Journal of Retailing 79 (2003) 213224

An empirically derived taxonomy of retailer pricing

and promotion strategies
Ruth N. Bolton a, , Venkatesh Shankar b,1
a
b

Owen Graduate School of Management, Vanderbilt University, Nashville, TN 37203, USA

Robert H. Smith School of Business, University of Maryland, College Park, MD 20742, USA
Accepted 2 September 2003

Abstract
Most research categorizes grocery retailers as following either an Every Day Low pricing (EDLP) or a High Low (Hi-Lo) pricing strategy
at a store or chain level, whereas this paper studies retailer pricing and promotions at a brand-store level. It empirically examines 1,364
brand-store combinations from 17 chains, 212 stores and six categories of consumer package goods in five U.S. markets. Retailer pricing and
promotion strategies are found to be based on combinations of four underlying dimensions: relative price, price variation, deal intensity and
deal support. At the brand-store level, retailers practice five pricing strategies, labeled Exclusive, Moderately Promotional, Hi-Lo, EDLP, and
Aggressive pricing. Surprisingly, the most prevalent pricing strategy is not Hi-Lo pricing strategy as is widely believed. It is one characterized
by average relative brand price, low price variation, medium deal intensity, and medium deal support. The findings provide some initial
benchmarks and suggest that retailers should closely monitor their competitors price decisions at the brand level.
2003 by New York University. Published by Elsevier. All rights reserved.
Keywords: Retailing; Pricing; Promotion; Strategy; Multivariate data analysis

Introduction
Most research categorizes grocery retailers as following
either an Every Day Low pricing (EDLP) or High Low
(Hi-Lo) pricing Strategy at the store or chain level (Bell, Ho,
& Tang, 1998; Bell & Lattin, 1998; Hoch, Purk, & Dreze,
1994; Lal & Rao, 1997; Partch, 1992). However, many market researchers have observed that grocery retailers pricing strategies and tactics are diverse and complex, including decisions on the depth, frequency, and duration of deals,
feature advertising, and displays for myriad brands and categories (Dhar & Hoch, 1997; Hoch et al., 1994; Levy &
Weitz, 1998). The purpose of this paper is to empirically
investigate grocery retailer pricing and promotion strategies
by analyzing pricing and promotion decisions for an assortment of brands and categories at different stores and markets. This study examines retailer promotion decisions
specifically deal intensity (depth of deal discount, frequency,
and duration) and deal support (features and displays)as

Corresponding author. Tel.: +1-615-322-5580; fax: +1-360-838-1844.

E-mail addresses: [email protected] (R.N. Bolton),
[email protected] (V. Shankar).
1 Tel.: +1-301-405-2175; fax: +1-301-405-0146.

well as pricing decisions (relative price and price variation).

We explore the underlying dimensions of retailer pricing decisions, classify the different types of pricing strategies, and
characterize their prevalence across the brands and stores in
our sample.
Our study investigates two broad research questions:
1. Are there a small number of stable underlying dimensions that characterize grocery retailers observed
pricing decisionsdespite the fact that these decisions
appear to be very complex and different across brands,
categories, and stores?
If the answer to this question is yes, the identification of these dimensions will significantly enhance
our understanding of retailer strategies. In particular, it
will shed insight into how retailers observed price and
price-promotion decisions are developed.
2. What are the different types of pricing strategies adopted
by grocery retailers?
Is a classification based on storewide policies (e.g.,
EDLP/Hi-Lo) sufficient for decision-making, or does a
classification scheme based on brand-store combinations
provide additional information? What pricing strategies are
most prevalent among retailers at the brand-store level?

0022-4359/$ see front matter 2003 by New York University. Published by Elsevier. All rights reserved.
doi:10.1016/j.jretai.2003.09.005

214

R.N. Bolton, V. Shankar / Journal of Retailing 79 (2003) 213224

How closely coordinated are pricing and promotion (both

deal intensity and support) decisions?
We address these questions by developing an empirically
based taxonomy for grocery retailers pricing and promotion
decisions. Prior research has developed theory based models that describe firms pricing behavior (cf. Tellis, 1986), as
well as normative models that prescribe how firms should
behave with regard to pricing (Lal & Narasimhan, 1995).
There are also descriptive studies of retailer pricing (e.g.,
Hulbert, 1981) and promotional strategies (e.g., Blattberg &
Neslin, 1989; Fader & Lodish, 1990). Recent research focuses on how retailers pricing decisions (especially pricing levels) are related to sales, customer variables, manufacturer (e.g., Farris & Albion, 1981; Lal & Villas-Boas, 1998),
category, competition, and market factors (e.g., Shankar &
Bolton, 2004). A few studies have also developed optimal price and promotion models (e.g., Achabal, McIntyre,
& Smith, 1990; Shankar & Krishnamurthi, 2003; Tellis &
Zufryden, 1995). In contrast, we do not develop a theory
of how retailers should make pricing decisions or identify
market outcomes (e.g., retailer price levels or dispersion).
Instead, the major contributions of our study are:
An empirical identification of the dimensions of grocery
retailer pricing strategy
Focusing on pricing and promotion (i.e., deal intensity
and support) decisions at the brand-store level, rather than
store/chain-wide or brand-specific pricing decisions
Yielding a classification of retailer pricing strategies and
a description of their prevalence within our study sample.
Our empirical analysis is based on grocery store-level
scanner data on 1,364 brand-store combinations from 17
chains, 212 stores and six categories of consumer package
goods in five U.S. markets. We conduct the analysis in two
stages. First, we analyze retailers observed pricing decisions for different brand-store combinations using principal
component analysis and find that they reflected four underlying dimensions: price variation, relative (brand) price,
deal intensity, and deal support. Second, we identify different types of retailer pricing strategy by grouping brand-store
combinations and stores along these dimensions using cluster analysis. In our concluding remarks, we discuss how
an understanding of underlying retailer pricing dimensions,
combined with our taxonomy of retailers pricing strategies,
can provide benchmarks for retailers and manufacturers to
evaluate their decisions.

Perspective on retailer pricing

Although market researchers have observed that retailer
pricing can be quite different across brands and stores, there
is little research on this issue. A notable exception is a conceptual article by Tellis (1986) that develops a taxonomy
that is intended to be applicable to all firms, not just retailers.
His pricing strategies are: differential pricing (random dis-

counting, periodic discounting, second market discounting),

competitive pricing (price signaling, penetration pricing or
experience curve pricing, and geographic pricing) and product line pricing (image pricing, price bundling or premium
pricing, and complementary pricing). He explicitly considers how these strategies differ depending on the existence
of consumer segments, competitors in the market, and product mix. He does not consider the complementary role of
promotion.
In this study, we are particularly interested in developing
a taxonomy based on grocery retailers actual pricing and
promotion decisions. Relevant empirical research on retailer
pricing and promotion can be grouped into two streams: (1)
studies that have examined the determinants of price and
promotional elasticities (e.g., Bolton, 1989; Hoch, Kim,
Montgomery, & Rossi, 1995; Kirande & Kumar, 1995;
Mulhern, Williams, & Leone 1998; Narasimhan, Neslin,
& Sen, 1996; Shankar & Krishnamurthi, 1996) and (2)
studies that have considered how retailer pricing and promotion tactics are related to purchase behavior, consumer
variables, competition, and market factors (e.g., Fader &
Lodish, 1990; Shankar & Bolton, 2004). Our research complements the latter research stream in two important ways.
First, we identify the underlying strategic dimensions of
retailer pricing and promotion decisions across multiple
brands, categories, stores, and markets. Second, since pricing decisions are made at different levelstypically the
brand level and sometimes the brand-size level (Kumar &
Divakar, 1999)we identify clusters of underlying pricing
and promotion dimensions at both the store and brand level
and describe their prevalence.
Although supermarket chains have category managers
(Basuroy, Mantrala, & Walters 2001), our study focuses on
price and promotion decisions at the brand level for each
store (rather than each category level for each chain).2 This
decisiona departure from most prior researchis based
on the following rationale. The most distinctive features of
market-driven organizations are their mastery of the market
sensing and customer linking capabilities that span the spectrum between external environment and the company, such
as pricing (Day, 1994). Hence, if retailers have a market orientation, their pricing decisions are likely to be customized
to reflect differences in the stores clientele and competitive
environment (e.g., Alba et al., 1994; Moriarty, 1985; Urbany
& Dickson, 1990, 1991), as well as for each brand and
category (e.g., Farris & Albion, 1981). Retailers may also
price and promote differently for store and national brands
(Ailawadi, Neslin, & Gedenk, 2001; Sethuraman, 1996), but
the differences between national and store brands are not the
focus of our study. Instead, we focus on how retailers typically formulate pricing strategies based on their knowledge
of consumer store choice, as well as brand choice and quan2 Since the prices of different flavors or colors of a brand (e.g., mouthwash or waffle SKUs, bathroom tissue SKUs) are often the same, price
decisions tend to take place at the brand level rather than the SKU level.

R.N. Bolton, V. Shankar / Journal of Retailing 79 (2003) 213224

tity decisions. Store choice is driven by store location, destination categories, store price perception, and store service
perception (Kumar & Leone, 1988). Some categories serve
as traffic-builders (Walters & Mackenzie, 1988) and others
act as cash cows. Thus, retailers are likely to have different
pricing strategies for different category-store combinations
rather than have just one storewide pricing strategy.
To verify our prediction that retailers are customizing their
pricing decisions for different brand-store combinations, we
conducted interviews with the marketing and category managers of a few retail chains. These managers suggested that
retailers are likely to make brand price decisions (as well
as category and store-level decisions) on the basis of visible
competitive activity, such as price and deal activity. Next,
we examined raw brand prices in a given week in our data
base (described in the following) and ascertained that they
vary across stores in the same chain, and across chains in
a market. We also examined the correlations among brand
prices in a given category across stores in the same chain, and
chains in the same market over the period of the data. These
correlations were significantly low, leading us to conclude
that the brand decisions are not always jointly determined
at the category level. Hence, our preliminary investigations
indicated that there is variation in retailer pricing decisions
at brand level, so our analyses begin by measuring price and
promotion decisions for each brand-store combination.

Underlying dimensions of retailers pricing decisions

This study begins by investigating whether retailers pricing decisionscomplex decisions that appear very different
across brands, categories and storescan be captured by a
parsimonious set of stable underlying dimensions. In this
section, we describe our database, explain how we calculated
measures of retailers pricing decisions, and then conduct a
principal components analysis (PCA) of these measures.
The database
The database consists of multi-brand, multi-category,
multi-store scanner data drawn from six categories of
consumer-packaged goods in five U.S. markets that include
information about pricing and promotion at the retail level.
The categories are spaghetti sauce, bathroom tissue, liquid
bleach, ketchup, mouthwash, and frozen waffles. The cities
are New York, Los Angeles, Chicago, Marion (IN), and
Pittsfield (MA) and are thus fairly representative of both
large and small markets in the United States. The database
describes all major brands and stores in these categories
and markets.3 There are 17 chains and 212 stores in the
3 To preserve the confidentiality of the data, this paper does not link
pricing strategies to particular brands and stores. However, the database
covers all major brands and stores in these categories and markets. Chains
include Albertsons, A&P, Dominicks, Food Lion, Jewel, Kroger, Lucky,

215

database. Altogether, the database yields 1,364 brand-store

combinations. The weekly store-level scanner data were
obtained from two sources, A.C. Nielsen Company and Information Resources, Inc. (IRI). Merging two different data
sources makes it possible to uncover systematic patterns
that exist across the different data collection and measurement conditionsincreasing our ability to generalize from
the study findings. To uncover stable underlying pricing dimensions and strategies, we study average pricing decisions
over a two-year period (a maximum of 121 weeks in any
particular store).4
The umbrella categories for these six categories are:
frozen breakfast foods (i.e., waffles), oral care (i.e., mouthwash), paper (i.e., bathroom tissue), laundry care (i.e.,
bleach), condiments (i.e., ketchup), and pasta (i.e., spaghetti
sauce). They are large categories and their roles represent
much of the spectrum of category roles in a typical store.
This notion is amply supported by penetration and frequency of purchase data for these categories among U.S.
households (IRI Category Report, 1998). It is also supported by qualitative information obtained in an interview
conducted with the Marketing Director of a leading grocery
chain in the United States. He described category roles in
terms of the combination of the importance of sales and
profit margins. The importance of sales and profits can be
low or high, yielding four combinations: (1) support role
comprising low sales and low profits, (2) preferred role
consisting of low sales, but high profits, (3) destination
role comprising high sales, but low profits, and (4) ideal
role consisting of high sales and high profits. As shown in
Table 1, these particular categories are distributed across
the four different category roles. Hence, we believe that
these brands and categories are somewhat representative of
retailers product assortments, allowing us to make useful
generalizations about retailers pricing strategies.
Conceptualization and measurement of pricing decisions
A proposed conceptual framework that identifies the underlying pricing dimensions to their measures is shown in
Fig. 1. We begin by developing granular measures that reflect retailers decisions concerning regular and deal prices.
We do not consider market measures such as absolute retail
price levels and price dispersion across retailers. Instead, we
focus on measures of retailer pricing policy or formatthat
Pathmark, Stop and Shop, Safeway, and Von. Bathroom tissue brands include Charmin, Cottonelle, Northern, Scott, Waldorf, and White Cloud.
Liquid bleach brands include Chlorox and Purex. Ketchup brands include Del Monte, Heinz and Hunts. Mouthwash brands include Listerine,
Plax and Scope. Spaghetti sauce brands include Prego and Ragu. Waffle
brands include Aunt Jemima, Downyflake, Eggo, and Roman Meal. The
database also includes private label and generic brands in these stores
and categories.
4 Average pricing measures are obtained by taking arithmetic means of
the weekly measures subsequently provided in Table 2 over the period
of the data.

216

R.N. Bolton, V. Shankar / Journal of Retailing 79 (2003) 213224

Table 1
Category roles in the store

Low/high is relative to other categories in the store. Sales are in dollars, but are typically correlated with unit sales. Profit margins pertain to the category.

is, on specific measures of price and price-related promotion decisions. Retailer pricing policy or format has typically
been labeled EDLP or Hi-Lo (Hoch et al., 1994). An EDLP
policy involves offering consistently low prices on many
brands and categories and is practiced by some supermarkets
(e.g., Food Lion and Lucky). A Hi-Lo policy is characterized
by steep temporary price discounts on high regular prices
for many brands and categories and is adopted by other supermarkets (e.g., Kroger and Safeway). An EDLP policy
tends to draw price sensitive shoppers, whereas a Hi-Lo policy often attracts cherry pickers (e.g., Lal & Rao, 1997). A

Dimensions

Price
Variation

store selects and communicates a pricing policy (EDLP or

Hi-Lo) to signal the underlying consistency of its prices to
consumers.
Our study departs from prior research concerning
retailers pricing strategies in four significant ways. First,
most prior studies consider pricing formats, such as
EDLP/Hi-Lo, as storewide policies, whereas we develop
measures of retailers pricing strategies that are specific to
the brand-store combination. Second, most prior studies
view pricing policy as a dichotomous variable (EDLP or
Hi-Lo). However, recent evidence suggests that EDLP and

Observed Decision Variables

Coefficient of Variation in Regular Price

Deal Depth

Deal
Intensity

Deal Frequency
Deal Duration

Feature Support

Deal
Support

Display Support
Feature and Display Support

Relative Price

Brand Price Relative to Weighted Average

Category Price

Fig. 1. Dimensions of retailer pricing strategy.

R.N. Bolton, V. Shankar / Journal of Retailing 79 (2003) 213224

217

Table 2
Retailer pricing dimensions: measures and descriptive statistics
Pricing dimensions

Measuresa

Relative price
Average actual price of the brand relative to other
brands in the category.

Single variable/measure
Average of brand price divided by the weighted average
category price (where the weights are market shares within the
store), over all the weeks.

Price variation
Extent to which a retailer follows a pricing
policy/format that is EDLP on one end and Hi-Lo
on the other end of the continuum.

Single variable/measure
Coefficient of variation: standard deviation of the brand price
divided by its mean over all the weeks (reverse coded in sign
so that large numerical values imply less variation).

Deal intensity
The depth, frequency, and duration of price cuts or
deal discounts for a given brand at the retail level.

Four variables/measures
(1) Deal depth 1: average deal depth (in cents) across all weeks,

Mean (SD)

(2) Deal depth 2: average deal depth (in cents) across only
deal weeks,
(3) Deal frequency: percentage of weeks with deals,
(4) Deal duration: average deal duration (in weeks).
Each brand-store average is normalized by dividing by the
category average to make it comparable across brand-stores.
Deal support
Complementary feature and/or display decisions for
a given brand.

Three variables/measures
(1) Feature and deal: percentage of weeks with feature and deal,
(2) Display and deal: percentage of weeks with display and deal,
(3) Feature, display and deal: percentage of weeks with
feature, display and deal.

1.00 (0.17)

0.05 (0.04)

0.12 (0.06)
0.38 (0.17)
0.35 (0.19)
0.11 (0.15)

0.08 (0.04)
0.06 (0.06)
0.03 (0.03)

Single overall measures for deal intensity and deal support are created by averaging the listed measures.

Hi-Lo maybe opposite ends of a continuum of pricing policy (e.g., Hoch et al., 1994), so we measure each retailer
pricing decision on a continuum. Third, although price variation refers to stable prices, many retailers that have stable
prices, have low stable prices, to stay competitive (e.g.,
Wal-Mart, Food Lion, and Lucky). Hence, we consider
the brands relative price level, as well as price variation.
Fourth, we believe that promotionboth deal intensity and
supportare important aspects of retailer pricing decisions.
Thus, we identify four pricing dimensions and develop multiple measures to describe decisions at the brand-store level.
Pricing dimensions, associated and descriptive statistics are
displayed in Table 2.

of the brands within the category in that store), consistent

with Bolton (1989). This measure can be interpreted as follows. Since we consider price level for a brand relative to
other brands in the category, we are implicitly choosing to
measure reference price effects that operate across brands
and stores within a category. This feature is useful because
consumer reference price is frequently considered to be a
critical component of the price level for a brand at a given
store over time (Winer, 1986), where reference prices influence brand choice at the retail level (Kumar, 1998). Note
that we do not measure absolute price levels because they
cannot be pooled and compared across the units of analysis
(stores, brands, etc.).

Relative price
To develop measures of retailer pricing decisions, we begin by distinguishing between pure price and promotion
decisions. A retailers pricing strategy for a brand includes
two pure price decisions: price level and price variation.
First, we consider how to measure the retailers decision
about a brands price level. Different stores have different price premiums or discounts for a brand relative to
(category-level) reference prices. For example, a supermarket located in an upscale neighborhood may have a different
price level for a particular brand than does a supermarket in
a blue-collar neighborhood relative to category prices at the
stores (Hoch et al., 1995). Hence, we measure the relative
price of a brand as the average (over weeks in the store) of
the ratio of the brand price divided by the weighted average of all brand prices (where the weights are market shares

Price variation
Second, we measure the price variation for a brand in a
store by calculating the coefficient of variation (i.e., the ratio
of the standard deviation of actual price over the mean actual
price) for the period of the data (Shankar & Krishnamurthi,
1996). The value for a brand-store pair is a dimensionless
ratio that enables to compare across different brand-store
combinations. The less variation in price for the brand-store
pair, the closer the ratio is to zero.
Deal intensity
A retailers pricing strategy for a brand includes two promotion decisions: deal intensity (deal depth, frequency, and
duration) and deal support (feature, display, both or neither)
decisions. First, manufacturers offer trade deals that chains
(or stores) may pass along to customersthereby influenc-

218

R.N. Bolton, V. Shankar / Journal of Retailing 79 (2003) 213224

Table 3
Correlation matrix of pricing measuresa

Relative price
Price variation
Deal depth (all weeks)
Deal depth (deal weeks)
Deal frequency
Deal duration
Feature support
Display support
Feature and display support
a

Relative
price

Price
variation

1
0.04
0.14
0.05
0.01
0.06
0.11
0.09
0.09

0.04
1
0.02
0.07
0.07
0.14
0.10
0.20
0.16

Deal intensity

Deal support

Deal depth Deal depth

Deal
(all weeks) (deal weeks) frequency

Deal
duration

Feature
support

0.02
0.07
1
0.45
0.36
0.17
0.17
0.10
0.17

1
0.46
0.12
0.07
0.15
0.19

Display
support

Feature and
display support

0.11
0.06
0.32
1
0.49
0.39
0.30
0.31

1
0.26
0.28
0.28

1
0.51
0.64

1
0.83

The numbers in bold represent the correlation among the four dimensions, namely, relative price, price variation, deal intensity, and deal support.

ing the pricing strategies of a store and its competitors.

Stiff competition and value-driven consumers have created
an environment marked by high markdowns and promotions
(Kumar & Pereira, 1997; Levy, 1999). Thus, retailers pricing strategies include decisions on deal intensitythat is,
the depth, frequency and duration of dealsultimately determining the final price paid by the consumers for a brand.
These decisions have important effects on the variability in
category sales. Higher deal depth, greater deal frequency,
and longer deal duration reflect higher overall deal intensity
for a brand in a given category and store. These tactical decisions are interrelated for a given brand or category (Alba
et al., 1994) and may be different for different brands within
a category (Tellis & Zufryden, 1995), across categories and
across stores. For example, deal frequency and deal magnitude may be negatively correlated for some brands (Alba
et al., 1994). Hence, we calculate four measures of deal intensity: two measures of average deal depth (average over
all weeks and average over only deal weeks), deal frequency
and deal duration.
Deal support
Second, a retailers pricing strategy for a brand includes
support of price discounts with newspaper features or displays (or both) during some weeks, but not in other weeks
(Blattberg & Neslin, 1989, 1990; Inman & McAlister, 1993).
Deals, if supported by features or displays, may benefit
both consumers (nearly half of whom are non-vigilant about
prices) and the retailer (Dickson & Sawyer, 1990). We believe that the deal support across multiple brands within a
category and across categories in a given store is an important complementary aspect of retailers pricing decisions.
Retailers who provide higher deal support for a brand have a
higher incidence of features and displays. We calculate three
different measures of deal support: proportion of weeks with
feature support, proportion of weeks with display support,
and proportion of weeks with feature and display support.
Thus, we are able to obtain nine retail pricing measures
from the data base: relative price, price variation, depth of
deals during all weeks, deal depth during promoted weeks,
frequency of deals, duration of deals, proportion of weeks

with feature support, proportion of weeks with display support, and proportion of weeks with feature and display support (see Tables 2 and 3). Since our focus is on stable pricing
dimensions and strategies, all nine measures are calculated
over a two-year period.
Underlying pricing dimensions identified from the principal
components analysis
Seven of the nine measures concern promotion decisions
(deal intensity and support), rather than price decisions.
Since there is likely to be redundancy in these measures, we
conducted a PCA, with a varimax rotation, to identify their
underlying dimensions. The results are displayed in Table 4.
As expected, we obtained two factors that explained 70% of
the variance. They are:
Deal intensity and deal depth (both measures), deal frequency, and duration of deals.
Deal support: frequency of deal and feature, deal and
display, and deal, feature, and display.
Our measures of deal intensity and support are similar to
those used by Kumar, Ghosh, and Tellis (1992) in their study
of repeat purchase behavior.
The PCA results indicate that retailers intensively promote some brands or categories (i.e., higher composite of
deal depth, frequency, and longer duration) and do not promote others. They also indicate that retailers coordinate the
price and promotional activities in some brands or categories
much more closely than they do in others. Thus, deal intensity and deal supporttogether with relative price and price
variation (each measured by a single item)can be considered to represent four underlying dimensions of retailing
pricing strategies.5 For the purposes of this study, we chose
to measure deal intensity and deal support by three and four
item additive indices, respectively. We used additive indices
5 PCA results need not have conformed to the deal intensity and support
classification scheme that was used in developing the measures. For
example, a single factor might have been uncovered, or display items
might have loaded together on one factor.

R.N. Bolton, V. Shankar / Journal of Retailing 79 (2003) 213224

219

Table 4
Principal components analysis results
Variable

Factor1 (deal intensity)

Factor 2 (deal support)

Average deal depth across all weeks

Average deal depth across only deal weeks
Deal frequency
Deal duration
Percentage of weeks with feature and deal
Percentage of weeks with display and deal
Percentage of weeks with feature, display, and deal
Eigenvalue
Percent variance explained

0.72
0.71
0.83
0.81
0.06
0.08
0.06
2.38
30

0.31
0.10
0.35
0.09
0.87
0.80
0.93
2.51
40

rather than factor scores because these are more easily interpretable by managers.
Examining the variability in the underlying pricing
dimensions

strategies may be specific to a brand-store combination or

simply to the store. Some strategies may be more prevalent
than others. To identify these strategies, we performed cluster analyses at both the brand-store and the store levels.
Brand-store-level strategies

We verified that there is considerable variability in

retailers positions on these four dimensions and that the
underlying pricing dimensions are relatively distinct (i.e.,
non-overlapping) in the following way. First, we conducted
an analysis of variance (ANOVA) for each pricing dimension to test for differences across stores. The F-statistic
for each dimension was significant at p < .001. The R2
ranged from a low of .20 for relative price to a high of
.43 for deal intensity. Thus, our ANOVA results indicated
the four retailer pricing dimensions (at the brand level) are
significantly different across the stores (including stores
within the same chain). We conducted the same analysis for
chains, and also found significant differences.
Next, we investigated the inter-relatedness of these four
pricing dimensions by calculating their correlations. The
correlations among the nine pricing measures, and the four
underlying pricing dimensions, are shown in Table 3. The
correlation of relative price with the other three underlying
dimensions is as follows: price variation (.04), deal intensity
(.02) and deal support (.11). The correlation of price variation with the remaining dimensions is as follows: deal intensity (.07) and deal support (.16). The correlation of deal intensity with deal support is the largest (.32) because the existence of some level of promotion is a necessary condition for
the existence of any deal support. Thus, the dimensions are
generally independent, although (not surprisingly) the correlation between deal intensity and deal support is the highest.6
Types of retailer pricing strategies
Retailers may choose different combinations of dimensions, resulting in different types of pricing strategies. These
6 If our analyses use factor scores (rather than additive indices) to
represent the pricing dimensions, we obtain substantively the same results.
The reason is that the additive indices are (almost) uncorrelated, as the
factors scores are (by definition).

The results from the k-means cluster analysis for

brand-store combinations appear in Table 5.7 The results of
an analysis of variance indicate that the means of all the
four dimensions are significantly different (p < .05) across
the five clusters.8
There are five clusters of pricing and promotion strategies
at the brand-store level, labeled: Exclusive, Moderately Promotional, Hi-Lo, EDLP, and Aggressive pricing strategies.
Table 5 shows a description of each strategy in terms of the
combinations of pricing dimensions. It also shows the distribution of brand-store combinations across the five clusters,
as well as each clusters mean scores on each of the pricing
dimensions. We classified each of the brand-store combinations as high, medium (average), or low on each of the four
pricing dimensions based on their median scores.
A Hi-Lo pricing strategy (11.2%) is characterized by average relative price, high price variation, high deal intensity,
and high deal support. We use this term because this strategy
is comparable to a storewide Hi-Lo pricing strategy, albeit at
the brand-store level. This combination of dimensions and
levels seems intended to make a retailer competitive with
its rivals primarily through promotions. In other words, retailers use in-store merchandising within categories to price
discriminate (Dhar & Hoch, 1997). Analogously, our EDLP
pricing strategy (45%) consists of average relative price, low
price variation, moderate deal intensity, and moderate deal
support. As above, we use this term because this pricing
strategy is comparable to a storewide EDLP strategy, albeit
at the brand-store level. This combination of dimensions
seems intended to offer value to customers.
7 Due to space limitations, only the aggregate results are shown. We
also did a hierarchical cluster analysis (Wards method). The results were
similar.
8 The F tests described in this paper are used only for descriptive purposes and strictly speaking, cannot be interpreted as tests of the hypothesis
that the cluster means are equal (Aldenderfer & Blashfied, 1984).

220

R.N. Bolton, V. Shankar / Journal of Retailing 79 (2003) 213224

Table 5
Pricing strategies and the mean scores on the dimensions (clustering by brand-store)a
Cluster
number (size)

Pricing dimensions
(pricing strategy/cluster)

Relative price

Price variation

Deal intensity

Deal support

1
2
3
4
5

Exclusive pricing (7.5%)

Moderately Promotional pricing (14.1%)
Hi-Lo pricing (11.2%)
EDLP pricing (44.9%)
Aggressive pricing (22.3%)

1.407
1.145
1.007
0.984
0.792

0.048
0.047
0.051
0.042
0.052

0.193
0.309
0.441
0.262
0.243

0.019
0.050
0.095
0.061
0.041

(102)
(192)
(153)
(613)
(304)
a

(high)
(average)
(average)
(average)
(low)

(medium)
(medium)
(high)
(low)
(high)

(low)
(medium)
(high)
(medium)
(lowmedium)

(low)
(medium)
(high)
(medium)
(medium)

Based on 1,364 brand-store combinations. Low, medium (average), and high labels are based on median scores on each dimension.

Hi-Lo pricing and EDLP pricing are used by about half

(56%) the brands in our database. However, the pricing and
promotional strategies of almost half (44%) the brands in
our stores do not correspond to practices currently recognized by marketing scholars and practitioners. For example,
Aggressive pricingwhich is utilized by nearly one-fourth
(22%) of all brands in our storesis not reported in the
business press. With an Aggressive pricing strategy, retailers
offer low prices and medium deal support, accompanied by
high price variation and lowmedium deal intensityprice
rather than deal is used to greater extent as the weapon in
this strategy. In summary, while chains and stores may use
these positioning or signaling strategies, retailers practice
different strategies at the brand-store level. Conventional
wisdom states that retailers practice two pricing strategies
along a continuum: EDLP and Hi-Lo pricing strategies.
However, we do not find that EDLP and Hi-Lo strategies
represent an underlying continuum. Instead, the strategies
we uncovered are combinations of the four independent
pricing dimensions, where each dimension is a separate
continuum.
EDLP pricing and Aggressive pricing are the most
commonly adopted pricing strategies at a brand-store
level. We believe this reflects the competitive nature of
the retailing landscape. Moderately Promotional pricing
(14.1%)corresponding to an undifferentiated strategyis
also fairly common. In contrast, Exclusive pricing (7.5%)
is the least adopted strategy. Since it is characterized by
low deal intensity, low deal support, and a high brand
premium, this strategy can only be profitable for a small
number of brands. We speculate that it is only appropriate for brands with high brand equity and manufacturer
advertising.

Store-level strategies
The results from the k-means cluster analysis for store
level appear in Table 6. There are five clusters of pricing and
promotion strategies at the store level. We have labeled them:
Exclusive, Premium, Hi-Lo, Low, and Aggressive pricing
strategies. Table 6 shows the combinations of pricing dimensions for each pricing strategy. It also shows the distribution
of stores across the five clusters, as well as each clusters
mean scores on each pricing dimension. Based on the results from an analysis of variance, the means of all the four
dimensions are significantly different (p < .05) across the
five clusters.
The five clusters of pricing and promotion strategies at
the store level are different from the five clusters identified
in the brand-store-level analysis. Two clusters are somewhat
similar at both levels: Hi-Lo (9.0%) and Exclusive pricing
strategies (2.3%). Note that both these strategies are infrequently practiced. The Hi-Lo strategy is characterized by
high deal intensity and support, but it has medium price
variation (vis--vis the brand-store case that is marked by
high price variation). As expected, Exclusive pricing (2.3%)
at the store level is the least adopted strategy because it
is likely to be appropriate only for stores with upscale image and high-end clientele. The remaining three store-level
strategies do not correspond to brand-store-level strategies.
Hi-Lo and Low pricing are used by about half (52%) the
stores in our database. Similar to the brand-store-level analysis, the pricing and promotional strategies of almost half
(48%) the stores do not correspond currently recognized to
practices. These other strategies include Premium pricing
(11.8%), Aggressive pricing (34.0%), and Exclusive pricing
(2.3%).

Table 6
Pricing strategies and the mean scores on the dimensions (clustering by store)a
Cluster
number (size)

Pricing dimensions
(pricing strategy/cluster)

Relative price

Price variation

Deal intensity

Deal support

1
2
3
4
5

Exclusive pricing (2.3%)

Premium pricing (11.8%)
Hi-Lo pricing (9.0%)
Low pricing (42.9%)
Aggressive pricing (34.0%)

1.402
1.167
1.062
0.922
1.019

0.012
0.044
0.043
0.045
0.049

0.150
0.275
0.380
0.259
0.265

0.023
0.052
0.082
0.058
0.050

(5)
(25)
(19)
(91)
(72)

(high)
(high)
(average)
(low)
(average)

(low)
(medium)
(medium)
(medium)
(high)

(low)
(medium)
(high)
(low)
(low)

(low)
(medium)
(high)
(medium)
(medium)

a Based on average scores of dimensions across brands in 212 stores. Low, medium (average), and high labels are based on median scores for each
pricing dimension.

R.N. Bolton, V. Shankar / Journal of Retailing 79 (2003) 213224

Low pricing and Aggressive pricing strategies are adopted

by over three-fourths of the stores, reflecting the competitive
retail environment. For example, there is no strategy corresponding to an EDLP pricing strategy cluster at the store
level. Among the five strategies identified at the store level,
the strategy closest to the commonly used term, EDLP strategy, is Low pricing (42.9%)but price variation is moderate, not low, as the phrase everyday low price would suggest. Again, we do not find that EDLP and Hi-Lo strategies
represent an underlying continuum. Instead, the strategies
we have uncovered are combinations of the four independent pricing dimensions, where each dimension is a separate
continuum.
The strategies identified herein are consistent with pricing
practices of retailers as gathered from depth interviews with
managers (who wish to remain anonymous) of four retail
chains in our sample. Based on our interviews, the pricing
practice at a retail chain can be generally summarized as
follows. A chain classifies its stores into geodemographic
groups primarily based on location, demographics, and competition. A chain has umbrella categories and categories (as
described earlier), with category managers for each. Each
category manager plans broad pricing strategies for her/his
categoryand brands within itincluding regular price
point and promotion type and level based on factors such
as past history, competition, and trade deals. These broad
strategies are made for key brands and stores, and include
such strategies or low stable prices (EDLP), or frequent promotions (Hi-Lo) or maintenance of exclusive image/price.
Although the retailers we interviewed did not quite use
terms like Exclusive pricing and Aggressive pricing, they
do think along these lines by brand and store. For example,
some of the terms that we heard category managers use to
denote Aggressive pricing included Sensitive pricing and
Deep pricing. These pricing strategies are then translated
into specific decisions such as regular price, deal discounts,
and feature for all brands and all stores within the chain at
a weekly level based on some decision calculustypically
through spreadsheets that involve adjustments to past period decisions. Thus, we concluded that there is some face
validity to the identified pricing strategies.
Summary
Both brand-store and brand-level cluster analyses reveal
that retailers adopt a variety of pricing strategies that extend
beyond the conventional Hi-Lo and EDLP strategies. Some
strategies are more frequently adopted at the brand-store
level than at the store level and vice-versa. The distinctive nature of the brand-store pricing strategies underscores the fact
that retailers customize their pricing strategies at the more
fundamental brand-store level, in addition to the store level.
Classification of retailer pricing strategies at the store level
cannot reveal these differences. Thus, our results suggest a
new level of complexity in pricing strategies, thereby extending Lal and Raos (1997) theoretical insight that retailers

221

pricing strategies are based on a basket of brands and categories. They also extend Dhar, Hoch, and Kumars (2001)
finding that retailer pricing is not just storewide, but is
category-specific to show that retailer pricing is brand-store
specific. Surprisingly, when retailer pricing is considered at
the brand-store level, the most prevalent strategy is not Hi-Lo
as is widely believed at the store level. It is one characterized
by average relative brand price, low price variation, medium
deal intensity, and medium deal support.

Discussion of managerial implications

We have extended earlier research on the nature of
retailers pricing strategies in three ways. First, we have empirically identified three new pricing dimensionsrelative
price, deal intensity, and deal supportthat complement
earlier research that has focused on pricing policy or price
variation (Hoch et al., 1994; Shankar & Krishnamurthi,
1996). The four pricing dimensions characterize pricing
strategy for a diverse set of brands, categories, stores and
geographic regions. Second, our results show that retailers pricing practices vary within the same storestores
do not follow uniform pricing practices across brands and
categories. Third, we have empirically described retailers
diverse pricing strategies (combinations of different levels
of dimensions)with value pricing and aggressive pricing
are most prevalentproviding some empirical support for
Levy and Weitzs (1998) observation of pricing diversity.
Pricing strategies are multi-dimensional
Prior research has focused exclusively on a single
dimensionprice variation (i.e., ELDP vs. Hi-Lo pricing)
implying a single pricing continuum. By examining a
broader set of measures, our results show that retailer pricing strategies reflect a richer set of dimensionsincluding
relative price, deal intensity, and deal support. Each of these
dimensions is continuous, and can be combined with any
level of another dimension. Depending on the combination of the levels of these dimensions, retailers can utilize
diverse pricing strategies at the brand-store leveli.e., an
undifferentiated strategy such as Moderately Promotional
pricing, niche strategies such as Exclusive or Aggressive
pricing, or mass-customized strategies such as Hi-Lo and
EDLP pricing.
Underlying pricing dimensions are stable, but pricing
strategies are brand-store specific
Although four pricing dimensions can be used to characterize all retail pricing decisions, retailers do not use
the same pricing strategies for different brands, categories,
stores, and geographic regions. For example, a retailer will
not necessarily offer consistently low prices for all brands
and categories in a given market place. This result implies

222

R.N. Bolton, V. Shankar / Journal of Retailing 79 (2003) 213224

that retailers are using their intimate knowledge of brands

and markets to customize their pricing strategieseither
to stimulate the purchases of promotion merchandise or to
encourage regular price merchandise purchases on the same
shopping trip. Hence, there is an opportunity for manufacturers to develop and exploit information about retailers
pricing strategies across brands and categories to become
a category captain, to support their brands with targeted
marketing efforts, and to build better relationships with
retailers. As a category captain for a retail chain, a manufacturer, with its resources, can help the retailer better plan
its pricing and promotions for all the brands in that category
(based on store-level data) that moves the retailer toward a
desirable pricing strategy for the manufacturers brands. For
example, Johnson (1999) makes some pragmatic recommendations regarding how manufacturers can manage their
brands when retailers begin to move toward a EDLP pricing
strategy. For example, she suggested the manufacturer avoid
price-related promotions and move toward on-pack promotions (e.g., collectibles linked to manufacturer advertising)
that would attract switching consumers.
Retailer pricing strategy is not restricted to EDLP or
Hi-Lo pricing strategies
Prior research and conventional wisdom assume that retailer pricing strategies fall under one of EDLP and Hi-Lo
pricing strategies. This view of retailer pricing strategy is
primarily at the store or chain level and is largely driven by
the store or chain positioning. Chains typically communicate
or signal their pricing policy as one of these two strategies.
For example, Wal-Marts positioning slogan, Low prices,
always, indicates an EDLP strategy. Similarly, Food Lion
and Lucky have also positioned themselves as EDLP chains.
However, the number of chains with EDLP positioning is
small. Most of the grocery retail chains in the United States
are positioned as Hi-Lo pricing chains (Partch, 1992). For
example, 15 out of the 17 chains in our data base are positioned as Hi-Lo chains (as determined by our examination
of the companys annual reports and other publicly available
documents).
Our analysis of pricing decisions in 17 chains, 212 stores,
six categories and five markets reveal some surprising insights about how retailers depart from overall Hi-Lo pricing
and promotion strategies when they customize their decisions for a particular brand and store. First, it shows that at
the brand-store level, retailers practice five types of pricing
strategies, which we label as Exclusive, Moderately Promotional, Hi-Lo, EDLP, and Aggressive pricingnot just two
types of pricing strategies as is widely believed. Second, an
interesting finding is that the most prevalent pricing strategy
is not any strategy close to Hi-Lo pricing strategy as casual
observation of chains and their pricing may suggest. It is
a pricing strategy that is closer to EDLP strategy than any
other strategy. The second most prevalent strategy, Aggressive pricing, is not close to a Hi-Lo pricing strategy either.

These findings point out that although retailers may signal

to consumers a positioning strategy of EDLP or Hi-Lo pricing strategy at the store or chain level, they actually engage
in different pricing strategies at the brand-store level.
This apparent contradiction can be explained by the fact
that EDLP is simpler to communicate internally and easier
to implement. However, Dolan and Simon (1996) observe
that pricing decisionsas opposed to other marketing
decisionsare the key to profitability for most companies, and nowhere is this more evident than in retailing
with its accompanying razor-thin margins. Thus, retailers
must become proactiverather than passive price-takers
customizing price at the brand-store level to local conditions. This argument is particularly compelling for retailers
who encounter dramatic differences in profitability associated with different store-wide pricing policies (Hoch et al.,
1994). Our study extends this argument by showing alternative ways that retailersand their competitorscan (and
do) customize their own pricing and promotion strategies to
different brands and stores. Retailers should closely monitor competitor behaviorat specific stores, for specific
brandsto see what pricing strategy is being adopted for a
particular brand at a specific store. Only then, they can form
reasonable managerial expectations about their competitor
pricing, and develop their own strategies. In particular, the
theoretical literature on price promotions emphasizes that
competition among retailers is a critical determinant of
optimal pricing strategies (cf. Pesendorfer, 2002).
This study can help retailers better understand their current pricing strategies across brands and stores. We believe
that our results generalize to other supermarket brands and
categories because our analyses are based on a census of
brands and stores in a representative cross-section of category roles and markets. Hence, these results provide a
benchmark for assessing an individual stores pricing decisions. For example, retailers can use our taxonomy to
classify their pricing strategy for a particular brand-store
combination (or store), and then compare it with the clusters of retailer pricing strategies described in Tables 5
and 6. This benchmarking procedure allows the retailer to
think about how his/her pricing strategies may differ from
competitors pricing strategies. Retailers have pricing latitude when they differentiate themselves along non-pricing
dimensions (e.g., by coordinating price and promotion, emphasizing different categories, serving different clientele).
Consequently, we observe a diverse set of pricing strategies
that are (apparently) successful in the marketplace.
A retail store can observe its closest competitor stores
pricing and promotion decisions over a period of time and
infer the competitors pricing strategy. For example, in our
data, a store in a non-metro market can observe that its
closest competitor store (midlarge sized store belonging to
a medium sized chain), for a leading brand of bleach, has
relative price of 0.993, price variation of 0.297, deal intensity
of 0.488, and deal support of 0.182. Using Table 5, the store
can infer that this observed pricing corresponds closest to a

R.N. Bolton, V. Shankar / Journal of Retailing 79 (2003) 213224

Hi-Lo strategy for this brand. However, there is an important

caveat to such inferences. The retailer must be sensitive to
the fact that retailer pricing strategies within a given market
are interdependent. In other words, when a retailer observes
its closest competitors pricing and promotion decisions, it
may well be observing some of the competitors reactions
to its own pricing (Coughlan & Mantrala, 1994; Dickson &
Urbany, 1994; Shankar & Bolton, 2004). This paradox raises
interesting questions for future research.
Nevertheless, based on the previous discussion, a retailer
can develop useful benchmarks for its pricing and promotion strategies for each category through the following approach that could be undertaken by the appropriate category
manager.
Observe and record the weekly prices and deal depth of
each brand-size within the category for each competitor,
over a reasonable long time frame (typically a year to
account for seasonality, holidays, and special events).
Compute measures of the underlying pricing dimensions
as specified in Table 2 (brand price can be computed as
an average of price per unit weight across all the brand
sizes. In the absence of market-share data of each brand
at competitor stores, market shares of those brands in own
store can be used as proxy).
Classify the pricing strategy for each brand at each competitor store based on Table 5.
Get a broader picture of the competitor pricing strategy for
that category by comparing the pricing strategies across
brands within the category across all competitor stores.
Choose own pricing strategy for each brand within the
category from the diverse set of possible combinations of
levels of pricing dimensions based on how the strategy will
match up with the closest competitors pricing strategies.
Observe and record changes in competitors pricing
strategies in response to own pricing strategies over the
long-term (at least one quarter).
Respond to competitors reactions for those brands that
may be appropriate in the long-term.

223

motions. This information could also be provided for categories and stores in different geographic regions, as well as
broken down by category types (e.g., destination category)
and store types (e.g., large urban store). Such benchmark
information would make it possible to say which pricing
strategies are most prevalent for which categories (e.g., destination or support categories)and describe exactly how
they are implemented by retailers.
Second, this study describes five retailer pricing
strategiesbut it does not assess their profitability. Subsequent research might explore when and how these strategies
are employedand assess the profitability of implementing
these strategies for different brands, categories and stores.
Third, it would be useful to develop a model of optimal
retailer pricing that extends Achabal et al.s (1990) and
Tellis and Zufrydens (1995) models of optimal depth and
timing of promotions to include regular price decisions.
The model by Shankar and Krishnamurthi (2003) is a step
in this direction.
Fourth, our ability to generalize from our findings is limited by our data sources. Although the categories and markets we studied were reasonably diverse and the retail chains
among the largest in the United States, it would be desirable
if future research could replicate the study using a probability sample of categories and markets. Fifth, in the same way
that Noble and Gruca (1999) broke new ground in industrial
pricing, it would be particularly useful for researchers to investigate new domains, such as the pricing and promotion of
non-grocery retailers, and of services. Sixth, retailer pricing
in response to major retail competitive events may be quite
different from retailer pricing in the stable environments that
we studied. For example, Kroger and Safeway retail chains
have drastically cut prices in response the entry of Neighborhood Markets grocery chain from Wal-Mart (Business
Week, 2002). Studying such pricing decisions will contribute
to a deeper understanding of retailer pricing strategies, and
the dynamic way that they unfold.

Acknowledgements
Limitations and future research
This study has limitations that suggest some interesting
opportunities for future research. First, we believe that an
important practical extension of our work would be the development of deal depth and deal support benchmarks for
manufacturers and retailers.9 These benchmarks might describe the relative incidence of different deal depths and deal
support levels. For example, a retailer might find it useful
to know the percentage of deals that have (a) a deal depth
of 5% or less and (b) are accompanied by features only.
This statistic could be accompanied by summary measures
of the average deal duration and average time between pro9

We are indebted to an anonymous reviewer for this suggestion.

We thank the editors, three anonymous reviewers, and

participants at presentations at Harvard University and INSEAD for useful comments. We also thank the Marketing
Science Institute (MSI) for financial support, and Information Resources, Incorporated and the A.C. Nielsen Company
for contributing the data sets. Our thanks are also due to
Valerie Durrant, Xing Pan, and Ying-Ping Yu for their assistance in assembling the data.

References
Achabal, Dale D., Shelby, McIntyre, & Smith, S. A. (1990, Winter). Maximizing profits from periodic department store promotions. Journal of
Retailing, 66(4), 383407.

224

R.N. Bolton, V. Shankar / Journal of Retailing 79 (2003) 213224

Alba, Joseph W., Broniarczyk, Susan M., Shimp, Terence A., & Urbany,
Joel E. (1994). The influence of prior beliefs, frequency cues, and
magnitude cues on consumers perceptions of comparative price data.
Journal of Consumer Research, 21(2), 219235.
Aldenderfer, M., & Blashfield, Roger. (1984). Cluster analysis. Beverly
Hills, CA: Sage Publications.
Ailawadi, Kusum L., Neslin, Scott A., & Gedenk, Karen. (2001). Pursuing
the value conscious consumer: Store brand versus national brand
promotions. Journal of Marketing, 65(1), 7189.
Basuroy, Suman, Mantrala, Murali, & Walters, Rockney. (2001). The
impact of category management on retail prices and performance:
Theory and evidence. Journal of Marketing, 65(4), 1632.
Bell, David R., & Lattin, James M. (1998). Grocery shopping behavior
and consumer response to retailer price format: Why Large Basket
shoppers prefer EDLP. Marketing Science, 17(1), 6688.
Bell, David R., Ho, Teck-Hua, & Tang, Christopher S. (1998, August).
Determining where to shop: Fixed and variable costs of shopping.
Journal of Marketing Research, 35, 352369.
Blattberg, Robert C., & Neslin, Scott A. (1989). Sales promotion: The
long and short of it. Marketing Letters, 1(2), 8197.
Blattberg, Robert C., & Neslin, Scott A. (1990). Sales promotions:
Concepts, methods, and strategies. Englewood Cliffs, NJ: Prentice
Hall.
Bolton, Ruth. (1989, Winter). Relationship between market characteristics
and promotional price elasticities. Marketing Science, 10(1), 2439.
Business Week. (2002, April 15). Wal-Mart is eating everybodys lunch
(p. 43).
Coughlan, Anne T., & Mantrala, Murali K. (1994). Dynamic competitive
retail pricing with uncertainty and learning. Managerial and Decision
Economics, 15(1), 320.
Day, G. S. (1994, October). The capabilities of market-driven organizations. Journal of Marketing, 58, 3752.
Dhar, Sanjay K., & Hoch, Stephen J. (1997). Why store brand penetration
varies by retailer. Marketing Science, 16(3), 208227.
Dhar, Sanjay K., Hoch, Stephen J., & Kumar, Nanda. (2001). Effective
category management depends on the role of the category. Journal of
Retailing, 77(2), 165184.
Dickson, Peter R., & Sawyer, Alan G. (1990, July). The price knowledge
and search of supermarket shoppers. Journal of Marketing, 54, 42
53.
Dickson, Peter R., & Urbany, Joel. (1994, Spring). Retailers reaction to
a competitors price change. Journal of Retailing, 70, 122.
Dolan, R. J., & Simon, H. (1996). Power pricing: How managing price
transforms the bottom line. New York: The Free Press.
Fader, Peter, & Lodish, Leonard M. (1990, October). A cross-category
analysis of category structure and promotional activity for grocery
products. Journal of Marketing, 54, 5265.
Farris, Paul, & Albion. (1981). The effect of manufacturer advertising on
retail prices (MSI Working Paper).
Hoch, Stephen J., Dreze, Xavier, & Purk, Mary. (1994). The EDLP,
Hi-Low, and margin arithmetic. Journal of Marketing, 58, 1627.
Hoch, Stephen J., Kim, Byung Do, Montgomery, Alan, & Rossi, Peter.
(1995). Determinants of store-level price elasticity. Journal of Marketing Research, 32, 1729.
Hulbert, J. (1981). Descriptive models of marketing decisions. Marketing
decision models (pp. 1953). New York: North Holland.
Inman, J. Jeffrey, & McAlister, Leigh. (1993, Fall). A retailer promotional
policy model considering promotion signal sensitivity. Marketing Science, 12(4), 339356.
Johnson, Maureen. (1999). From understanding consumer behavior to
testing category strategies. Journal of the Marketing Research Society,
41(3), 259288.
Kirande, Kiren W., & Kumar, V. (1995). The effect of brand characteristics
and retail policies on response to retail price promotions: Implications
for retailers. Journal of Retailing, 71(3), 249278.

Kumar, Pankaj, & Divakar, Suresh. (1999). Size does matter: Analyzing brand-size competition using store level scanner data. Journal of
Retailing, 75(1), 5976.
Kumar, V. (1998). The impact of internal and external reference prices
on brand choice: The moderating role of contextual variables. Journal
of Retailing, 74(3), 421426.
Kumar, V., Ghosh, Amit, & Tellis, Gerard J. (1992). A decomposition of
repeat buying. Marketing Letters, 3(4), 407417.
Kumar, V., & Pereira, Arun. (1997). Assessing the competitive impact of
type, timing, frequency and magnitude of retail promotions. Journal
of Business Research, 40(1), 114.
Kumar, V., & Leone, Robert. (1988). Measuring the effect of retail store
promotions on brand and store substitution. Journal of Marketing
Research, 25(2), 178185.
Lal, R., & Narasimhan, . (1995). The inverse relationship between manufacturer and retailer margins: A theory. Marketing Science, 15(2),
132151.
Lal, R., & Villas-Boas, J. Miguel. (1998). Price promotions and trade deals
with multiproduct retailers. Management Science, 44(7), 935949.
Lal, Rajiv, & Rao, Ram C. (1997). Supermarket competition: The case
of everyday low pricing. Marketing Science, 16(1), 6080.
Levy, Michael. (1999, September 20). Revolutionalizing the retail pricing
game. Discount Store News, 38(19), 15.
Levy, Michael, & Barton A. Weitz (1998). Retailing management (3rd
ed.). New York: Irwin/McGraw-Hill.
Moriarty, Mark M. (1985). Retail promotional effects on intra- and interbrand sales performance. Journal of Retailing, 62(3), 2747.
Mulhern, Francis, Williams, Jerome, & Leone, Robert. (1998). Variability
of brand price elasticities across retail stores: Ethnic, income, and
brand determinants. Journal of Retailing, 3, 401426.
Narasimhan, Chakravarthi, Neslin, Scott A., & Sen, Subrata K. (1996).
Promotional elasticities and category characteristics. Journal of Marketing, 60, 1730.
Noble, Peter M., & Gruca, Thomas S. (1999). Industrial pricing: Theory
and managerial practice. Marketing Science, 18(3), 435454.
Partch, Ken. (1992, October), Why the issue of EDLP wont go away.
Supermarket business (pp. 1922).
Pesendorfer, Martin. (2002). Retail sales: A study of pricing behavior in
supermarkets. Journal of Business, 75(2), 3366.
Sethuraman, Raj. (1996). A model of how discounting high-priced brands
affect the sales of low-priced brands. Journal of Marketing Research,
33, 399409.
Shankar, V., & Bolton Ruth, N. (2004). An empirical analysis of determinants of retailer pricing strategy. Marketing Science.
Shankar, Venkatesh, & Krishnamurthi, Lakshman. (1996). Relating price
sensitivity to retailer promotional variables and pricing policy. Journal
of Retailing, 72(3), 249273.
Shankar, Venkatesh, & Krishnamurthi Lakshman. (2003). RETPRICE:
A retailer pricing and promotion decision support model (Working
Paper). University of Maryland, College Park, MD.
Tellis, Gerard J. (1986, October). Beyond the many faces of price: An
integration of pricing strategies. Journal of Marketing, 50, 160246.
Tellis, Gerard J., & Zufryden, Fred. (1995). Tackling the retailer decision
maze: Which brand to discount, how much, when and why? Marketing
Science, 14(3), 271299.
Urbany, Joel E., & Dickson, Peter R. (1990). Prospect theory and pricing
decisions. Journal of Behavioral Economics, 19(1), 6980.
Urbany, Joel E., & Dickson, Peter R. (1991). Competitive price-cutting
momentum and pricing reactions. Marketing Letters, 2(4), 393402.
Walters, Rockney, & Mackenzie, Scott B. (1988). A structural equation
analysis of the impact of price promotions on store performance.
Journal of Marketing Research, 25(1), 5163.
Winer, Russell. (1986). A reference price model of brand choice for
frequently purchased products. Journal of Consumer Research, 13,
250256.