Model-building for better media deployment

Robert Stratton

For the marketing science industry, the fragmentation of media channels, and the increasing role of interactive and social media pose new challenges. In particular, the question of causality – what 'caused' a sale or response is more complex to pinpoint than ever before. Our previous definitions of causality no longer work as well in the context of on- and offline interactions and intermediate customer behaviours. In addition, analytical and scenario planning methods that have a proven history when used on high-reach, high-frequency channels like TV are less suited to handling smaller channels, social media and word of mouth.

In order to understand the direction of causality and deconstruct a campaign's performance we need a reliable conceptual model of the underlying consumer process – a framework that represents the real mechanisms at work. For a broadcast medium this used to be straightforward – methods like adstock were widely used and accepted. For a campaign involving digital media the underlying process is more complex.