Understanding the relationship strengths in users' activities, review helpfulness and influence

The activities of users on social media lead to the establishment of different types of relationships with variable strengths. This study aims to investigate how activity level and review helpfulness affect users' influencing power, and how the relationships are moderated by tie-strength. In this regard, first a conceptual model is introduced. Second, we define measures to capture activity of users, and discuss the helpfulness of their reviews, their relationships and their influencing powers. Based on their strength relationships, the users are divided into two groups: high-strength and low-strength. To evaluate the significance of differences between groups, partial least squares–based multi-group analysis is used. We have assigned different metrics to provide better comprehension of the relation between users' activities and their influence. The analysis results from Epinions.com indicate that the type of a relationship and its strength affects the association between activity and influencing power in multiple ways. The experimental results show the effect of activity level of a user on influencing power is stronger for those with high strength of transactional connectivity than for those with low strength. Also, the effect of the users' review helpfulness on influencing power is stronger for those with high similarity strength than for those with low similarity strength. This study can facilitate the examination of the effects of different characteristics of influential power through the strength of relationships among users. The findings can also benefit online market owners to create better strategies through better understanding of the characteristics of influential users. •The moderating role of strength of transactional connectivity on users' activities and influence associations.•The moderating role of strength of transactional connectivity on users' review helpfulness and influence associations.•The effect of strength of similarity on users' activities and influence associations.•The effect of strength of similarity on users' review helpfulness and influence associations.•Estimate stronger predictor for identifying influential users.