國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
Chapter 5 Discussion
5.1 Overview of Research Finding
In our study, we detected various influential factors affecting customer resonance on social networks and analyzed customer resonance by different types of user
behavior based on demographic information for customer relationship management.
As for resonance on social-network platforms, the research results indicated that all gratifications such as self-presentation, strength of social ties, and utilitarian value of user-generated content are strongly related to customer resonance. In addition, we also found that customer resonance is significantly related to purchase intention. Then we explained the clustering of resonance on different types of attributes. Finally, we investigated the clustering results based on management issues.
5.2 Influence of Social-Network Sites on Customer Resonance
The results of the structural-equation model showed that factors in eachgratification construct are related to customer resonance on the social-network sites.
The constructs in our model are content gratification, social-relation gratification, and self-presentation gratification. We separately explain each factor in the constructs and its numerical results obtained by using statistical software.
In content gratification, the utilitarian value of content has a significant effect on resonance, but hedonic value does not. The possible explanation is that people may spend some time browsing user-generated content when they obtain some useful information, and they will have more willingness to share the posts. One of the reasons for users to share their posts is to enrich others’ lives. On the other hand, the hedonic value of content does not have a significant influence on resonance. The possible explanation is that people sharing the posts is not only for positive feelings but also for negative feelings. As long as the post can arise the strong feedback after people browsing the post, people will do resonance behaviors on post.
In social-relation gratification, the tie strength, normative influence, and information influence have significant impacts on customer resonance. As for tie strength, our data analysis showed that people have more willingness to express resonance behavior such as replying to posts in environments of strong social ties.
With regard to normative influence, it is an important factor to make resonance behavior happen on social-network sites because people desire to obtain
recommendations from others (Dholakia et al. 2004). These demands often dominate
‧
our actions, opinions, and decisions. For example, we tend to like a post from others after we see the post liked by others because it will make us conform with others. As for informational influences, people tend to comment on a post in order to obtain more information when resonance has happened. On the other hand, homophily and trust in social-relation gratification are not influence factors on resonance. With regard to homophily, it is an important factor for people engaging in virtual community because people like to interact with others who have certain similar characteristics. However, homophily don’t have directly impact on doing resonance behaviors between individual and a post. People will comment or share the post whether the post have similar characteristics with individuals. It is possible for people to do resonance behaviors on the condition people agree with the post or not agree with post. As for trust, it doesn’t have effect to arising resonance on social network sites. Traditionally, trust is an important factor when people engage in e-commerce website since individuals are relatively invisible than face-to-face interaction on networks. Trust can create a reliable atmosphere in e-commerce platform and played a major factor for people especially in exchanging benefit. People have more
willingness to do transaction because of the exchange of benefit between individuals on e-commerce platform. However, trust is not a key factor to arise resonance on the social network sites because it doesn’t involve exchanging benefit between individual and the post.
In self-presentation gratification, the construct of self-presentation is
significantly related to customer resonance because this is another space for users to manage their images. User behavior such as liking, replying, and sharing may represent their image. For example, people will share a post related to travel information in order to create an image of themselves as people who like traveling.
Hence, we will like a post, comment on posts, or share posts in order to present our identities to others.
5.3 Effects of Clustering on Customer Resonance
We partitioned the data into two clustering and labeled each clustering based on the degree of resonance. Cluster A is labeled as having a high status of resonance, indicating that people have more willingness to engage in resonance behavior such as liking a post, replying to the content, or sharing the information. In addition, people are greatly influenced by self-presentation, utilitarian value of content and social tie.
In this cluster, the percentage of female is higher than that of males, and they mainly
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
have low revenue. Lastly, the phenomenon of using platform experience is whether in rich experience in using platform or not rich experience in using platform.
On the other hand, cluster B is labeled as having a low status of resonance, which represents people who are not likely to engage in resonance behavior on social-network platform sites. Also, the three highest values separately are social tie, self-presentation and normative influence, indicating that people are in high degree of social relation in environments of social network platforms. In the cluster, the
percentage of males is much greater than the percentage of female, and has high revenue compared with high-resonance clustering. Finally, the platform experiences in using social-network sites are mainly from 1 year to 5 years.
From the perspectives of two clustering, we found that the social tie and
self-presentation play an important factor regardless in high resonance clustering and low resonance clustering. Besides, we also realized that homophily doesn’t have influence on resonance because homophily are not significant neither on low degree of homophily in high resonance nor high degree of homophily in low resonance. The phenomenon of trust is same with homophily in our research findings. In conclusion, we can summarize that high resonance clustering mainly focus on self-presentation gratification, content gratification and social relation gratification. On the other hand, the low resonance clustering mainly focus on self-presentation gratification and social relation gratification.
‧
國立 政 治 大 學