The demographic variables selected are ones which consistantly impact customer and prospect segmentation in our most complex models. We often look at over 200 variables, but these area the ones that matter most. We are continuously working on interesting new variables. We built "country club density" for an upscale menswear catalog. We also built "graphic artists density" for a business-to-business catalog. To keep things simple, we are just using these 12 but we're open to questions.