Dancing with wolves: how value creation and value capture dynamics affect complementor participation in industry platforms

ABSTRACT

Industry platforms shape the interactions among participating firms, especially between the platform owner and its complementors. However, little research has examined how such dynamics would influence the initial participation by complementors. Based on the coopetition perspective, we develop hypotheses about the relationship between platform owner’s value appropriation potential and complementor participation, at platform initiation. Using a comprehensive dataset of open source software (OSS) platform initiations, this study empirically examines how the resource profiles of platform owner firms may affect complementor participation in the platform. We find that when the platform owner has higher percentage of sales from complementary market segments, the number of complementors for the platform is lower. Furthermore, the platform owner’s sales growth positively moderates this relationship. The findings shed new lights on how the value creation and value appropriation dynamics between the platform owner and its complementors may shape the outcome of industry platform launching.

KEYWORDS:

• Open platform
• industry platform
• platform complementor
• open source software (OSS)

Acknowledgments

We thank Prof. Joseph Mahoney, Prof. Yong Li, Prof. Douglas Miller for their great help in making this research paper a reality. We also thank Prof. Glenn Hoetker, Prof. Deepak Somaya, Prof. Kevin Chastagner, and attendees of macro research consortiums at HKUST and CUHK who provided insightful inputs for the improvement of this paper.

Notes

¹ http://opensource.org/osd.

² IBM News Release. 2001. ‘IBM Donates $40 Million of Software to Open Source Community’. November 5. https://www-03.ibm.com/press/us/en/pressrelease/1025.wss. Accessed on 21 May 2017.

³ See: https://www.gale.com/c/business-insights-global.

⁴ See: http://english.ckgsb.edu.cn/content/gale-business-insights-global.

⁵ The announcements included in our final sample result from the cross-checking of these three different databases. The concern over selection bias for one database can be mitigated by the consistent search results from three difference sources.

⁶ See: https://opensource.org/licenses.

⁷ We have also conducted robustness check with three-year time lag for our dependent variable, and the results are robust.