Getting Real About Prediction in Marketing Research
By Jan Hofmeyr, International Director of Innovation, Synovate Brand and Communications Practice, Cape Town
I begin this paper with a meta-analysis of the predictive performance of customer satisfaction, purchase intention, and advertising recall. I show that it is poor. Developments in cognitive neuroscience and evolutionary psychology help us to understand why. I suggest two routes to improvement: first, through the development of better advertising diagnostics using a new measure of emotional engagement coupled to text mining tools; and second, through the development of a new measure of attitudinal equity based on the Zipf distribution. I finish the paper with a brief discussion about the conservatism of our industry and suggest the need for some changes in orientation.
Download white paper (PDF, 202 Kb)