Leveraging social networks for analytical customer relationship management

Meer over het boek

CRM departments often utilize data mining to predict customer behavior, typically overlooking the customer's position within a social network. However, many companies possess not only individual customer behavior data but also insights into customers' social networks. The telecommunications sector, for instance, generates vast amounts of calling records that reveal both individual usage patterns and customer interactions. This thesis aims to harness network data for marketing purposes, focusing on two main areas. First, it examines how incorporating social network information can enhance the predictive accuracy of classification tasks. Second, it explores various centrality measures in the context of viral marketing, which seeks to leverage existing social networks to amplify brand awareness through processes akin to epidemic spread. When customer network data is available, centrality measures can effectively facilitate the dissemination of viral marketing campaigns. The thesis conducts computational experiments to compare different centrality measures for marketing message diffusion. Results indicate a notable improvement when targeting central customers for message distribution, alongside significant variations in effectiveness among centrality measures based on the network's topology and the diffusion process employed.