Differences between consumers in sensitivity to marketing mix variable
s have been extensively documented in the scanner panel data. All stud
ies of consumer heterogeneity focus on a specific category of products
and ignore the fact that the purchase behavior of panel households is
often observed simultaneously in multiple categories. If sensitivity
to marketing mix variables is a common consumer trait, then one should
expect to see similarities in sensitivity across multiple categories.
The goal in this paper is to measure the covariance of both observed
(linked to measured characteristics of households) and unobserved hete
rogeneity in marketing mix sensitivity across multiple categories. Mea
surement of correlation in sensitivities across categories will serve
to guide the interpretation of the literature on household heterogenei
ty. If there is a large correlation, one can be more confident that se
nsitivity to marketing variables is a fundamental household property a
nd not simply a category-specific anomaly. Detection of correlation in
sensitivities across categories requires an appropriate methodology t
hat can handle the high dimensional covariance structures and properly
account for uncertainty in estimation. For example, a simple approach
might be to fit a brand choice model to each of the available categor
ies in turn, ignoring the data in the other categories. For each categ
ory, household parameter estimates could be obtained for the parameter
s corresponding to price, display, and feature sensitivity. These para
meter estimates could be viewed as data and the correlations across ca
tegories could be computed. Such a procedure could induce a downward b
ias in the estimation of correlation due to the independent sampling e
rrors, which are present in each parameter estimate. We develop a hier
archical model structure that introduces an explicit correlation struc
ture across categories and utilizes the data in multiple categories at
the same time. To reduce the size of the covariance matrix, we use a
variance components approach. We introduce household-specific demograp
hic variables to decompose the correlation across categories into that
which can be ascribed to observable and unobservable sources. Shoppin
g behavior variables such as shopping frequency and market basket size
as well as intensity of shopping in a category are also included in t
he model. Using data on five categories, we find substantial and stati
stically important correlations ranging from .32 for price sensitiviti
es to .58 for feature sensitivity. These correlations are much larger
than the correlations obtained with the state-of-the-art techniques av
ailable prior to our work. We attribute our ability to detect substant
ial correlations to our method, which involves the joint use of multip
le category data in a parsimonious and efficient manner. Unlike previo
us studies with panel data, household demographic variables are found
to be strongly related to price sensitivity. Higher income households
are less price sensitive and large families are more price sensitive.
Shopping behavior variables are also important in explaining price sen
sitivity. Households that visit the store often are more price sensiti
ve. Households with larger market baskets are less price sensitive, co
nfirming the view of Bell and Lattin (1998). Heavy user households ten
d to be both less price sensitive and less display sensitive. The evid
ence presented here of substantial correlations validates, in part, th
e notion that sensitivity to marketing mix variables is a consumer tra
it and is not unique to specific product categories. It also opens the
possibility of using information across categories in making inferenc
es about consumer brand preference and marketing mix sensitivity, prov
iding a richer source of information for target marketing.