A quantitative approach to media planning

Abstract

This dissertation develops methodologies for solving the media planning problem, keeping in view the three-fold requirements of fidelity, analytical rigour, and practicality. A mathematical programming model has been formulated which develops a minimum press media plan subject to the requirements of the desired minimum reach and average frequency (opportunities-to-see i.e. OT. S.) in several target audience group. The model, called the MRF model (Minimum Reach and Frequency model), uses the existing press audience data available in India from the NRS report. Real life applications of the model have demonstrated its superiority, over the methods used by practitioners, in drawing up press media plans. The problem of obtaining the range, given by an upper and a lower bound, of the reach of a media plan, when only the individual media vehicle audience and the common audiences of any two of them are known, has been formulated as a linear programme (LP). The LP being large, even for media plans of moderate size, alternatives formulations and solution procedures have been developed to obtain the bounds. The procedure, developed in the dissertation, for obtaining the micro lower bound has been proved to be superior to the existing methods. This procedure is used in the MRF model to obtain the micro lower bound. The macro bounds, both, upper and lower, (based on aggregate media audience data) developed in the dissertation have been proved to be superior to the currently known macro bounds. Keeping in view the future requirements of media planners in India, a model has been formulated which considers the scheduling of advertisements over time and the target audience response. In the model, the effects of one additional exposure (O. T. S.) and non-exposure during a time period have been treated as two first order Markov Chains. The Makovian assumptions have been tested in real life situations and are found to hold reasonably well. In its present form, the model can be used by practitioners to cross check their own judgements about the desired minimum reach and frequency, required for the MRF model.

Finally, the dissertation has suggested areas for future research in the field.