Modelling in a multi-media environment

The purpose of media behaviour models is to provide the media planner with a decision making tool. We have had models for several decades that work at the simplest level to support questions like ‘should I buy title A or title B’, or more complex ‘should I buy this combination C of newspapers/insertions or combination D of newspapers/insertions’. But the complexity of the model compounds at the multi-media level in supporting questions like ‘should I buy this combination E (TV channels/spots) or combination F (magazines/insertions) or ‘should I mix media by taking a percentage out of combination E (TV) and placing it in magazines and if so what percentage’?
The first step in modelling media behaviour is normally to provide the planner with comparative statistics on the reach and frequency (OTS) of selected schedules. In the case of multi-media evaluation this becomes a multi-dimensional matrix with as many dimensions as there are media. From the matrix we will know what proportion of the target receive only TV OTS, what proportion receive 3 print OTS, 2 TV and 6 radio OTS and so on. The interpretation of such a matrix is no longer the simple choice of the schedule that delivers the most at (3+) OTS.

Download Paper (PDF)