Article

Spatial Models in Marketing

Marketing Letters (Impact Factor: 1.06). 02/2005; 16(3):267-278. DOI: 10.1007/s11002-005-5891-3
Source: RePEc

ABSTRACT Marketing science models typically assume that responses of one entity (firm or consumer) are unrelated to responses of other entities. In contrast, models constructed using tools from spatial statistics allow for cross-sectional and longitudinal correlations among responses to be explicitly modeled by locating entities on some type of map. By generalizing the notion of a map to include demographic and psychometric representations, spatial models can capture a variety of effects (spatial lags, spatial autocorrelation, and spatial drift) that impact firm or consumer decision behavior. Marketing science applications of spatial models and important research opportunities are discussed. Copyright Springer Science + Business Media, Inc. 2005

Download full-text

Full-text

Available from: Neeraj Arora, Aug 06, 2015
3 Followers
 · 
168 Views
  • Source
    • "For decades, management and marketing researchers have been increasingly using spatial statistical techniques to solve marketing problems (Bradlow et al. 2005). These applications of spatial statistics can be divided into two distinct categories — those based on physical geographical correlations and those based on more abstract inter-personal correlations. "
    [Show abstract] [Hide abstract]
    ABSTRACT: Our research utilizes revenue–business-based relationships and data to expand the donor bases of non-profit organizations. Fundraisers desire to predict who will donate and how much to allocate their marketing resources effectively. To answer both questions, we develop the Spatial Tobit Type 2 (ST2) model that integrates the auto-Logistic (AL) and auto-Gaussian (AG) models into the Tobit type 2 framework. The AL component is used to predict who is likely to donate by inferring inter-client similarities based on the clients' transaction information from the revenue businesses. Similarly, the AG component is used to predict how much based on a similar measure of inter-client similarities. The Tobit type 2 framework combines both components into the single framework of ST2. Our empirical application linking a veterinary school's medical treatment records to its donation records demonstrates that clients' relationships built through their medical treatments at the school hospital positively contribute to their donation decisions.
    Journal of Interactive Marketing 05/2013; 27(2):112–129. DOI:10.1016/j.intmar.2012.10.002 · 1.68 Impact Factor
  • Source
    • "Several factors not explicitly included in the model and thus relegated to the error terms introduce spatial dependence (Bradlow et al. 2005; Bronnenberg 2005; Chintagunta, Dubé, and Goh 2005). Such unobserved factors emerge because neighboring " regions … often share climate, resources, history and sociodemographic and economic makeup. "
    [Show abstract] [Hide abstract]
    ABSTRACT: This article develops a method for optimal allocation of resources based on an empirically validated model of how national and regional advertising generate sales over time. The authors derive the profit-maximizing total budget, its optimal split between national and regional spends, and its optimal allocation across multiple regions. They formulate a spatiotemporal model that accounts for spatial and serial dependence, spatial heterogeneity, neighborhood effects, and sales dynamics. Because of spatial and serial dependence, correlated multivariate Brownian motion drives the sales dynamics, resulting in a second-order differential equation for the Hamilton—Jacobi—Bellman (HJB) equation with multiple states (i.e., regional sales) and multiple controls (i.e., regional and national advertising expenditures). By solving the HJB equation analytically, the authors furnish closed-form expressions for the optimal total budget and its regional allocations. In addition, they develop a method to estimate the proposed model and apply it to market data from a leading German cosmetics company. Using the estimated parameters, they evaluate the optimal budget and allocations. Comparing them with actual company policy, the proposed approach enhances profit by 5.07%, and it not only identifies which regions under-or overspend but also reveals how much budget to shift from national to regional advertising (or vice versa).
    Journal of Marketing Research 02/2012; 49(1):1. DOI:10.2307/23142854 · 2.52 Impact Factor
  • Source
    • "Several factors not explicitly included in the model and thus relegated to the error terms introduce spatial dependence (Bradlow et al. 2005; Bronnenberg 2005; Chintagunta, Dubé, and Goh 2005). Such unobserved factors emerge because neighboring " regions … often share climate, resources, history and sociodemographic and economic makeup. "
    [Show abstract] [Hide abstract]
    ABSTRACT: How should brand managers determine the optimal advertising budget to generate sales and maximize profit from multiple regions and over time? How much of it should be set aside for national advertising? How should they allocate the rest across multiple regions? This paper addresses these questions by developing a method for optimal allocation of resources based on an empirically validated model of how national and regional advertising generate sales over time. We derive the profit-maximizing total budget, its optimal split between national and regional spends, and its optimal allocation across multiple regions. To this end, we formulate a spatiotemporal model that accounts for spatial and serial dependence, spatial heterogeneity, neighborhood effects and sales dynamics. Because of spatial and serial dependence, correlated multivariate Brownian motion drives the sales dynamics, resulting in a second-order differential equation for the Hamilton-Jacobi-Bellman (HJB) equation with multiple states (i.e., regional sales) and multiple controls (i.e., regional and national advertising expenditures). By solving the HJB equation analytically, we furnish closed-form expressions for the optimal total budget and its regional allocations. In addition, we develop a method to estimate the proposed model and apply it to market data from a leading German cosmetics company. Using the estimated parameters, we evaluate the optimal budget and allocations. Comparing them with actual company policy, the proposed approach enhances profit by 5.07%. Finally, the proposed method not only identifies which regions under- or over-spend, but also reveals how much budget to shift from national to regional advertising (or vice versa).
    SSRN Electronic Journal 04/2011; DOI:10.2139/ssrn.1873570
Show more