Your community of customers, like any other community, consists of a collection of individual parties that can also be referred to as “actors” who are related to one another. These relationships can be modeled using the graph abstraction, in which every actor is represented as a node, and every connection between two actors is represented as a link or an edge between two nodes.

For example, if one actor is friends with another actor, it would be represented as two nodes with one undirected edge between them.

However, not all relationships are necessarily reciprocal. For example, actor A may be the father of actor B, while actor B is the daughter of actor A. These relationships would have to be modeled slightly differently to show the direction of the relationship with a directed edge between them. This also allows you to refine the nature of the relationships between actors within a community.

This abstraction can be used to represent any relationship among a set of actors, such as households, schoolmates, people who work in the same industry, etc. Not every relationship is between two customers, either. One customer has purchased one of your company’s products – one node represents the customer, the other represents the product, and the edge between them would indicate that the customer purchased that product. In this way we can build up a social network representing the ways that the different actors are connected; this social network model becomes the basis for a number of different analyses for understanding and taking advantage of ways some customers can influence the behaviors of others.