Model of Brand Choice with a No-Purchase Option Calibrated to Scanner-Panel Data

Abstract

In usual practice, researchers specify and estimate brand-choice models from purchase data, ignoring observations in which category incidence does not occur (i.e., no-purchase observations). This practice can be problematic if there are unobservable factors that affect the nopurchase and the brand-choice decisions. When such a correlation exists, it is important to model simultaneously the no-purchase and the brand-choice decisions. The authors propose a model suitable for scanner-panel data in which the no-purchase decision depends on the price, feature, and display of each brand in the category and on household stock of inventory. They link the no-purchase model to the brandchoice outcome through marketing-mix covariates and through unobservables that affect both outcomes. The authors assume that model parameters are heterogeneous across households and allow for a flexible correlation structure between the coefficients in the no-purchase model and those in the brand-choice model. The model formulation is more general than what is possible from either a nested logit model or a translog utility model and from models in which the no-purchase outcome is an additional outcome with the deterministic component of its utility set equal to zero. The authors estimate the proposed model using Bayesian Markov chain Monte Carlo estimation methods. They then apply the estimation methods to scanner-panel data on the cola product category and compare the results with those from the widely used nested logit model. L L J J J J ( )~( ) , † †

Estimation with the nested logit model: specifications and software particularities

“…Not every single scaling parameter is estimated, but the parameters of their distribution instead. The nested logit approach is predominantly used in the field of transportation research and logistics (Train, 1980;Bhat, 1997;Knapp et al, 2001), but can also be appropriate for marketing issues (Kannan and Wright, 1991;Chintagunta, 1993;Chintagunta and Vilcassim, 1998;Guadagni and Little, 1998;Chib et al, 2004). The nested logit model is the most often used hierarchical model in marketing (Suárez et al, 2004) and can be used for modelling in any situation where subsets of alternatives share unobservable utility components (Ben-Akiva and Lerman, 1985).…”

“…The nested logit model is the most often used hierarchical model in marketing (Suárez et al, 2004) and can be used for modelling in any situation where subsets of alternatives share unobservable utility components (Ben-Akiva and Lerman, 1985). In the field of marketing the nested logit model is mainly applied in brand choice modelling (Kamakura et al, 1996;Ailawadi and Neslin, 1998;Guadagni and Little, 1998;Sun et al, 2003;Chib et al, 2004), where brands are nested, for example, regarding manufacturer (Anderson and de Palma, 1992); in a purchase incidence decision (Chintagunta, 1993;Chintagunta and Vilcassim, 1998); or regarding brand type (Baltas et al, 1997). One important point to make is that the nested logit model is a combination of standard logit models.…”

Purchase-Frequency Bias in Random-Coefficients Brand-Choice Models

“…For example, a segmentation study of customers of Mobil Oil (Forsyth, Gupta, Haldar, Kaul, and Kettle 1999) found that price shoppers spent an average of $700 annually, whereas price-insensitive, heavier users spent as much as $1,200. Similarly, Chib, Seetharaman, and Strijnev (2004) found that the parameters explaining category purchases are correlated with the parameters explaining brand-choice decisions.…”

“…The literature describes two approaches to modeling this dependence. One of these approaches is to jointly model brand-choice and category-purchase incidence (Chib et al 2004;Chintagunta 1993;Chiang 1991) or purchase timing (Chintagunta and Prasad 1998), and the other is to jointly model brand-choice and purchase frequencies (Dillon and Gupta 1996). We propose a "conditional likelihood" approach to modeling the dependence.…”

“…Another stream of work has modeled the brand choice and category purchase (incidence) decisions jointly (e.g., Chib et al 2004;Bucklin and Gupta 1992). A natural question of interest is whether there is a similar source of bias in those models.…”