Catalog creative design and consumer demand: A spatial distance-metric approach

Retail mail-order catalogs seek to define a niche and develop a brand in order to set themselves apart from the competition. Development of catalog niches and brands are typically achieved by unique selection of merchandise and effective use of catalog creative designs. In this paper, we focus on the unexplored problem of catalog product demand modeling in a spatial context and examine methods to study how catalog creative can be optimized to increase a catalog's profitability and cross-merchandising potential.; We propose a reduced-form, aggregate-level demand specification that addresses the unique context of catalog product marketing and creative design issues with multiple catalog versions, multiple product categories, and hundreds of item alternatives. The demand model accounts for variety in consumer choices, and allows for flexibility in inter-dependency effects between substitute or complementary products sold. Product characteristics are specified in the model such that cross-price elasticities of demand depend on the distance between products in characteristics space. To model products' spatial differentiation, we operationalize a set of continuous and discrete distance metrics that can be used to quantify aspects of catalog creative designs and products. Based on findings from behavioral marketing research, we lay out propositions on how the proposed distance metrics may influence price responses. We utilize a two-stage estimation procedure to estimate our model parameters while accounting for spatial correlations and price endogeneity problems in a catalog retailing context.; We estimate our demand model for catalog merchandise using a data set from a catalog-marketing firm selling fashion apparel. Model estimation results show that most own-price and cross-price effect variables in our demand specification have significant coefficients that are in line with prior propositions. The price coefficient is negative and significant, while own-price sensitivity increases with the Euclidean distance of an item's price to photograph location. The most significant measures of product rivalry or complementarity are those related to apparel category discrete metrics. The strengths of these competitive or complementarity relationships are moderated by products' within-spread and between-spread distances apart in the catalog. Model implications with regard to price elasticities of demand and creative optimization impacts are further discussed.