Inter-organizational governance and trilateral trust building: a case study of crowdsourcing-based open innovation in China
- 365 Downloads
In a case study of a Chinese crowdsourcing intermediary, we explore the impact of inter-organizational governance on trilateral trust-building. We show that formal control and relational governance mechanisms are essential for swift and knowledge-based trust in R&D crowdsourcing. The case also indicates that Chinese businesses continue to use guanxi (informal personal connections) as a relational and contingent mechanism to maintain affect-based trust, but guanxi is shown to inhibit the growth of Internet-based crowdsourcing for open innovation in China.
KeywordsGuanxi Chinese context Crowdsourcing Open innovation R&D Trust
Scientists and organizations are increasingly and broadly collaborating in engineering, computer sciences, chemistry, life sciences, and physical sciences research and development (R&D) projects through crowdsourcing-intermediary (CI) platforms such as innocentive.com, innovationexchange.com, and yourcencore.com (Chesbrough 2006). Companies are now widely accepting R&D crowdsourcing as a primary means for reducing the costs of R&D for innovation needs (West and Bogers 2014). As China steps into the Internet era, many e-commerce players, Internet users and websites have emerged to perform tasks such as logo and website designs and translation services. Although China’s rapid economic growth has increased the demand for R&D, Internet-based open-innovation platforms present a dual challenge (Gassmann and Han 2004). Most small-and-medium enterprises (SMEs) lack internal R&D workforces, but China has a large pool of talent being able to conduct essential R&D projects. However, China’s companies and research institutions resist placing high-value R&D projects on crowdsourcing websites, and few researchers have systematically addressed how such resistance might affect the growth of R&D crowdsourcing for innovation in the Chinese context.
Crowdsourcing is an effective approach to innovation, but to avoid exchange hazards, build trust in online environments, and form customer–client relationships, crowdsourcing sites need formal controls such as written contracts, feedback, escrow services, and credit-card guarantees (Ulset 1996; Willamson 1985). They also need relational governance emphasizing mutual trust, relational capital, communication, and collaboration (Dyer and Singh 1998; Lee and Cavusgil 2006). Inter-organizational trust is essential for business transactions and building successful long-term relationships and alliances among companies (Zaheer et al. 1998; Zaheer and Venkatraman 1995). The extant literature mainly focuses on trust-building in vendor–client or alliance bilateral relationships; few studies examine trust-building in trilateral relationships comprising intermediaries that facilitate activities between two companies (Gefen and Straub 2004; Pavlou and Gefen 2004; Stewart 2003). Because Internet intermediaries replace traditional agents in connecting geographically separated vendors and clients, we lack sufficient knowledge about trust development in the virtual world among clients, online intermediaries, and vendors in the context of crowdsourcing-based open innovation.
Drawing from the literature on inter-organizational governance and trust development, we address a research question: how can inter-organizational governance mechanisms be adapted to enhance trilateral trust-building in crowdsourcing-based innovation in China? By focusing on the Chinese context and by using a case study, we provide additional insights into how Chinese firms utilize inter-organizational governance mechanisms for open innovation. We find that Chinese CI differs from western practices, where CI is an open platform for inter-organizational exchanges to accomplish R&D crowdsourcing tasks. Instead, Chinese CIs more intensively govern crowdsourcing processes, trilateral interactions, and trust-building between users, but lack proper formal control and relational mechanisms to manage swift and knowledge-based trust. More importantly, we find that Chinese CIs utilize guanxi, or informal personal connections, to maintain affect-based trust. Based on the case findings, we propose that guanxi-based relational governance challenges open innovation and prevents organizations from utilizing online crowdsourcing for R&D projects.
This study makes several important theoretical contributions to the open-innovation literature. Foremost are theoretical contributions that incorporate the literature of inter-organizational governance and trust development. Our study underscores that CIs must have appropriate governance mechanisms to develop swift and knowledge-based trust among crowdsourcers, crowdsourcees, and CIs. We also contribute to the literature of open innovation by identifying effective governance mechanisms specific to the Chinese context (Byrum and Bingham 2016). Second, extending the inter-organizational governance literature and e-commerce trust concept (Ba and Pavlou 2002; Benbasat et al. 2008) to the crowdsourcing context, we contribute to understandings of trust-building in trilateral business relationships (Svensson 2004) and show that swift and knowledge-based trust is essential if crowdsourcing is to grow. Third, we offer additional understanding of guanxi’s (Chen et al. 2013) double-edged functions in governing relations, but hindering the growth of crowdsourcing-based open innovation in China.
The paper is organized as follows: First, we discuss the related literature highlighting the roles of crowdsourcing-based open innovation. Then we discuss inter-organizational governance and trust development in the crowdsourcing environment. We introduce our research context and methods, followed by our main findings and discussion.
Crowdsourcing-based open innovation
Open innovation is defined as an approach that firms use to acquire external knowledge across industries to complement their internal knowledge base, to advance their technology, and to acquire competitive advantage (Chesbrough 2003, 2006). Firms often use internal employees or the community for open innovation, but most frequently rely on online crowdsourcing for sundry purposes (Cabiddu et al. 2013; Sieg et al. 2010) from simple letter-writing or graphic-design projects to complex R&D innovation (Andriole 2010; Erickson 2012).
Online crowdsourcing-website stakeholders include crowdsourcers, crowdsourcees, and crowdsourcing intermediaries (CIs) (Brabham 2008; Cabiddu et al. 2013; Majchrzak and Malhotra 2013). Crowdsourcers initiate the crowdsourcing process by submitting task requests and specifying the acceptance criteria. Crowdsourcees are registered members who are paid by clients to undertake and execute tasks. CIs are the online brokerage companies acting as intermediaries or virtual knowledge-brokers (Frey et al. 2011; Howells 2006; Katzy et al. 2013; Verona et al. 2006), ensuring that crowdsourcees successfully complete projects and that crowdsourcers pay for the services (Zogaj et al. 2014).
Because CIs are increasingly important for driving open innovation, management scholars have turned attention to research on crowdsourcing (Chesbrough 2006) to learn motivations for participating (Frey et al. 2011), the design of open-innovation contests (Cabiddu et al. 2013; Terwiesch and Xu 2008), managerial challenges and remedies for solving R&D problems through innovation intermediaries (Sieg et al. 2010), product-solution capabilities and added value for product-development processes (Tran et al. 2011), and ways to attract participants who can substantially contribute to innovation challenges (Frey et al. 2011). Nevertheless, research needs more emphasis on governance, technology, and crowdsourcing outcomes (Pedersen et al. 2013). In summary, the literature has recognized that CIs are essential for driving open innovation, but we need further understanding of how the Internet environment can successfully manage R&D projects.
Inter-organizational governance mechanisms and trust-building in crowdsourcing
Studies of inter-organizational sourcing relationships have focused on two governance mechanisms: formal controls and relational governance. Formal controls are “the written contracts and institutional mechanisms designed to guide behaviors toward desired objectives, whereas relational governance is unwritten, work-based mechanisms designed to influence inter-organizational behavior” (Goo et al. 2009, p. 120). Contracts are useful for preventing exchange hazards (Ulset 1996; Willamson 1985). In addition, CIs, as transaction intermediaries, can use “third-party institutional structures that provide a rational basis for interaction among marketplace participants” (Pavlou and Gefen 2004, p. 41), such as feedback, escrow services, and credit-card guarantees for gaining trust.
Mutual trust, relational capital (Dyer and Singh 1998; Lee and Cavusgil 2006), and social actions such as open communication and cooperation (Goo et al. 2009; Lacity et al. 2009) are essential for forming buyer–supplier relationships. Relational governance improves the performance of inter-organizational exchanges in general (McEvily et al. 2003) and IT outsourcing in particular (Sabherwal 1999). Even the most fundamental discrete exchanges include some relational elements (Macneil 1980).
Mutual trust is essential for alliance success and performance (Zaheer et al. 1998; Zaheer and Venkatraman 1995). To maintain long-term and healthy relationships between organizations, vulnerability must be reduced (Bachmann and Inkpen 2011; Doney and Cannon 1997; Gambetta 1988; Katsikeas et al. 2009; Kee and Knox 1970). Trust is even more important for enabling online transactions among geographically dispersed users who lack prior history in information and communication technologies (ICT).
Before interactions occur, CIs can build swift trust based on third-party guarantees and recommendations (Shapiro 1987; Zucker 1986). Institutional warrants ease the development of swift trust (Kramer 1999). In the crowdsourcing marketplace, third-party recommendations act as institution-based swift trust that users can generalize and transfer to related entities (Hamilton and Sherman 1996; Stewart 2003). Swift trust then transfers from CI to crowdsourcers and crowdsourcees who lack previous transaction history. After initial contacts, however, knowledge-based trust depends on personal knowledge about past behaviours (Robert et al. 2009). Online members may need a longer time to assess whether they can trust other members.
First, both crowdsourcers and crowdsourcees must trust the CI. CIs must provide reliable, secure, and fair online platforms with open rules and procedures. They must use accredited sourcers/sourcees who register online and encourage benevolent transaction norms. If sourcers and sourcees lack initial trust, they will not sign up with the CI or continue operating in the crowdsourcing environment.
Second, as intermediaries, CIs must trust sourcees for having the ability, skills, and competence for which they claim qualification, and must trust that sourcers who initiate the sourcing process will honour their promises to pay the sourcees once the tasks are completed satisfactorily. Thus, all three entities must trust each other to complete the dynamic transaction system. Any missing links will cause the crowdsourcing project to fail.
Guanxi and affect-based trust-building
Swift trust is institution-based (Kramer 1999); knowledge-based trust is cognition-based (Chua et al. 2009). Affect-based trust is unique in the Chinese context (Chen et al. 2013; Park and Luo 2001). China is generally less trusting than comparable countries (Bjørnskov 2007). From individualist and collectivist country perspectives, Chinese people tend to have lower levels of trust than citizens of the United States and Germany (Huff and Kelley 2003). Chinese business people are less able to overcome the sense of distrust that naturally exists between potential partners, especially in the initial stages of business relationships (Davison and Ou 2008). Guanxi, an embedded and sophisticated concept in the Chinese trust context, plays an essential role in trust-building. Guanxi indicates pre-existing personal connections and informal social networks (Hennart 2015) with trust, bonding, reciprocity, and empathy dimensions (Park and Luo 2001; Wong 2007), based at interpersonal/individual levels rather than being in “the system” (Martinsons 2008). In general, guanxi is a major factor for successfully conducting business and building partnerships in China (Millington et al. 2005; Park and Luo 2001; Wong 2007). Guanxi networks are essential for developing successful inter-company relationships (Hutchings and Murray 2002). Guanxi causes affect- and cognition-based trust to be more intertwined for Chinese managers in business transactions (Chua et al. 2009).
In the context of online intermediaries in China, B2B platforms such as alibaba.com explicitly represent trustworthy knowledge, enhance supplier selection and diminish the salience of guanxi networks (Davison and Ou 2008). However, high-stake R&D crowdsourcing projects are more complex than B2B platforms and require higher trust levels (Das and Teng 1998; Economist Intelligence Unit, 2008). It is unclear whether Chinese companies can transact with distant parties and develop new partnerships without utilizing guanxi.
Recognizing the lack of prior research, we chose case study (Miles and Huberman 1994) as our method for observing the uptake and development of R&D crowdsourcing in China. The case-study approach is particularly useful for observing, explaining, and exploring new phenomena within their real-life setting, especially for answering why and how questions (Verner and Abdullah 2012) and for investigating unclear boundaries between contemporary phenomena and their context (Myers 2013; Yin 2003).
The unit of analysis is an R&D crowdsourcing project between the QC Corporation as crowdsourcer, cs.com as the CI intermediary, and Jack and Rachel, two actual crowdsourcees. The cs.com company was expanding its CI business into R&D crowdsourcing, and asked the authors to be outside observers for identifying the opportunity and challenges of R&D crowdsourcing activities. In response, the authors gathered feedback regarding the launch, implementation, and transaction of the project, in accordance with previous studies in which the authors had active and full access to the research site and with data sources, and were able to engage in extensive observation despite having minimal research resources (Heiskanen et al. 2008).
The data were collected after the event and activities had already occurred and the outcomes were known. Retrospective researchers can observe final process outcomes. Consequently, retrospective designs are popular for learning about the past in organizational and strategic management research (Golden 1997; Miller and Glick 1997) and organization-theory research (Bourgeois and Eisenhardt 1988; Feeser and Willard 1989; Huber and Power 1985).
The primary data came from eleven written questionnaires with open questions (See “Appendix”) Questionnaires were mailed to participants, because geographic and time-zone differences prevented face-to-face interviews. Participants included the general manager of cs.com, responsible for the general operation and management of the CI; the chief technical officer managing the IT-infrastructure and technical background of the CI; one crowdsourcer; and two crowdsourcees involved in the crowdsourcing project. Cs.com initially invited six other companies to use crowdsourcing, and they also answered questionnaires after we reached them through the personal networks of the manager of cs.com. The six companies were from knowledge-intensive industries potentially needing R&D outsourcing such as IT, telecommunications, and aerospace. Approximately 1 month passed before all participants returned the questionnaires. The questionnaires were originally conducted in Chinese, but we then translated the key points into English. The questions included aspects of overall experiences and concerns about participating in R&D crowdsourcing projects. The participants were also asked whether the Chinese context affected their trust. They also gave suggestions for developing R&D crowdsourcing in China.
For triangulation purposes, we collected additional documents (Myers 2013) such as archival project documents, memos, reports, intermediary webpages, and trial working documents stored on the website. We also conducted real-life observations throughout the project’s lifecycle to observe how the events evolved from initiation to completion.
We examined the data using thematic analysis, widely used in qualitative research (Saldaña 2015) to create meaningful patterns. Researchers study the data, generate initial codes, search for themes; review, define, and name the themes; and produce a final report.
In addition to thematic analysis, we adopted “reflection in action” (Schön 1983) in which hypotheses are generated based on data acquired during case development (Heiskanen et al. 2008). The idea was later developed to mean that researchers report their direct experience in a way that makes sense to both academic and practical audiences (Heiskanen 1995; Heiskanen and Newman 1997). The goals are to reflect on how a problem is solved, to observe the event procedures and sequences, and to rethink any presuppositions.
The case of R&D crowdsourcing at cs.com
The high-tech QC Corporation (QC) is located in Henan Province, China. Founded in 2006 with the support of the Development and Reform Commission of Henan Province, QC’s primary business consists of electric automobiles, energy conservation, and alternative-energy products. As a pioneer in promoting and producing electric automobiles in China, QC saw a need for a comprehensive analytical report examining the whole industry.
Cs.com, a crowdsourcing intermediary start-up based in Shanghai, recently launched its CI website, aiming to position itself as an innovation platform for finding external talent for highly complex and challenging projects. In December 2010, QC licensed cs.com to crowdsource the QC Corporation Electric Automobile Production Feasibility Analysis Report, for ¥150,000 RMB (approximately $24,000 USD). To advertise the project, the staff of cs.com extensively publicized it and the CI platform through posts in public forums, university forums, university teach-ins, and emails from December 2010 through January 2011. About twenty crowdsourcees submitted proposals to cs.com. Rachel, a graduate student from a prestigious university in China, with little working experience, won the bid and contracted with cs.com to finish the report between 1 February 2011 and 31 May 2011.
Around 15 April 2011, the cs.com staff reminded Rachel of the approaching deadline and prompted her to hand in her half-finished proposal. Surprisingly, she had failed to make substantial progress because she had received other unexpected work. On 1 May 2011, she hastily submitted an unsatisfactory draft report. According to her contract, she should have received no payment, but cs.com recognized the work she had done by giving her 30% of the total payment.
Cs.com staff were highly frustrated. The manager then used his personal guanxi to find Jack, an expert with more than twenty years of experience working at a research institute, and hired him to complete the work by the end of June. The manager of cs.com negotiated with QC to extend the project deadline so that cs.com received no fine. Jack submitted the report at the end of June. Luckily, QC was satisfied with the final report and did not impose a penalty for the one-month delay; they executed full payment on 15 July 2011. Jack got 70%, but cs.com received nothing for facilitating the project.
Formal control mechanisms and swift trust
In the crowdsourcing context, swift trust is established before anyone has knowledge about members’ prior behaviour (Robert et al. 2009). As the starting point for sourcers and sourcees, building swift trust and eliminating scepticism about opportunism are essential. Control mechanisms can reduce the threat of opportunism and governance costs (Hödl and Puck 2014), so CIs should establish formal control mechanisms to develop swift trust in the Internet environment. Our case study shows that the CI failed to provide the necessary formal control mechanisms to promote swift trust.
Now it’s really key to establish a large talent pool that can meet all kinds of needs. The company might not be looking for the top one among the talent pool, but they surely would look for the most appropriate one to do their work.
Usually, we prefer to outsource a task to people that we know or we have cooperated with before. It would be hard for us to accept the new model [crowdsourcing platform] first, while at the same time outsourcing R&D to someone we don’t know. We need to escrow a big amount of money on the website and wait for a result that we are uncertain of. Although I know the manager well, I still have doubts about whether the firm can solve my problem.
Because the sourcers and sourcees lack prior knowledge of each other, speculation and mistrust inhibit initial contact. Rachel and Jack expressed the same concerns as sourcees. They mentioned the uncertainty factor: working hard but not getting paid when the job was done.
Authentication (Luo 2002) in an e-commerce setting is necessary for R&D crowdsourcing. Consequently, both sourcers and sourcees must be authentically registered so that the CI will trust them both. In our case, cs.com recruited the candidates through an online job posting without a valid authentication system to verify sourcee’s qualifications, and failed to initiate swift trust between the sourcer and sourcee.
In China, crowdsourcers are commonly required to guarantee payment by depositing the full charge. When the task is completed successfully, the CI then releases the payment to the sourcee after collecting the CI commission. In our case, both QC and the manager of cs.com indicated concerns about the large deposit, but cs.com still asked QC to deposit the full amount. Because R&D projects are usually quite costly, crowdsourcers take financial risk when making a full deposit. However, the crowdsourcees were assured payment by seeing that the CI held the full deposit. It is more difficult to build trust under high uncertainty and risk (Dong and Glaister 2007). Indeed, CIs must establish effective formal control mechanisms tailored to R&D projects and minimizing e-commerce uncertainties through third-party recommendations, escrow services, and mutual ratings (Pavlou 2002; Pavlou and Gefen 2004). In our case, cs.com failed to implement appropriate formal control mechanisms and to build swift trust. Hence, we propose several effective control mechanisms that CIs can adopt to drive swift trust:
CI-based formal control mechanisms can build trilateral swift trust between crowdsourcers, CIs, and crowdsourcees.
Talent pools and mutual recommendation systems will help crowdsourcers trust CIs and crowdsourcees.
Authentication systems will help crowdsourcers and crowdsourcees build mutual trust.
Partial payment escrows and CI warranty systems will help crowdsourcers trust CIs, crowdsourcees trust CIs, and crowdsourcees trust crowdsourcers.
Relational mechanisms and knowledge-based trust
The manager of cs.com also said:
It’s not that we are not trusting QC; it’s not that we intentionally delayed the job; it’s purely our fault because of poor time management. We should have planned well. However, it would be helpful if the website could in some way monitor each project that is being crowdsourced on the platform.
We didn’t do a good job of monitoring the process. For sure, we have to reconsider the crowdsourcing mechanisms and website features.
This quote suggests that CIs must do more than just act as intermediaries to connect sourcers with sourcees. To keep the platform in business, they must also be facilitators, using relational mechanisms to enhance knowledge-based trust between sourcers and sourcees. Performance management often requires formal control mechanisms. In the crowdsourcing context, a performance-tracking application, such as a status report, could be the relational mechanism that enhances knowledge-based mutual trust. A status report allows the sourcee to provide regular progress reports that keep the sourcer informed so that both can track the progress and ensure that it is completed on time. In our case, QC never actually engaged in performance management. In such circumstances, it may not be altogether surprising if a crowdsourcee fails to complete a task.
Cs.com also failed to provide a platform to allow QC to communicate directly with the sourcee. We suggest a workroom platform as a CI-based communication application that can allow sourcees and sourcers to directly interact. Both status reports and workrooms can facilitate open communication and timely information exchange to build the necessary mutual knowledge-based trust (Goo et al. 2009). The manager of cs.com later realized that the design of performance-tracking and communication applications is essential for R&D CI settings.
Third, arbitration is necessary when sourcers and sourcees cannot resolve contract disputes. Arbitration strives for harmonious conflict resolution (Goo et al. 2009) and should be highlighted in the R&D CI setting. In the case of cs.com, arbitration was more guanxi-based than legal-based, which might apply well in the Chinese context. However, a mature CI would require arbitration justice as an effective way of managing outsourcing risks if contractor performance is unsatisfactory (Vining 1999). An arbitration mechanism will increase mutual trust between sourcers and sourcees, and their mutual trust in the CI.
Relational governance mechanisms can help sourcers, sourcees, and CIs build trilateral knowledge-based trust.
Performance-tracking systems will help sourcers and sourcees share knowledge-based trust.
CI-based communication systems will help sourcers and sourcees share knowledge-based trust.
CI-based arbitration systems will help crowdsourcers/crowdsourcees trust the CI and will help crowdsourcers and crowdsourcees build mutual trust.
Guanxi and affect-based trust
People do business in China by guanxi in most cases, so the job is not outsourced to the most appropriate person, but to the most familiar person. So the key to success is to get to know the person.
Companies have difficulties all the time, big or small. Sometimes they cannot solve them because of limited resources, like us. Our city is small, and we are short of talents; for China, the most excellent talents flow into big cities like Beijing and Shanghai. Therefore, if a mature crowdsourcing intermediary with solid mechanisms can protect its users from inappropriate behaviors or harm in the meantime, and if they can help us save costs, shrink working periods, and find better solutions for difficulties, why not try it?
The Internet has been changing the structure of Chinese society, including people-to-people trust. But it’s a slow process. What we can do is to take full advantage of the surging trend of crowdsourcing and figure out the mechanisms to make it work, just like when e-commerce sprang up decades ago.
His statement shows that guanxi connections are still significant in the R&D crowdsourcing market, and that relationship-based management capabilities and strategies potentially enhance product innovation (Xin and Pearce 1996). In fact, QC hired cs.com to crowdsource the project mainly because their managers knew one another personally. Cs.com selected Rachel as the service provider, but when they realized that she could not accomplish the task as expected, the cs.com manager used his personal relationships to find someone outside the sourcing pool to complete the task. In other words, he used guanxi connections as a contingent relational mechanism to ensure that the crowdsourcing project succeeded, to maintain the relationship, and, particularly, to restore affect-based trust with QC.
Intermediaries go beyond simply linking parties; they search for and transform ideas, and combine solutions to fit their clients (Hargadon and Sutton 1997). In the Chinese context, guanxi connections can be contingencies or supplements for finding the right crowdsourcers or crowdsourcees and enforcing transactions (Hennart 2015). Indeed, in China, guanxi is more influential in the search for suitable suppliers than other information sources such as company websites, trade fairs, and the trade press (Millington et al. 2005). Guanxi-based transactions help maintain affect-based trust, but they can lower levels of swift and knowledge-based trust. In our case, cs.com retained their guanxi within the company, and provided no online path allowing Rachel and QC to directly communicate. Instead, cs.com monitored the progress and performance. When the project was failing, cs.com utilized their guanxi, which prevented the sourcers and sourcees from building swift and knowledge-based trust or a continual working relationship. Nevertheless, open innovation should allow direct choice and engagement between sourcers and sourcees. Consequently, guanxi-based transactions violate the nature of open innovation and are a major hurdle for the organic growth of open innovation.
Our analysis leads to the following propositions:
In China, guanxi is a complementary and contingent relational governance mechanism for providing open innovation and maintaining affect-based trust.
In China, guanxi inhibits the development of swift and knowledge-based trust and the organic growth of crowdsourcing-based open innovation.
Discussion and conclusions
In our case study, we incorporate the literature of inter-organizational governance and trust development as a theoretical contribution to the open-innovation literature. Although crowdsourcing can be used to drive open innovation (Zhao and Zhu 2012), few researchers have explored how innovation projects are successfully managed in the Internet environment. Our study underscores that Internet crowdsourcing intermediaries (CIs) are critical for providing open-innovation opportunities to manage R&D projects, but they must utilize appropriate governance mechanisms to facilitate trust development among crowdsourcers, crowdsourcees, and CIs. The case highlights the additional importance of relational governance mechanisms for managing highly complex R&D crowdsourcing, such as through open and timely communication between sourcers and sourcees, mediation from the platform for resolving conflicts, and performance-management tools. Considering that most open-innovation studies have focused on western practices, we contribute by identifying effective governance mechanisms specific to the Chinese context (Byrum and Bingham 2016) and show that Chinese CIs should safeguard transactions and ensure that innovation projects will grow through customized technical functions such as partial payment and CI warranty systems.
Second, we extend the inter-organizational governance literature and e-commerce trust concept (Ba and Pavlou 2002; Benbasat et al. 2008) to the R&D crowdsourcing context, and thus contribute to understandings of trilateral trust-building with intermediaries in business transactions (Svensson 2004). Businesses traditionally use relational governance to enhance trust between intermediaries and buyers or between intermediaries and suppliers. They have no need to govern or enhance trust between suppliers and buyers. However, crowdsourcing involves a different business environment: crowdsourcing intermediaries are facilitators and must provide formal and relational governance mechanisms to enhance swift trust and knowledge-based trust between crowdsourcers and crowdsourcees. Swift trust encourages their initial contact; knowledge-based trust further ensures that they will return for future transactions. The proposed trilateral trust-building is therefore critical to provide a healthy environment for the growth of crowdsourcing-based open innovation.
Third, we offer additional understanding of China’s guanxi tradition and provide evidence that it can have double-edged functions (Chen et al. 2013) for the growth of crowdsourcing. Guanxi provides social capital that substitutes for formal institutional support (Xin and Pearce 1996), enforces transactions (Carlile 2002; Hennart 2015), and provides business opportunities (Wong 2007). Therefore, guanxi connections are contingent formal control and relational governance mechanisms for maintaining affect-based trust. However, they can inhibit swift and knowledge-based trust development among key players in China’s R&D crowdsourcing marketplace. In our case study, guanxi prevented the CI from providing open and transparent selection of crowdsourcers and crowdsourcees, which indicates that ventures pursuing new technological innovations may find guanxi to be an ineffective strategy (Li and Atuahene-Gima 2001). We suggest that the cultural tradition of guanxi as a prerequisite for business relationships violates the needs of R&D crowdsourcing to provide open biddings and open communications, and is, consequently, a major barrier to the growth of open innovation in China.
Limitations and future studies
Our study has limitations that should be considered. We urge practitioners and researchers in the field to further examine crowdsourcing in other areas. First, we examined the issue from the perspective of trust-building, but Chinese businesses may avoid crowdsourcing R&D tasks for other reasons, such as the characteristics of the tasks and management issues. We challenge future research to follow up the study and offer more insights, particularly by exploring non-intermediary and re-intermediary processes (Sun 2006) and comparing the new online crowdsourcing model with traditional intermediary services. Second, our single case study helps to explain the research context, but case studies lack transferability. Future studies could incorporate additional cases to allow ‘replication’ logic to confirm or refute conceptual insights (Brown and Eisenhardt 1997; Yin 2003). Third, we conclude with managerial implications to be drawn from the analysis. Future study could use a quantitative approach to test the concept empirically and continue to enhance understanding.
China will gradually and inevitably accept open innovation, and it thus provides a dynamic setting for examining firms as they adopt crowdsourcing as an open-innovation approach. Therefore, our study provides practical implications for the development of crowdsourcing in China. That is, we show that trust is essential for all three entities involved in R&D crowdsourcing. To meet the demand for innovation through Internet-based crowdsourcing in China, CIs should build trust by incorporating and promoting various website features to build personal and social relationships among crowdsourcers and crowdsourcees.
Clearly, guanxi practice continues to affect innovation crowdsourcing in China, although it is declining as modern China transitions towards legal contracts (Guthrie 1998). Organizations with substantial resources, minor agency problems, and well-structured or formalized routines are likely to use contracts as their basic governance mechanism. In contrast, organizations with few resources, acute agency problems, and informal relation-oriented organizational routines are more likely to use guanxi as a governance mechanism (Zhang and Keh 2010). To minimize reliance on guanxi and develop the marketplace for open innovation, Chinese CIs must gradually build their talent pools for crowdsourcing R&D projects and strengthen their formal control, relational mechanisms and performance-management tools, for example, through escrow systems and arbitration processes.
Our case study allowed us to observe the developing trend of Internet-based crowdsourcing to achieve open innovation in China and to identify challenges facing R&D crowdsourcing. We identify inter-organizational governance mechanisms for building trilateral trust among crowdsourcers, crowdsourcing intermediaries and crowdsourcees. We propose that formal control and relational governance mechanisms must be adapted to the Chinese context. We highlight that guanxi is a temporary relational source, but a long-term barrier to the growth of open innovation in China.
This work was supported by the National Natural Science Foundation of China (NSFC) Grants Nos. 71171131, 71471141, 71503108 and 91646205. It was also funded by the Chinese Scholarship Council. The authors would also like to acknowledge the valuable feedback from the editor and two anonymous reviewers.
- Brabham, D.C. 2008. Crowdsourcing as a model for problem-solving: An introduction and cases. Convergence 14 (1): 75–90.Google Scholar
- Byrum, J., and A. Bingham. 2016. Improving analytics capabilities through crowdsourcing. MIT Sloan Management Journal 57 (4): 43–48.Google Scholar
- Cabiddu, F., M. Castriotta, M.C. Di Guardo, and P. Floreddu. 2013. Open innovation and crowdsourcing communities design: A cross-case analysis. In Designing organizational systems: Lecture notes in information systems and organisation 1, ed. R. Baskerville, et al., 143–155. Berlin: Springer.CrossRefGoogle Scholar
- Chesbrough, H.W. 2003. Open innovation: The new imperative for creating and profiting from technology. Boston: Harvard Business Press.Google Scholar
- Chesbrough, H.W. 2006. Open innovation: A new paradigm for understanding industrial innovation. In Open innovation: Researching a new paradigm, ed. H. Chesbrough, W. Vanhaverbeke, and J. West, 1–12. Oxford: Oxford University Press.Google Scholar
- Das, T.K., and B.S. Teng. 1998. Between trust and control: Developing confidence in partner cooperation alliances. Academy of Management Review 23 (3): 491–512.Google Scholar
- Dyer, J.H., and H. Singh. 1998. The relational view: Cooperative strategy and sources of interorganizational competitive advantage. Academy of Management Review 23 (4): 660–679.Google Scholar
- Economist Intelligence Unit. 2008. The role of trust in business collaboration: An Economist Intelligence Unit briefing paper sponsored by cisco systems. http://graphics.eiu.com/upload/cisco_trust.pdf. Accessed 10 April 2016.
- Erickson, L. 2012. Hanging with the right crowd: Matching crowdsourcing need to crowd characteristics. Paper presented at the 18th Americas conference on information systems, August 9–11, Seattle, USA.Google Scholar
- Gambetta, D. 1988. Trust: Making and breaking cooperative relations. Cambridge: Blackwell.Google Scholar
- Gefen, D., and D.W. Straub. 2004. Consumer trust in B2C e-commerce and the importance of social presence: Experiments in e-products and e-services. International Journal of Management Science 32 (6): 407–424.Google Scholar
- Goo, J., R. Kishore, H.R. Rao, and K. Nam. 2009. The role of service-level agreements in relatio¤ment of information-technology outsourcing: An empirical study. MIS Quarterly 33 (1): 119–145.Google Scholar
- Heiskanen, A., and Newman, M. 1997. Bridging the gap between information systems research and practice: The reflective practitioner as a researcher. Paper presented at 18th international conference on information systems, 15–17 December, Atlanta, GA, USA.Google Scholar
- Macneil, I.R. 1980. The new social contract: An inquiry into modern contractual relations. New Haven: Yale University Press.Google Scholar
- Miles, M.B., and A.M. Huberman. 1994. Qualitative data analysis: An expanded sourcebook. Thousand Oaks: Sage.Google Scholar
- Myers, M.D. 2013. Qualitative research in business and management. London: Sage.Google Scholar
- Pedersen, J., Kocsis, D., Tripathi, A., Tarrell, A., Weerakoon, A., et al. 2013. Paper presented at 46th Hawaii international conference on system sciences (HICSS), 7–10 January, Wailea, HI, USA.Google Scholar
- Saldaña, J. 2015. The coding manual for qualitative researchers. Thousand Oaks: Sage.Google Scholar
- Schön, D. 1983. The reflective practitioner: How professionals think in action. New York: Basic Books.Google Scholar
- Sun, K-q. 2006. Non-intermediary and re-intermediary in e-commerce. Journal of Yunnan University of Finance and Economics 22 (1): 50–53.Google Scholar
- Willamson, O. 1985. The economic institutions of capitalism. New York: Free Press.Google Scholar
- Yin, R.K. 2003. Applications of case study research. Thousand Oaks: Sage.Google Scholar
- Zucker, L.G. 1986. Production of trust: Institutional sources of economic structure, 1840–1920. Research in Organizational Behavior 8 (2): 53–111.Google Scholar