Comparación critica con la literatura existente

Organizational pressures impacting innovation practice transfer cited by respondents primarily involved the speed of the innovation cycle, and the structure of the organization, especially that of the innovation network. MNEs which operated in faster cycle industries, such as Software, acknowledged that when establishing a new subsidiary, it was difficult to transfer a standard set of innovation practices, as these were constantly evolving:

The development of technology is an intensely, quickly moving process.

The way you develop technology now may be very different in two year‟s time, and setting up a lab can be a long term process in itself. By the time you set the lab up, the technology that you‟re using and developing on may well have changed itself. We quite regularly change how the developer‟s code, and the technology that they use. So you can‟t plan this year for something that‟s going to happen in 2013.

Given the nature of this industry, central administration of practices is required, but constantly challenged as technology and software development techniques change frequently. Hence, a degree of organic diffusion may be desirable, as it is more efficient.

However, this may also lead to many adaptations of practices, and hence inconsistencies throughout the innovation network.

While the structural elements of integration and specialization were not specifically attributed as the cause of adaptations within each MNE, some evidence was nonetheless inferred. At Aero, for example, the degree of integration amongst

subsidiaries is quite high, as each project is carried out at multiple sites. As evidenced by the fact that adaptations to the ERP system requested by Aero Latin America had to be reflected throughout the overall organization, this integration increased resistance to adaptations of the technical dimension of practices:

We were told by the SAP implementation team 90% of the system is standard, your 10 % is going to fit your requirements. In some instances we went against the standard because it didn‟t fit the business requirements of the local unit and we had to justify to the implementation team. [They told us] this is going to impact the operations of all sites if we decide to go this way. In the end, the business case was approved and we did not follow the standards.

Telecom, according to one manager in the US, had become less centralized over time as it grew, and made many acquisitions. Efforts were being made to integrate different subsidiaries, but due to the acquisition strategy used, there remained a fair degree of separation amongst them. As a result, there were far fewer standard innovation practices as might be observed in Aero. This lack of integration made driving common practice challenging:

Definitely there are different ways of doing things. We are still a million-dollar start-up, so there's entrepreneurial ways of doing things all over the place, but we're trying to wrangle that in a little bit.

(Senior Program Mgmt. Office Manager, Telecom Co., US)

Likewise, although projects involved multiple sites at Software, the extent of integration amongst the different subsidiaries was limited by the software development process and federated structure. At Software, although the subsidiaries were not as integrated as those of Aero, major global changes in software development practices were driven by headquarters, which essentially transferred the conceptual aspect of this development paradigm, leaving details of the social and technical implementation up to each subsidiary. This offers further support for the negative relationship between integration and technical adaptation of innovation practices, and further suggests that integration is negatively related to adaptations of the social dimension of the practice:

It could become bottom up, but [an initiative like lean development]

initiative has to come top down. If you want to do it around the world it definitely has to come top down, so that everybody in all the locations would sing from the same sheet.

(Lab Head, Software Co., Canada)

Finally, at Telecom, ideas are shared in a competitive format. Different subsidiaries, which operate with a fair degree of autonomy and have limited shared history and context, propose different ways of doing things and the „best‟ are chosen.

However, as each idea is likely to be interpreted within the context of a particular subsidiary‟s mandate and history, the conceptual dimension is left open for interpretation.

Hence, like the social and technical dimensions, the conceptual dimension of a practice is likely to be more highly adapted when subsidiaries are less integrated within the MNE

network. This latter point is also consistent with the finding that conceptual adaptations are made very early during the transfer process (Yu and Zaheer, 2010).

Proposition 8. Integration between innovation practice providers and recipients result in lower levels of adaptation of the technical, social and conceptual dimension of transferred practices.

Although highly integrated, the subsidiaries at Aero are also somewhat specialized in that each has a particular area in which they focus their capability development. For example, flight controls are a specialty of the Eastern European unit while India is the primary test site outside of the US. However, the degree of specialization is not as high as in the case of the research labs of IT Services Co. which each have their unique “character” defined by the knowledge and venture capital resources in their location. This character is further made unique by the fact that they are often managed by an individual with a unique set of capabilities and expertise.

It appears that higher degrees of specialization limit the extent to which common innovation practices can be implemented across different subsidiaries. At IT services, the practices which are implemented tend to involve the operational and reporting aspects surrounding how a lab is run. These practices are negotiated by the technology board which includes the heads of each lab. Once the practices are transferred to the labs, there are substantial pressures for adaptation along technical and social lines. Hence, adaptations to the conceptual dimension of the practice occur prior to transfer, and due to

high degrees of specialization in the labs, there is substantial adaptations made to the technical and social dimensions.

At Aero, limited specialization amongst different subsidiaries results in limited levels of adaptation to innovation practices. Hence, specialization tends to increase the extent to which the technical and social dimensions of each innovation practice are adapted, but may have no effect on the conceptual dimension. As alluded to earlier, once practice transfer has commenced, it is more likely that the technical and social dimensions will face adaptation pressures than the conceptual dimension. However, if a practice is designed to address a certain aspect of innovation, such as how the capabilities of each lab are assessed, then adaptations to the technical dimension can lead to difficulties in interpretation. Overall, the conceptual dimension of practices does not appear to come under adaptation pressure as a direct result of specialization.

Proposition 9. Specialization between innovation practice providers and recipients result in increases in adaptation to the technical and social dimension of transferred practices.