure. Kusunoki Kusunoki and Chesbrough (2001) points out a dynamic shift of product architecture that alternates between times when integral architecture and modular architecture are effective. In dealing with this dynamic shift, CI is effective when there has been a shift toward integral architecture and the total product innovation is required, so that vertical integration or close collaboration and coordination among firms is efficient. Subsequently, OI becomes effective when there is a shift toward modular architecture when vertical disintegration occurs, which enables innovations that are specialized for each part. However, the analysis in Chesbrough (2003) is incomplete in displaying the effectiveness of OI as it focuses on vertically-integrated firms and only analyzes shifts from CI to OI.
Premised on a dynamic shift of product architecture, companies that have been successful with OI in modular architecture are also required to encompass the capability to escape from the modularity trap, such as moving toward vertical integration or building close collaboration between companies, for redesigning the entire product (Nakagawa, 2006;Wi, 2002). If companies do not possess this capability, even if they have been temporarily successful with OI, they will lose their competitive advantage when the shift to integral architecture occurs.

Discussion
Beside issues posed by these two dilemmas and one trap, it can also be pointed out that there is a problem with the analysis unit in Chesbrough (2003). Vanhaverbeke (2006) points out that the unit of analysis for OI is the firm and the discussion on OI emphasizes close ties and a network among firms. If an innovation involves the participation of several firms, even if capital ties and personnel exchanges will exist among those firms, or product development occurs through integral architecture that involves close coordination, this will be considered OI, not CI. Kusunoki and Chesbrough (2001) analyzed changes in patterns of innovation from the perspective of product architecture. This enabled the authors to look at substantive activities to ascertain parts that were being coordinated, the scope of integration of relevant knowledge, and changes in the scope of coordination in product development activities. On the other hand, Chesbrough (2003) analysis is based on firm, which is an institutional unit, and it is not possible to understand the activities in practice. It seems that this led to a surge in cases that were CI in substance but were classified as OI.
As pointed out in Takahashi (2014), the firm as an institution is merely the concept of a boundary or partition. To ascertain the scope of knowledge integration and the scope of coordination in product development activities, should we define whether they are CI or OI, depending on whether the participating firms are in close coordination on a set basis? 2 In other words, it seems that innovations taking place internally in a transfirm organization theory in Takahashi's study should be defined as CI and those that are not should be defined as OI.
2 Although they do not include the term "OI," Kokuryou (1995Kokuryou ( , 1999 are Japanese studies discussing the merits of being open from the perspective of business networks. These studies define the formalized business relationships like those of the Japanese corporate keiretsu as kakoikomigata (closed-type) management. In contrast, open-type management is defined as a situation in which knowledge can be freely consolidated internally and externally via modularization.