24-08-2012, 02:53 PM
A Market-Wide Analysis Accounting for Endogeneity
A Market-Wide Analysis.pdf (Size: 227.57 KB / Downloads: 74)
ABSTRACT
This paper studies the effects of loyalty programs on behavioral loyalty using market-wide
panel data on supermarket purchases. We present a conceptual framework relating loyalty
program membership to share-of-wallet. To account for the two-sided causality between these
variables, we specify a model for the loyalty program decision and use instrumental variables
such as privacy concerns and loyalty program enjoyment. In general, loyalty programs are
shown to influence share-of-wallet positively; ignoring endogeneity leads to substantially
overestimate of effects. Not all loyalty programs are effective, however; four out of seven
loyalty programs under study give greater value away to their members than they earn back in
terms of additional customer revenues. The effectiveness of loyalty programs increases with
value given through saving features. In contrast, the effectiveness of a program diminishes
with higher price discounts to program members.
INTRODUCTION
In recent years many retail companies have introduced loyalty programs with the goal of
enhancing customer revenues. Loyalty programs are currently available in many industries --
such as supermarkets, gasoline stations, and clothing stores -- and have become almost
common property in some of them (Lewis 1997). Loyalty programs provide members with
benefits such as promotions and saving features, which make these programs popular among
consumers (Liebermann 1999). In the United States, almost 80% of all households have at
least one supermarket loyalty card, and in Canada this is even 90% (ACNielsen, Consumer
Insight 2002). In addition, many households are members of multiple loyalty programs
simultaneously (ACNielsen, Consumer Insight 2002). Although loyalty programs have
attained a prominent position in the marketing-mix of retailers, and loyalty cards have
infiltrated customers’ wallets, it is still not evident whether loyalty program memberships
actually affect purchasing behavior (Dowling and Uncles 1997). Extant empirical research
provides mixed evidence of their effectiveness. Some previous studies found positive effects
of retail loyalty programs on customer purchasing (Bell and Lal 2003; Mägi 2003), whereas
others report the absence of any effects (DeWulf, Odekerken-Schröder, and Iacobucci 2001;
Sharp and Sharp 1997). This ambiguity relates, at least partly, to the fact that data and
methodology limitations hinder proper assessment of the effects of loyalty programs.
CONCEPTUAL FRAMEWORK
This section presents a framework that makes it possible to study empirically the
effectiveness of loyalty programs in a retail market. In line with previous research (e.g. Sharp
and Sharp 1997; Yi and Jeon 2003), we define a loyalty program as an integrated system of
marketing actions that aims to make member customers more loyal. To take advantage of the
loyalty program, a customer must become a member and must identify himself as such at
every purchase occasion. In retailing, most loyalty programs use loyalty cards for
identification and registration.
Customer Loyalty
Loyalty programs aim primarily to stimulate customer purchasing at the retail company in a
structural way, which comes down to improving behavioral loyalty. The marketing literature
provides a wide range of loyalty measures (Odin, Odin, and Valette-Florence 2001), the
usefulness of which depends on the specific market and study objective. In grocery retailing,
purchasing behavior is characterized by high buying frequency and variation in basket sizes
(Kahn and Schmittlein 1992). Further, consumers are often regular buyers at different
companies (Kahn and McAlister 1997), a phenomenon referred to as polygamous loyalty
(Dowling and Uncles 1997). Given these characteristics, share-of-wallet is the most suitable
measure for behavioral loyalty (Berger et al. 2002; Mägi 2003). Share-of-wallet measures the
share of category expenditures spent on purchases at a certain company, which integrates
choice behavior and transaction sizes during a certain time period into one single measure.
Equivalently, in brand management the share-of-category requirements is often used,
calculated as the average proportion of the brand’s customers’ category purchases over all
customers, as a measure of brand loyalty or even brand health (Ailawadi, Lehmann, and
Neslin 2001; Bhattacharya et al. 1996).