DANIEL BIRKE

Diffusion over networks: Modelling the spread of innovations and customer churn over social networks

The aim of this paper is to introduce a general concept of how to model diffusion of innovations on networks. Traditionally, most diffusion studies focus on aggregate processes. The model presented here is micro-funded in that it takes the complete network structure of a large scale network into account. In the second part of the paper, we are empirically testing the proposed model with the help of usage data from a large European mobile phone operator. We show that switching of subscribers from one operator to another operator is contagious and spreads over the social network. The methodology presented in this paper can also be used for modelling the spread of innovations in general and over a variety of different networks, like the one created by social network software.