The Effect of Social Capital and Outcome Expectations on Product Recommendations in Social Networks: An Empirical Comparison of Facebook and ASmallWorld

Diplomarbeit von Deniz Yildirim

Introduction:

Problem Outline:

The importance of information exchange through the Internet is growing and social networking sites are having a profound impact on the evolution of Internet business and e-commerce. According to Comscore, Facebook, the world’s most popular and well-known social networking site, was the fourth largest site worldwide with 340 million unique users in July 2009 right after Google, Microsoft and Yahoo (Internet World Business 2009) and the number one website in the US as of March 2010. Facebook’s growth underscores the ongoing shift away from algorithm-based search engines to social search, which are results based on interpersonal interaction. Interestingly, according to Nielsen, the growth of social networking for information searches coincides with an increase of global consumer trust in online channels with 70% trusting consumer opinions posted online in 2009 compared to 60% in 2007. Both of these current trends have been greatly facilitated by social networking sites.

Most consumers are members of multiple social networks and use these networks to fulfill different needs and desires. Social networking sites are generally classified into two categories: open social networks (OSNs) and invitation-only social networks (ISNs). Social networks are also organized around either a specific topic or general demographic such as friends or business partners. OSNs have no entry restrictions, whereas ISNs as private social networking sites require invitation, which results in a limitation in its membership base.

Whether social networking sites are OSNs or ISNs, previous research has found that qualitative and rich knowledge contribution is essential to successful online communities. Chiu et al. suggest that knowledge contribution is influenced by social capital and outcome expectations. Social capital is the network of relationships possessed by an individual or a social network and the set of resources embedded within it. Outcome expectations represent the expectations that an individual has towards the possible outcome of knowledge contribution for him/herself or for the community as a whole. From a marketing perspective, an important possibility to utilize member participation in terms of knowledge contribution is the encouragement of product recommendations between social networking site members, which may strongly influence the purchasing decisions of other members. A key aspect in this case is, whether or not recommenders receive monetary or non-monetary rewards. At first glance, one may think that a monetary reward should increase a member’s willingness to participate. However, this may not be always the case because trust-based relationships within the community and among its members greatly impact consumer behavior and should be cultivated to influence a member’s participation. The following scenario will show that the decision about implementing a monetary or non-monetary product recommendation system in different social network types is important since different approaches can lead to very different outcomes: ISNs have entry restrictions and are therefore smaller, which increases trust since relationships are based on authentic connections and true information. An ISNs focus is to provide a confidential platform for like-minded individuals who already have strong connections with each other. As such, a monetary incentive for a recommender should not influence the acceptance of recommendations because members trust each other and are intrinsically motivated. In contrast, OSN members usually have weaker connections and therefore, may trust other members less. In this environment with minor trust, it could be expected that some OSN members only recommend products to gain monetary rewards without consideration of whether their input helps the receiver of their recommendation. OSN members will be less trusting in product recommendations than ISN-members because they cannot distinguish between valuable and invaluable products recommendations. This results in a decreasing influence on the purchasing decision, making a monetary rewarded product recommendation system less successful than initially believed.

The scenario outlined above shows that the decision about monetary or non-monetary rewards is not trivial and sometimes not intuitive as different setups may lead to different outcomes depending on whether it is an open or invitation-only social network. Therefore it must first be empirically validated which product recommendation system will be successful prior to implementation. Then, as a next step, social capital and outcome expectations as important drivers of knowledge contribution in terms of product recommendation should be analyzed for both OSNs and ISNs.

Objective:

The aim of this thesis is to measure and analyze the differences in the impact of social capital and outcome expectation elements on monetary and non-monetary rewarded recommendations between open and invitation-only social networks and to identify which product recommendation system should be implemented for ISNs and OSNs.

Procedure:

This thesis is organized as follows: Chapter 1 gives a brief introduction to the objective of this research. Chapter 2 defines open- and invitation-only social networking sites and compares the social networking sites. Further, the motivations for product recommendations in social networking sites, as well as, the impact of social capital and outcome expectations are discussed. Chapter 3 defines the research model for this study and sets up hypotheses based on previous research in the area of social networking sites. Chapter 4 empirically compares and contrasts factors that influence knowledge contribution in open- and invitation-only social networking sites, using structural equation modeling with data from a survey conducted for this study. Based on the differences between open- and invitation-only social networking sites, this thesis will then identify the appropriate product recommendation system needed for each social network type in chapter 5.

	Table of Contents:
	List of Figures	IV
	List of Tables	V
	List of Abbreviations	VI
1	Introduction	1
1.1	Problem Outline	1
1.2	Objective	2
1.3	Procedure	3
2	Theoretical Background	3
2.1	Social Networking Sites	3
2.1.1	Definition	3
2.1.2	Previous Research	4
2.1.3	Invitation-only Social Networks (ISNs)	6
2.1.4	Open Social Networks (OSNs)	9
2.1.5	Theoretical Comparison of ISNs and OSNs	11
2.2	Previous Research about Product Recommendations	12
2.2.1	Traditional Word-of-Mouth	13
2.2.2	Electronic Word-of-Mouth	14
2.2.3	Comparison of WOM and eWOM	17
2.2.4	eWOM in Social Networking Sites	18
2.3	Previous Research about Social Capital and Outcome Expectations	19
2.3.1	Structural Capital	20
2.3.2	Relational Capital	21
2.3.3	Cognitive Capital	24
2.3.4	Outcome Expectations	25
3	Conceptual Model and Hypotheses	27
4	Empirical Research	35
4.1	Research Methodology	35
4.2	Description of the Sample	38
4.3	Common Method Bias Analysis	41
4.4	Measurement Model Validation	42
4.5	Structural Model Validation	46
4.6	T-tests for the Comparison of OSN and ISN	50
4.6.1	Differences in Relational Capital	52
4.6.2	Differences in Outcome Expectations	53
5	Discussion and Implications	54
5.1	Findings	54
5.2	Research Implications	58
5.3	Managerial Implications	60
5.3.1	Product recommendation systems for OSNs	61
5.3.2	Product recommendation systems for ISNs	64
6	Conclusion	67
	Appendix	69
	References	77

Text Sample:

Chapter 3, Conceptual Model and Hypotheses:

In this section the research model is being developed. This thesis investigates the influence of social capital and outcome expectations on product recommendations in social networks and compares the influence of relational capital and outcome expectations on ISNs and OSNs. Chiu et al. (2006) followed a similar approach to this thesis to analyze knowledge sharing in online communities in terms of quantity and quality. Their model is also based on Nahapiet and Ghoshal’s (1998) three dimensions of social capital and outcome expectations on a personal and community level suggested by Compeau and Higgins (1995). Prior research showed that although the three dimensions of social capital are interrelated with each other to some degree, they should be analyzed independently of each other.

There are three notable differences between Chiu et al.’s (2006) approach and this study:

They analyzed the effects that impact the quantity and quality of knowledge contribution. In contrast, this thesis focuses more on a special case of knowledge contribution, the quality of product recommendations as a part of product-related knowledge contribution.

Chiu et al. (2006) observed non-monetary rewards for knowledge contribution. In this area, the focus is on the relationship between recommendations and monetary and non-monetary rewards.

The specific focus in this study is on social networking sites rather than online communities in general.

However, the focus for the comparison of ISNs and OSNs lies only on the relational dimension of social capital to predict product recommendations. Following Wiertz and Ruyter (2007) and Casaló et al. (2008) there are generally three arguments for this reduction. First, social networking sites are platforms to maintain and develop relationships among its members. Therefore the influence of relational capital on a member’s knowledge contribution behavior should be studied. Second, relational capital refers to the assets that are contained in a relationship. In contrast, structural and cognitive capital, do not contain these assets but rather facilitate the achievement. Relational capital has been identified as a facilitator of a member’s actions within the group and strongly influences an individual’s member behavior. According to Kollock (1993), the quality of the relationship impacts a user’s willingness to share valued information. Structural capital facilitates the obtainment of information or the access of specific resources and cognitive capital facilitates social relations and the integration and combination of resources. Third, prior research analyzed firm-hosted online communities or other online community types and found mixed results for the importance of relational capital. As a result, the impact of relational capital in the context of social networking sites needs to be studied further. Wasko and Faraj (2005) did not find a significant positive impact of relational capital on knowledge contribution in online communities. In their study, they hypothesized that relational capital may not develop due to insufficient shared history, high interdependence, frequent interaction and co-presence. However, Wasko and Faraj (2005) found that relational capital does develop over time as efforts spent on the relationship with other members begin to create a shared history. Members then have access to new information, knowledge and opportunities that are embedded in the network through these close relationships, which then makes the network more valuable for the individual user. However, relational capital is network specific and therefore cannot be transferred from one network to another. This increases the opportunity costs for leaving a social network where relational capital is already developed.

Based on the preceding review of the literature, hypotheses for recommendations with monetary and non-monetary rewards and considering recommendations for ISN and OSN are postulated.