A Comparison of Algorithms for TV Ad Targeting

Television is the largest advertising category in the United States with 70 billion spent by advertisers per year. We compare a variety of different targeting algorithms, ranging from the traditional age-gender targeting methods employed based on Nielsen ratings, to new approaches that attempt to target high probability buyers using Set Top Box data. We show the performance of these different algorithms on a real television campaign, and discuss the advantages and limitations of each method. In contrast to other theoretical work, all methods presented in this paper are implementable on current television delivery systems.