An evaluation and optimization model for media planning

A major objective of any media planner is to decide on the ideal combination of advertisement insertions in various media so as to maximize some measure of advertising effectiveness. To achieve this, there is a need for an appropriate method of evaluation which would derive parameters such as Reach, Average Opportunities To-See (AOTS), Cost Per Person Reached (CPPR) and such other factors. This paper presents an evaluation model based on the binomial distribution for computing these basic parameters. An optimization model using dynamic programming is developed for determining the ideal combination of advertisement insertions that maximizes Reach. The models are developed for use in conjunction with the National Readership Survey (NRS) database of India but can be easily adapted to other databases also. Algorithms of these models have been effectively utilised as software on personal computers by a leading advertising agency in South India.