Continued from news and insight page:

Agent Based Modelling of Customer Behaviour in Telecoms and Media Markets

A bottom-up approach to understanding systems which provides a powerful tool for analysing complex, non-linear markets.

One of the most prominent features of the telecom, IT and media markets over the last ten years has been the tremendous amount of change across the various dimensions of the markets: including technical, regulatory and demand aspects. More than ever, the famous proclamation of the ancient Greek philosopher Heraclitus is apt:  the only constant is change. Moreover, these transformations have not always occurred in a proportionate or linear manner. Discontinuity has characterised many of these changes. Working in this environment has provided great challenges for analysts and managers alike.

While the traditional statistical and equation-based modelling techniques continue to be powerful tools, in certain markets their appropriateness can be called into question; especially for those markets characterized by non-linearity and complexity. In the past, the high degree of structural stability of the telecommunications market allowed econometricians to make reasonably confident predictions of the demand effect of a small change in price. Markets for new telecommunication products, however, have often proved to be highly non-linear, as witnessed, for example, by the explosive growth of pre-paid mobile services and fixed-rate Internet access. Traditional linear demand equations are not appropriate in such situations.

This article presents a new method for modelling complex systems: agent based modelling (ABM). The article covers the theoretical background to ABM and presents a brief case study of its usage in analysing the Pay-TV subscriber market.