Background

Network analysis in marketing refers to a range of emerging applications of graph theory. In particular, network analysis combined with theories and models of social influence is important for understanding the roles and effects of influencer marketing, the diffusion of innovations within a population, and other aspects of modern marketing. 

The two core components of networks are nodes typically representing people but could be other entities (e.g., firms, stores, hashtags, etc.), and links (also called ties, edges) representing relationships between nodes. Links could represent various types of relationships such as acquaintance/friend, business connections, familial bonds, influencers-followers, donors-recipients, communication/conversations (e.g., tweets, re-tweets), co-occurrence of words in product reviews, etc.  Social networks are a special kind of network that have distinct characteristics based on the tendency of human beings to form clusters (families and friends) and communities (e.g., interest groups, acquaintances, cities). The growth of social media is making visible different types of nodes and the links between them, providing the impetus for social network analysis in marketing.

There are several areas of marketing where network analysis is particularly useful, including the following.

• Identifying influencers based on their position in the social network associated with a focal organization (i.e., the influence potential is based, in part, on the social network and not just on intrinsic individual characteristics).
• Gaining an understanding of how an influence process will propagate through a network based on different sets of influencers (i.e., alternative seeding of the process).
• Identifying important concepts within semantic networks constructed from user-generated content. 

Network analysis is a vast topic that could (and should) be a separate course by itself. With the Enginius software, you will be able to visualize a network‘s structure, summarize the structural properties of nodes as well as the structural properties of the entire network, including structural components such as clusters and communities, and finally, you can assess how a diffusion process might evolve through a social network under different strategies for seeding the process.