NISTIR 7323 - The Determinants of Success in R&D Alliances
Part 6 - Conclusion
This study examines a complete set of alliance design factors and alliance management factors that may influence R&D alliance success. Relative to the hypotheses presented above, we derive the following conclusions from our empirical analysis:
- Number of alliance partners. We find that the number of alliance partners has a
weakly negative effect on patent application, and no effect on overall value or com
mercialization. The evidence suggests that alliance designers are largely successful in
achieving an "optimal" number of alliance partners, balancing the marginal benefit of
adding an additional partner to the marginal cost associated with an additional partner.
- Presence of competitors. Simple correlations show that alliances with competitor firms
have less frequent communication and lower levels of goodwill trust. But in multiple
regression analysis we find that the presence of competitors in an alliance has no effect
on alliance outcomes. Again, this suggests that alliance designers are largely successful
in "optimizing" the structure of the alliance.
- Geographic distance between alliance partners. Simple correlations show that alliances
with greater geographic distance between partners have less joint work interaction,
and lower levels of governance effectiveness and goodwill trust. But in multiple regres
sion analysis, we do not find that geographic distance has any effect on alliance out
comes. The simple correlations also show that frequency of communication between
alliance partners has no correlation with geographic distance, so it appears that geo
graphic distance does not affect the communication, knowledge sharing, and coordina
tion necessary for successful alliance outcomes.
- General and partner-specific alliance experience. We find that general alliance experi
ence and partner-specific experience have a negative effect on patent application as a
performance outcome. This may suggest that creativity and invention are more likely
when new partners come together in new collaborations to combine ideas, different
approaches, and complementary knowledge.
- R&D capability. We find that R&D employment at the company location has a nega
tive effect on overall value as a performance outcome, and no effect on either the
patent application or commercialization measures. We use R&D employment as a
proxy measure of the R&D capability of the firm. One explanation for the finding of negative or no effect of R&D employment on outcomes is that R&D employment is also a measure of firm size, and R&D personnel in large firms have greater institutional hurdles to overcome in order to realize the potential of their research results.
- Number of technical personnel. We find that the number of technical personnel
resources has a positive effect on patent application and on commercialization of technology. This result supports the view that innovation and learning outcomes depend
on the number of technical personnel engaged in the effort.
- Frequency of communication. We find that frequency of communication has a strong
statistically significant positive effect on all three outcomes measures for R&D alliance
performance. The finding suggests that alliance managers can increase the likelihood
of alliance success by establishing routines that encourage frequent communication.
Frequent communication facilitates the knowledge sharing and coordination that is
critical to alliance success.
- Governance arrangements. We find that effective governance arrangements have a
positive effect on the perceptual measure of overall value, and a weakly positive effect
on the patent application measure. This finding highlights the importance of establishing
contractual provisions and governance procedures as a factor in alliance success,
for example, in protecting intellectual property rights, monitoring task performance,
and resolving disputes.
- Goodwill trust. We find that goodwill trust among alliance partners has a weakly neg
ative effect on the perceptual measure of overall value, and a negative effect on the
patent application measure. The results suggest that successful alliances do not depend
only on goodwill trust, but develop contractual-based trust that is generated through
contractual provisions and effective governance arrangements. In other words, in
alliance management, one would do well to "Trust, but verify."
Finally, we find that "reaching for the stars" is a strong predictor of alliance success. More ambitious projects with farther reaching goals demonstrate greater success on all three measures of R&D alliance performance. More ambitious projects have intrinsically greater potential value and potential impact. In addition, more ambitious projects are likely to mobilize greater commitment and effort on the part of both the partner companies and the individual participants.
For strategy and innovation scholars, this study demonstrates the usefulness of employing multiple measures of success in the analysis of R&D alliances. Each type of measure—perceptual measure of overall value, patent application, and financial value from commercialization—has drawbacks, but by considering them together we are able to gain a richer understanding of the factors that explain R&D alliance success. Researchers may gain additional insight into the determinants of alliance success by analyzing which factors are common to various outcome measures, and which factors influence only a specific outcome. Some factors, such as prior experience or goodwill trust, may well be determinants of success in some types of alliances, such as manufacturing or marketing alliances, but not in other types of alliances, such as R&D alliances where creativity and innovation are central. Our research suggests that there are fundamental differences between R&D alliances and other types of alliances, and consequently, the factors that influence success may be quite different. For example, R&D alliances involve high risk and uncertainty as a fundamental characteristic, and so there are inherent limitations to applying past experience in future projects. Though many research studies combine R&D alliances and other types of alliances as a single subject for analysis, we conclude that R&D alliances should be treated as a separate and independent type of collaborative activity.
For companies engaging in R&D alliances, and government agencies that support collaborative R&D, this study shows that alliance management is important for alliance success. While outcomes of research projects are inherently uncertain and determined in part by chance, there are managerial factors which can increase the odds of success. In particular, managers can increase the likelihood of success by: (a) carefully considering the composition and structure of alliances, in terms of number of partners, involvement of competitors, attributes of alliance partners, etc.; (b) establishing effective contractual provisions and governance procedures; and (c) managing R&D alliances to ensure frequent communication among participants.
This study includes only government-sponsored R&D alliances in the United States. One might ask whether the results of this study generalize to other R&D alliances that do not have government support, or to R&D alliances in different institutional settings. In order to evaluate whether findings from this study also hold for privately-funded R&D alliances or R&D alliances in other settings, further research in broader samples may be conducted.
Return to Table of Contents or go to next section of interim report.
Date created: August 29, 2006
Last updated:
September 13, 2006
|
