1 6 What little I know about marketing hen they told me I had won the Parlin Award for Marketing, I W thought they were kidding. I cannot even understand what the experts write on the subject. Stuff like this from Professor Paul Warshaw of McGill: Though use of sample cross-validated correlations is acceptable, the infrequently used squared population cross-validated correlation 2 coefficient ( ) is a more precise (although slightly biased) measure (Cattin 1978a, b; Schmitt, Coyle, and Rauschenberger 1977). It utilizes all available data simultaneously rather than bisecting the sample into arbitrary estimation and holdout components. Because 2 of these comparative advantages, is used in the present analysis. Though several versions are available, Srinivasan’s (1977) 2 formulation is acceptable for models containing fixed predictor 1 variables. If you can understand this kind of thing, you may find it useful to look up other models of consumer behavior, such as Lavidge and Steiner, Andreason, Nicosia, Engel-Kollat-Blackwell, Howard and Sheth, and Vaughan. All double Dutch to me. However, thirty odd years of rubbing shoulders with marketing practitioners has taught me some things which have helped in my work. New products About 35 per cent of supermarket sales come from products which did not exist ten years ago. You can judge the vitality of a company by the number of new products it brings to market. I have known Chief Executive Officers who made enough profit from the products they inherited from their