The impact of category management on reatiler pricing

7/29/2019 The impact of category management on reatiler pricing.

1/17

16 / Journal of Marketing, October 2001Journal of Marketing

Vol. 65 (October 2001), 1632

Suman Basuroy, Murali K. Mantrala, & Rockney G. Walters

The Impact of Category Managementon Retailer Prices and Performance:

Theory and EvidenceCategory management (CM) is a recent retail management initiative that aims at improving a retailers overall per-formance in a product category through more coordinated buying, merchandising, and pricing of the brands in thecategory than in the past. Despite tremendous retailer and manufacturer interest in the process of CM and its rapidadoption in the industry, much uncertainty exists about the consequences of CM for channel members. The pres-ent study focuses on how a shift to CM by a retailer affects its equilibrium prices, sales, and profitability in a com-petitive retail setting. On the basis of an analysis of a model of two competing national brand manufacturers thatsupply two competing common retailers, the authors find that one retailers adoption of CM increases its averageunit price of the category and reduces its sales volume and revenues. However, this retailer can still enjoy anincrease in its gross margin profits as competing manufacturers wholesale prices fall in the process. Also, the CMadopters profits are greater than those of a symmetric competing retailer that follows the traditional brand-centeredmanagement of a product category when the interbrand competition is high but interstore competition is low. Apply-ing the intervention analysis methodology, the authors empirically test several of these analytical findings, employ-

ing a unique data set that contains information about a supermarket chains weekly average unit prices and salesof the laundry detergent category before and after this product category was moved to CM by the retailer. Thepropositions that adoption of CM will lead to higher retail prices and lower sales are upheld in this empirical study.The authors discuss the implications of these findings for practitioners and researchers, the limitations of the study,and directions for further research.

Suman Basuroy is Assistant Professor of Marketing, University at Buffalo.

Murali Mantrala is Manager, ZS Associates. Rockney Walters is AssociateProfessor of Marketing, Kelley School of Business, Indiana University.Suman Basuroy thanks the Office of Sponsored Research Programs, Rut-gers University, for partially supporting this study through a research grantin 1998. Murali Mantrala thanks the College of Business Administrationand the Center for Retailing Education and Research at the University ofFlorida for supporting this research through summer research grants in1996 and 1997. Rockney Walters thanks the Kelley School of Business atIndiana University for supporting the study through a summer researchgrant in 1997. All of the authors thank Information Resources Inc. for thedata and Yusuf Khanji for his help with the research.

Afundamental change is taking place in the retail gro-cery and drug store industries as retailers and manu-facturers begin to embrace a process called category

management (CM). Traditionally, retailers assigned buyersto purchase brands of specific manufacturers, instead ofmaking all purchases within a particular product category.

Individual brand-oriented buyers sought to improve theireconomic performance by procuring large quantities ofproduct on deals and then relying on retail pricing, promo-tions, and merchandising activities to deplete brand-levelinventories as quickly as possible. In contrast, CM recog-nizes the interrelatedness of products in the category andfocuses on improving the performance of whole product cat-egories rather than the performance of individual brands.Under CM, traditional brand (vendor)-oriented buyers arereplaced with category managers who are responsible forintegrating procurement, pricing, and merchandising of all

brands in a category and jointly developing and implement-ing category-based plans with manufacturers to enhance theoutcomes of both parties (Pellet 1994; Progressive Grocer1995a, b; Supermarket News 1997).

Retailer interest in CM is high. For example, accordingto one recent industry report, 83% of grocery retailers sur-

veyed view CM as the most important issue facing them(Progressive Grocer 1996), and another study shows CMinitiatives to be the most important reason that retailers areimproving their information technology systems (ChainDrug Review 1997). Despite the interest in CM and its rapidadoption in the industry (ACNielsen 1998), however, muchuncertainty exists about the consequences of CM for retail-ers, manufacturers, and consumers. For example, beyondanecdotal reports, few studies have rigorously investigatedhow a retailer shift from brand-centered management(BCM) to CM affects retail prices or retailer and manufac-turer profits as its proponents maintain (Harris and McPart-land 1993; Category Management Report1995). The objec-

tive of the present research is to investigate the impact of aretailers shift from BCM to CM on retail and wholesaleprices, sales, and profits in a competitive decentralizedchannel setting. Adoption of CM results in many changes inthe retailers operations and management. We restrict ourinquiry to pricing decisions and their outcomes, however,because one of the key benefits of CM is a more profitablepricing structure (ACNielsen 1998; Category ManagementReport 1995; McLaughlin and Hawkes 1994). Examiningpricing under CM is important because changes in theretailers approach to pricing can directly affect manufactur-


2/17

The Impact of Category Management on Retailer Prices and Performance / 17

ers, competing retailers, and consumers, who are believed tobenefit from CM adoption.

For the purpose of this study, CM is defined as a situa-tion in which a category manager jointly sets the prices of allbrands in the category so as to maximize total category prof-its. Traditional BCM of a category is defined as a situationin which each brands retail price is set independently so asto maximize its own profit contribution and the prices ofcompeting brands in the category are taken as given. Thesedefinitions are consistent with the basic notion that CM

involves more coordinated management of brands in a cate-gory, including price setting, than in the past (CategoryManagement Report1995; Food Marketing Institute 1995).We recognize that in practice, some level of coordinatedprice setting takes place under BCM. Clearly, CM calls fora high level of price coordination, which is the phenomenonexamined in the present research. In this study, we first layout the strategic approach that surrounds the CM businessprocess. On the basis of this discussion, we analyze how aretailers shift from BCM to CM alters equilibrium prices,sales, and profits within the context of a model of a two-level (competing national brand manufacturers/competingcommon retailers) channel system. More specifically, we

derive and compare equilibrium retailer prices, sales, andprofits as functions of demand function parameters in a cat-egory composed of two national brands sold by two com-

peting common retailers. Comparisons of prices, sales, andprofits are made under three commonly occurring scenarios:(1) Both retailers practice BCM, (2) one retailer practicesCM and the other employs BCM, and (3) both retailers prac-tice CM. The comparative analyses produce several propo-sitions about how the adoption of CM by one retailer affectsits prices, sales, and profitability compared with a competi-tor that stays with BCM or might also shift to CM. Severalinteresting implications of the retailers adoption of CM forthe manufacturers and consumers in the channel also emerge

from this investigation.We test the implications of the key analytical propositions

using store-level scanner data in the laundry detergent categoryobtained from Information Resources Inc. The database con-tains information on average prices, sales volumes, and rev-enues of brands in this category collected from 21 stores thatare affiliated with a large supermarket chain over a three-yearperiod, January 1993 to December 1995. These 21 stores areall located in one major Midwestern metropolitan market. Thesupermarket chain switched the laundry detergent categoryfrom BCM to CM in February 1994, allowing an examinationof the category before and during CM. Information about theaverage prices and sales for competitors to the chain who did

not switch to CM are also contained in the database, providingresearchers and practitioners with a rare look at CM effects forthe retailer and its suppliers, competitors, and patrons.

The Strategic Framework for CMand Theoretical Analysis

CM Framework

In 1995, the Category Management Subcommittee of theECR Best Practices Operating Committee and the PartneringGroup Inc. published an important study: Category Man-agement Report: Enhancing Consumer Value in the Grocery

Industry. This report is basically the how-to of CM and laysout eight critical steps that are necessary for a proper imple-mentation of CM by a retailer. The basic steps in the CMprocess are outlined in Figure 1. It is important to understandthe strategic structure and process surrounding CM to evalu-ate the outcomes of its implementation effectively.

Step 1: category definition. This is the first step in thecategory planning process. This step determines the prod-ucts that constitute a category, subcategory, and major seg-mentation. The category definition should include all prod-ucts that are either highly substitutable or closely related,subject to operational constraints.

Step 2: category role. This step assigns the category role(purpose) based on a cross-category analysis that considersthe consumer, distributor, supplier, and marketplace. Desig-nating a role also helps the retailer allocate resources amongvarious categories.

Step 3: category assessment. This step involves gather-ing and analyzing historical data and relevant informationand then developing insights for managing the category.

Step 4: category scorecard. In this step, performancemeasures are established to evaluate program execution,

FIGURE 1

The Category Management Process

Category definition

Category role

Category assessment

Category scorecard

Category strategies

Category tactics

Category implementation

Categoryreview


3/17

18 / Journal of Marketing, October 2001

FIGURE 2

The Competitive Retail Structure

including target gross margins, return on inventory goals,service levels, and so forth.

Step 5: category strategies. Typical category strategiesinclude cash generating, excitement creating, profit generat-ing, traffic building, and so forth. For example, a traffic-building strategy is focused on drawing consumer traffic tothe store and into the aisle, and a profit-generating strategyseeks to increase category gross margin percentage andgross profit dollar.

Step 6: category tactics. This step involves the determi-nation of optimal category pricing, promotion, assortment,and shelf management that are necessary to achieve theagreed-on role, scorecard, and strategies. Pricing policiesshould be applied to the current prices to develop pricechanges and set overall price changes for the category. Pro-motional policies should be applied in the development of apromotional plan that includes frequency of promotions andrecommended price points (Category Management Report1995, p. 45).

Step 7: plan implementation. An implementation plangenerally includes what specific tasks are to be done, wheneach task should be completed, and who is to accomplish

each task. The plan should also note the start date of eachtask.

Step 8: category review. This step involves the regularmanagement of the intended results of the overall plan.Reviews should be scheduled at established intervals andlisted in the implementation plan.

An inspection of the CM framework reveals that oncethe category definition (Step 1) and the category role (Step2) are chosen, the bulk of the action lies in determining thecategory strategy (Step 5) and then executing the specificcategory tactics (Step 6). Although different strategies maybe appropriate for different categories, retailers predomi-nantly practice CM to increase profits and sales. As

ACNielsen (1998, p. 5) notes in itsEighth Annual Survey ofTrade Promotion Practices, Retailers practice categorymanagement with several ends in mind, but increasing prof-itability, increasing revenue and optimizing item mix are the most important motivators. For example, 97% of retail-ers surveyed indicated that the top priority for practicingCM is increased profitability. Similarly, the retailer exam-ined in this study employed a variety of strategies, includinga profit-generating strategy, consistent with CM. For model-ing purposes, we assume that the retailer sought to buildprofits by CM.

We use the components of the strategic framework ofCM to develop and analyze a model of a decentralized dis-

tribution channel that consists of two competing retailers,Retailers A and C, each carrying two differentiated nationalbrands that are produced by competing manufacturers M1and M2, as shown in Figure 2.

Theoretical Analysis

Note that most trading areas are more complex than the oneshown in Figure 2 and are composed of multiple retailers(e.g., Winn-Dixie, Kroger, Safeway, Albertsons) that sellmultiple brands in a product category, which are producedby multiple manufacturers. However, replacing the 2 1 2

structure (2 manufacturers, each produces 1 brand, 2 retail-ers) with a more complex structure (e.g., a 2 3 2 struc-ture) would not change the substantive nature of the resultsfrom the present modeling framework. Because the benefitsof greater realism in the form of more manufacturers and

brands are outweighed by the costs of a more analyticallycomplex model that does not alter our predictions about theeffects of retailer adoption of CM, we opted for a simplerstructure. The study focuses on the demand-side implica-tions of a retailers shift from BCM to CM. In keeping withthis thrust, we assume that the manufacturers are symmetricwith respect to their costs of production and that the manu-facturersmarginal costs of production are constant; for easeof exposition, manufacturers marginal costs of productionare set equal to zero. We also assume that the manufacturerssell their brands to the retailers at a constant per-unit charge(i.e., the wholesale price) and the retailers incur no othercosts of acquisition. Last, we assume that each manufacturer

sells its brand at the same wholesale price to each retailer.This is in keeping with legal restrictions against discrimina-tory price discounts by manufacturers that exist in practice(see, e.g., Ingene and Parry 1995; Kotler and Armstrong1996, p. 88).

Our investigation employs the traditional game-theoreticapproach to analyzing problems of channel price coordina-tion and competition (e.g., Choi 1991; Coughlan and Wern-erfelt 1989; Ingene and Parry 1995; Jeuland and Shugan1983; McGuire and Staelin 1983; Raju, Sethuraman, andDhar 1995; Trivedi 1998; Zenor 1994). As do many previ-ous researchers who model decentralized channels in thisstream of literature, we assume, first, that each manufacturer

determines the wholesale prices to maximize its profits.Given these wholesale prices, managers at the two retailersdecide on the retail prices to maximize their respectiveobjective functions. The manufacturers know each retailerspricing decision rule and take these into account when set-ting their wholesale prices. That is, we assume that the inter-action between the firms is such that each manufacturer actsas a Stackelberg leader in setting its wholesale price, and theretailers follow with their retail price decisions. Adopting amanufacturerStackelberg rather than retailerStackelbergperspective is reasonable, because no store brands are


4/17


involved in our analysis. Second, we assume that the retail-ers face symmetric brand-level demand functions that arelinear in the brands prices. More specifically, we denote thequantities of manufacturers brand, k (k = 1, 2), demandedat retailers j (j = A, C) by qjk, respectively. Then we specifythe corresponding demand functions as follows:

where the demand function parameters [0,1) and [0,1), respectively, denote the interbrand cross-price sensi-tivity (degree of product differentiation) and the interstorecross-price sensitivity (degree of store differentiation), andpjk is the price of brand k (k = 1, 2) at retailer j (j = A, C).Thus, the quantities of brand k demanded at retailer/store jare directly affected by the brands own price at store j, thedifference between the two competing brandsprices at storej, and the difference between the brands price at store j andits price at the competing store. A value of close to zeroimplies that the two national brands are highly differenti-ated, whereas 1 implies that the brands are highly sub-stitutable. Similarly, as increases from zero to one, demandfor a brand at one retailer is increasingly influenced by itsprice at the other retailer; that is, store competition for con-sumers is more intense. Last, note that aggregate demand inthe product category at zero prices for the brands is scaledto 1.

Previous models of decentralized channel price compe-tition involving a common retailer (i.e., an independent

retailer carrying brands of different manufacturers) thatfaces a linear demand function structure include those ofChoi (1991, 1996), Zenor (1994), Raju, Sethuraman, andDhar (1995), and Trivedi (1998). Among these authors, onlyChoi (1996) and Trivedi (1998) analyze a duopoly commonretailer model similar to ours. However, the focus of theseauthors is to compare the effects of varying channel powerrelationships on profits and prices, assuming that both retail-ers are natural category managers. In contrast, our focus isthe competitive effects of a retailers move from BCM toCM, and we derive and compare equilibrium prices, sales,and profits for the following three retail competitive scenar-ios: (1) The BCMBCM scenario, in which both retailers

practice BCM; (2) the CMBCM scenario, in which oneretailer (Retailer A) shifts to CM and the other (Retailer C)stays with BCM; and (3) the CMCM scenario, in whichboth retailers adopt CM. None of these scenarios in a two-level duopoly channel structure has been previously ana-lyzed in the literature. Indeed, only Zenor (1994) hasfocused on CM issues. However, Zenor concentrates on thepricing and profit benefits of CM by competing multibrandmanufacturers marketing to a monopolistic retailer. In con-trast, our analysis focuses on the effects of the adoption ofCM by a retailer in a competitive retail setting. We now

( ) [ ( ) ( )],

( ) [ ( ) ( )],

( ) [ ( ) ( )],

11

41

21

41

31

41

1 1 2 1 1 1

2 2 1 2 2 2

1 1 2 1 1 1

2

q p p p p p

q p p p p p

q p p p p p

q

A A A A C A

A A A A C A

C C C C A C

C

= + +

= + +

= + +

=

and

(4)11

41 2 1 2 2 2[ ( ) ( )], + + p p p p pC C C A C

describe the three scenarios that will generate the proposi-tions that we subsequently subject to empirical testing.

Analysis of the BCMBCM scenario. In this scenario,we assume that each retailer has separate managers forprocuring and pricing the two manufacturers brands. Con-sidering the known wholesale prices, wi, i = 1, 2, of each

brand, each of these managers sets price so as to maximizeown-brand profits, taking the price of the other brand at thesame store as well as the two brands prices at the compet-

ing retailer as fixed. Thus, the objective functions of the twobuyers at Retailer A are, respectively, Max(pA1BCM

w1BCM)qA1

BCM w.r.t. pA1BCM and Max(pA2

BCM w2BCM)qA2

BCM w.r.t.

pA2BCM. Similarly, the objective functions of the two buyers at

Retailer C are Max(pC1BCM w1

BCM)qC1BCM w.r.t. pC 1

BCM and

Max(pC2BCM w2

BCM)qC2BCM w.r.t. pC2

BCM. Substituting the cor-

responding demand functions in Equations 14 into theseobjective functions and simultaneously solving the four buy-ers first-order conditions for profit maximization gives us

the (Nash) equilibrium retail prices, pA1BCM, pA2

BCM, pC1BCM, and

pC2BCM as functions of the given wholesale prices, w1

BCM and

w2BCM, and parameters and . Substituting these retail pric-

ing decision rules into the demand equations, Equations14, we obtain the corresponding demands qA1

BCM, qA2BCM,

qC1BCM, and qC2

BCM as functions of the wholesale prices w1BCM

and w2BCM and the demand parameters. Then, considering

the retailers conditional pricing decision rules, Manufac-turer is (i = 1, 2) problem is to determine the wholesale

price, wiBCM, that maximizes its profit. We assume that Man-

ufacturer i does so taking the other manufacturers whole-sale price as fixed; that is, the manufacturers are themselvesengaged in Nash competition. Thus Manufacturer i solves

Max{wiBCM[qA1

BCM(wiBCM, wl

BCM) + qCiBCM (wi

BCM, wlBCM)]}

w.r.t. wiBCM, i, l = 1, 2, and l i . Simultaneously solving the

manufacturersfirst-order conditions gives the Nash equilib-

rium wholesale prices, w1*BCM and w2

*BCM. With these solu-

tions in hand, it is straightforward to derive the expressionsfor the equilibrium retail and wholesale prices, retaildemands, each retailers total category profits, and the man-ufacturers brand profits as functions of the demand para-meters. The analytical results for this scenario, theCMBCM scenario, and the CMCM scenario may beobtained from the authors.

Analysis of the CMBCM scenario. In a fairly commonsituation, one retailer adopts CM and a competitor in thesame trading area remains with BCM (e.g., SupermarketNews 1997b). More specifically, assume that Retailer A

replaces its two separate national brand buyers with one cat-egory manager who jointly sets the two brands prices so asto maximize total category profit, taking into account theannounced wholesale prices and treating Retailer Cs brandprices as fixed. This category managers objective function

is then Max{(pA1CMBCM w1

CMBCM)qA1CMBCM + (pA2

CMBCM

w2CMBCM)qA2

CMBCM} w.r.t. pA1CMBCM and pA2

CMBCM. In contrast,

the objective functions of the two buyers of Retailer C are

the same as in the previous scenario; that is, Max(pC1CMBCM

w1CMBCM)qC1

CMBCM w.r.t. pC1CMBCM and Max(pC2

CMBCM


5/17


FIGURE 3Difference Between Retailer As Equilibrium

Prices in CMBCM and BCMBCM Scenarios as a

Function of Cross-Price Sensitivities

PriceDifference

.2.4

.6.8

.8

.6

.4

.2

.06.04.02

0

w2CMBCM)qC2

CMBCM w.r.t. pC2CMBCM, respectively. Considering

the change in Retailer As objective function and followingthe same game solution approach as already described, theexpressions for the equilibrium retail and wholesale prices,demands, and the retailers and manufacturers profits arederived.

Analysis of the CMCM scenario. In this scenario, weassume that each retailer has a category manager who jointlysets the prices of the brands in the category so as to maxi-

mize total category profit, taking into account theannounced wholesale prices and treating the competingretailers prices as fixed. Retailer As category managers

objective function is then Max{(pA1CMCM w1

CMCM)qA1CMCM +

(pA2CMCM w2

CMCM)qA2CMCM} w.r.t. pA1

CMCM and pA2CMCM, and

Retailer Cs category managers objective function is

Max{(pC1CMCM w1

CMCM)qC1CMCM + (pC2

CMCM w2CMCM)qC2

CMCM}

w.r.t. pC1CMCM and pC2

CMCM. Following the same game solution

approach as already described, the expressions for the equi-librium retail and wholesale prices, demands, and the retail-ers and manufacturers profits are derived.

Comparative AnalysesTo gain insight into the results obtained in the previous sce-narios, we numerically evaluate the behavior of the retailersequilibrium prices, sales, and profits in these scenarios andthe differences between them as the values of the demandfunction parameters (measuring intracategory brand com-petition) and (measuring interstore brand competition) areeach varied over the range [0,1). Next, we summarize ourfindings in the form of propositions with accompanyingexplanations.

Impact of CM on Category Prices

The initial set of issues examined involves the CM retailers

pricing decisions. A comparative analysis between variousscenarios suggested by our model leads us to the followingproposition, which deals with pricing changes within aretailer:

P1: All else being equal, the retail price of competing brands ina product category will increase when a retailer (RetailerA) moves the category from BCM to CM. This is true irre-spective of whether the competing retailer (Retailer C)remains with BCM (i.e., the CMBCM scenario) or shiftsto CM (i.e., CMCM scenario).

Figure 3 displays the computed difference betweenRetailer As equilibrium prices in the CMBCM andBCMBCM scenarios at various values of and in the

specified range. We see that the price difference is positiveeverywhere. This difference in prices increases as the valueof increases and that of decreases. However, theretailers prices in the two scenarios are the same when =0, whatever the value ofis. The rationale for the results isthat in the CM regime, Retailer A engages in coordinated orcooperative pricing of brands in the category to maximizetotal category profits as opposed to the competitive pricingof brands that occurs under BCM. Coordinated pricingresults in higher prices. There is no difference between coor-dinated and competitive pricing outcomes when the brands

are perfectly differentiated, that is, when their demands areindependent of each others prices ( = 0). However, whenthe brands are substitutable ( > 0), coordinated manage-ment has the effect of dampening this natural price compe-tition that exists between the brands, which results in higherprices. This dampening effect becomes more significant asthe brands become more substitutable, leading to larger dif-ferences between coordinated and competitive pricing out-comes as 1. However, the price increase effect ofwithin-store pricing coordination is tempered by the need to

be competitive with the lower prices of the other retailer,which stays with BCM. This competitive effect becomesmore pronounced as shoppers propensity to engage incross-store shopping increases. Thus, as interstore cross-price sensitivity increases, the difference between RetailerAs prices in the two scenarios diminishes. The analysis of aretailer move from BCMBCM to CMCM is analogous.The following proposition relates to pricing changes andcomparisons across retailers:

P2: All else being equal, the increase in the retail price ofbrands in a product category moved to CM by a retailer(Retailer A) will be higher than the increase in retail priceof that category at the competing retailer (Retailer C) that

continues with BCM (i.e., CMBCM scenario).

Figure 4 shows differences between the retailers equi-librium prices within the CMBCM scenario when values of

FIGURE 4Difference Between Retailer As and Retailer Cs

Equilibrium Prices in CMBCM Scenario as a


PriceDifference

.2.4

.6.8

.8

.6

.4

.2

.01.0075

.005.0025


6/17


and are varied. The intuition is similar to that for P1. Theretailer that sets prices so as to maximize total category prof-its will have a higher price level than a symmetric retailerthat sets each competing brands price so as to maximize itsown profit contribution. However, although the basic eco-nomic explanation for P1 and P2 is fairly straightforward, thepractical implications of these results are significant. Specif-ically, a higher average retail price in a category as a resultof CM is hard to reconcile with the efficient consumerresponse (ECR) objective of providing higher consumer

value to attract and keep customers. We return to this issuesubsequently.

Impact of CM on Category Sales Volume andRevenues

Industry CM experts argue that adoption of CM shouldimprove overall category sales for a retailer. For example, aleading trade journal (Progressive Grocer1995, p. S4) statesthat Category managements ultimate objective ought to beto increase total store sales. Similarly, another report (Cat-egory Management Report 1995, p. xvii) maintains thatCategory Management represents a significant and results-proven opportunity to achieve substantial business improve-

ments. However, as stated in the following proposition, the

present analysis suggests a decline in category sales when aretailer adopts CM.

P3: All else beingequal, the total unit sales in a product categorywill decrease whena retailer (RetailerA) moves the categoryfrom BCM to CM, irrespective of whether the competingretailer (Retailer C) remains with BCM (i.e., a CMBCMscenario) or shifts to CM (i.e., a CMCM scenario).

Figure 5 shows the decrease in Retailer As unit sales uponmoving from BCM to CM as values of and are varied.

These outcomes are not surprising given the earlier observa-tion that prices increase under CM. Note that the decline insales is greatest when is high and is low, that is, in theconditions under which Retailer As price increase is great-est. Furthermore, a higher interstore cross-price sensitivitywill temper Retailer As price increase somewhat but notenough to prevent a loss in sales to lower-priced BCM retail-ers such as Retailer C in a CMBCM scenario. Therefore,we give the next proposition, which is illustrated in Figure6. The analysis of a retailer move from BCMBCM toCMCM is analogous.

P4: All else being equal, the total unit sales of a product cate-gory moved to CM by a retailer (Retailer A) will be lowerthan the total unit sales of that category at a symmetriccompeting retailer (Retailer C) that continues with BCM(i.e., a CMBCM scenario).

These analytical results are consistent with the finding of arecent survey conducted in several large U.S. cities that sig-nificant numbers of shoppers have switched away fromstores practicing CM (Cottrell 1995).

Next, turning to sales revenues, the analytical resultslead to the following proposition (see Figure 7):

P5: All else being equal, the sales revenues of the category willdecrease when a retailer (Retailer A) moves the categoryfrom BCM to CM, irrespective of whether the competingretailer (Retailer C) continues with BCM (i.e., a CMBCMscenario) or shifts to CM (i.e., a CMCM scenario).

Revenues decline because the price increase under CM byRetailer A does not compensate for the resulting reductionin consumer demand at this retailer. Next, we turn to cate-gory profits.

FIGURE 5

Difference Between Retailer As EquilibriumCategory Unit Sales in CMBCM and BCMBCM

Scenarios as a Function of Cross-PriceSensitivities

Category Sales

Difference

.2.4

.6.8

.8

.6

.4

.2

0

.01.02.03


Equilibrium Category Unit Sales in CMBCM

Scenario as a Function of Cross-PriceSensitivities

CategoryUnit SalesDifference

.2.4

.6.8

.8

.6

.4

.2

0.01.02

FIGURE 7Difference Between Retailer As Equilibrium

Revenues in CMBCM and BCMBCM Scenariosas a Function of Cross-Price Sensitivities

RevenueDifference

.2.4

.6.8

.8

.6

.4

.2

0

.005

.01


7/17


FIGURE 9

Difference Between Manufacturers EquilibriumWholesale Prices in CMBCM and BCMBCM

Scenarios as a Function of Cross-PriceSensitivities

WholesalePriceDifference

.2.4

.6.8

.8

.6

.4

.2

0.01.02.03.04

ProfitDifference

.2.4

.6.8

.8

.6

.4

.2

.006.004.002

0

FIGURE 10

Difference Between Retailer Cs EquilibriumProfits in CMBCM and BCMBCM Scenarios as a


CM and Category Profits

So far in our analysis, the move to CM appears to offer lit-tle benefit to Retailer A. However, this conclusion changeswhen we examine this retailers category profits (average

gross margin times sales in dollars) after CM adoption.From its inception, CM has been touted as a mechanism forretailers (and manufacturers) to build overall category prof-its (see, e.g., Category Management Report 1995; FoodMarketing Institute 1995). Our findings, summarized in thefollowing proposition, support this contention and are con-sistent with the retailers strategies in the category.

P6: All else being equal, the profits of a product category willincrease when a retailer (Retailer A) moves the categoryfrom BCM to CM, irrespective of whether Retailer Cremains with BCM (i.e., a CMBCM scenario) or shifts toCM (i.e., a CMCM scenario).

Figure 8 displays the positive difference between

Retailer As equilibrium profits in the CMBCM andBCMBCM scenarios at various values of and . The real-ization of higher profits despite a decline in Retailer As unitsales as well as revenues implies a significant increase inthis retailers unit gross margin. A contribution to thisincrease in unit gross margin comes from a decline in thebrands equilibrium wholesale prices. As illustrated in Fig-ure 9, our analytical results indicate that equilibrium whole-sale prices in the CMBCM scenario are lower than the cor-responding levels in the BCMBCM scenario.

The equilibrium wholesale price is lower in theCMBCM scenario because of the reduction in the overalldemand for the manufacturers products caused by RetailerAs price increase. More precisely, in the face of loweredtotal market demand, the competing manufacturers maxi-mize their individual profits at a lower wholesale price.Effectively, therefore, the move to CM helps Retailer A gainprofits at the expense of the suppliers.

The lower wholesale price induced by Retailer As moveto CM also benefits the competing Retailer C that stays withBCM. Figure 10 displays Retailer Cs increase in profitswhen Retailer A moves from BCM to CM. Thus, within thecontext of our model, both retailers gain profits at theexpense of the manufacturers, even though only one retailer

adopts CM. The analysis of a retailer move fromBCMBCM to CMCM is analogous.

Adoption of CM will always increase Retailer As profit,as stated in P6. However, because As action also enhancesRetailer Cs profitability, it would be interesting to comparethe relative profits. An evaluation and comparison of theequilibrium profits of Retailers A and C in the CMBCMscenario (Figure 11) shows that Retailer As equilibriumprofits are greater than those of Retailer C only under certaincircumstances. Retailer As profits are higher if the inter-brand cross-price sensitivity is high (i.e., close to unity)and if the interstore cross-price sensitivity, , is close to zero.Conversely, Retailer Cs equilibrium profits can dominatethose of Retailer A when is large and is small. That is,although Retailer As profits under CM are higher than itsown profits under BCM, relative to Retailer C, Retailer Acan enjoy higher profits if few consumers visit the compet-ing retailer, which has lower prices. But if cross-store shop-ping is significant, Retailer As improved margin under CMdoes not adequately compensate for its loss in demand toRetailer C. This leads us to the following proposition:

FIGURE 8

Difference Between Retailer As EquilibriumProfits in CMBCM and BCMBCM Scenarios as a


Profit

Difference

.2.4

.6.8

.8

.6

.4

.2

.01.0075

.005.0025

0


8/17


ProfitDifference

.2.4

.6.8

.8

.6

.4

.2

.002.001

0.001

FIGURE 11

Difference Between Retailer As and Retailer CsEquilibrium Profits in CMBCM Scenario as a


P7: Retailer As category profits are greater than those ofRetailer C when interbrand cross-price sensitivity, , ishigh and interstore cross-price sensitivity, , is low, and

Retailer As category profits are lower than Retailer Csprofits when is low and is high.

P7 echoes some conclusions of Cottrells (1995) study:It is still not clear that those [retailers] who dont practice[CM] find themselves at any competitive disadvantage. Infact, the study found many [retailers] to be performing bet-ter than those competitors who were practicing it (Progres-sive Grocer1995, p. S8).

Empirical Tests of the Propositions

Data

Empirical tests of the propositions require time series datafrom a trading area where a retailer is known to have movedto CM. More precisely, time series data covering pre-CMand CM regimes for a retailer and its competitors are neededto test the propositions. However, such data on the effects ofCM are difficult to obtain because CM is a fairly recent phe-nomenon, which limits the number of retailers that havedatabases comprehensive enough to capture category per-formance before and during CM. Fortunately, the data setused in this study overcomes this hurdle, providingresearchers and practitioners with a rare opportunity to mea-sure CM effects.

Aggregate store-level weekly scanner data from 21stores of a national supermarket retail chain, hereafter calledRetailer A, that had moved its laundry detergent category toCM were obtained from Information Resources Inc. (IRI).The stores are all located in one defined Midwestern U.S.urban market and together account for 35% share of laundrydetergent sales in this market. The data set also containsaggregated information from the major competitors ofRetailer A, hereafter labeled Retailer C. The data include theweekly average (weighted by stockkeeping unit sales) unitretail prices as well as the weekly unit sales in the laundrydetergent category for Retailers A and C for 156 weeks:

from January 1, 1993, through December 31, 1995. In gen-eral, implementation of CM in a product category by aretailer occurs over a period of time rather than during a par-ticular week. The intervention analysis methodologyemployed here requires a specific date for the switchover toCM, however. The managers of the supermarket chain indi-cated to the authors that CM began in earnest around the57th week of the data. Therefore, for analytical purposes wehave assigned this week as a switchover date, acknowledg-ing fully that CM implementation was spread over a period

of time around that date. None of the competing retailers inRetailer C adopted CM during this three-year period; allstayed with traditional management approaches (BCM)throughout. Thus, the data set covers 56 weeks of the pre-CM regime (i.e., BCM) and 100 weeks of the CM regime forRetailer A and 156 weeks of only BCM regime for RetailerC. Unfortunately, the data set does not include informationon wholesale prices or the retailers margins. Therefore, weare not in a position to rigorously test our analytical propo-sitions pertaining to retailers profits.

Methodology

To test the propositions related to CMs effects on average

prices and sales, a simple t-test might be used to comparethe means of the dependent measures (average prices, sales,revenues) before and after adoption of CM by Retailer A.However, the use of a t-test to determine whether a signifi-cant difference exists between the means of the two regimesis not appropriate when the dependent measures come fromthe same time series. This is because a t-test assumes thatdata for each group are independently generated from sam-ples with normal distributions and constant variances. Thetime series data used in the study does not meet theseassumptions because of the presence of autocorrelation.Therefore, we employ intervention or interrupted timeseries analysis (e.g., Box and Tiao 1975; McDowall et al.

1980) to study the impact of Retailer As adoption of CM onprices, sales, and revenues. Although applications of inter-vention analysis in the marketing literature are few (e.g.,Krishnamurthi, Narayan, and Raj 1986; Mulhern and Leone1990; Wichern and Jones 1977), the technique is widelyused in other social sciences. Figure 12 (see McCain andMcCleary 1979) summarizes the details of the four stagesautoregressive integrated moving-average method(ARIMA) identification, estimation, diagnosis, and inter-vention hypothesis testingof the intervention analysisprocedure.

Empirical Analyses

We now investigate P1 through P6, which are derived fromour analytical model, using the intervention analysisapproach. However, to begin with, we perform a rough testof whether the laundry detergent category at Retailer A ischaracterized by conditions favoring CMa high interbrandcross-price sensitivity at Retailer A and demand that is rela-tively insensitive to other retailers prices. Following Raju,Sethuraman, and Dhar (1995), we use the category own-price sensitivity as a surrogate for national brand cross-pricesensitivity. This approach is reasonable given the one-to-onerelationship between own-price sensitivity and cross-


9/17


FIGURE 12

Intervention Analysis: Modeling Strategy

Diagnosis

An ACF and PACF are computed from theresiduals of the estimated tentative model, andthese statistics are used to decide whether thetentative model is adequate. If there are nospikes at lag 1 and the seasonal lags (if any) ofthe ACF and PACF, and the Box-Ljung statistics

are not significant, the model is adequate.

Intervention Hypothesis Testing

Specify a tentative transfer function model. Estimate the joint transferfunctionARIMA model parameters, , , and . If the size ofthe intervention effect is large, diagnose the adequacy of the jointtransfer functionARIMA model. Interpret the intervention effectparameters.

Estimation

The parameters of the tentativeARIMA model are estimated.

Identification

An ACF and PACF are computed from the timeseries observations. If these indicate that theseries is nonstationary, the series is differenced,and the new ACF and PACFs are computed fromthe differenced series. These statistics are used toselect a tentative ARIMA model for the series.

price sensitivity , namely, = (1 + ), in our specificationof the demand functions (Equations 14). Consistent withthe linear form of the demand functions assumed in the ana-lytical model, we estimate the following equation:

where QtA = Retailer As weekly unit sales of laundry deter-gent, PtA = Retailer As average weekly unit price, = intra-store own-price sensitivity, P tC = Retailer Cs averageweekly unit price, = interstore cross-price sensitivity, =intercept, and eA = the error term. The regression results are

reported in Table 1.The results indicate that the own-price price sensitivity

of the average item in the category, , is negative and sig-nificant for Retailer A. The nonsignificant but positive esti-mate of the interstore cross-price sensitivity () suggeststhat little price-based cross-store shopping takes place forbrands in the category. Overall, the results suggest that thelaundry detergent category possesses the competitive char-acteristics that support Retailer As move to CM for thiscategory.

(5) Q = + P + P + e , t = 1, 2, ..., 56,At

At

Ct

A

Testing Implications of Propositions on CM andRetail Price

A testable implication of P1 is that Retailer As average unitretail price of laundry detergents in the CM regime will behigher than the average unit retail price level in the pre-CMregime. A testable implication of P2 is that the increase inthe average unit retail price of Retailer A as it moves frompre-CM to CM will be higher than the increase, if any, of theaverage unit retail price of Retailer C. Summary descriptivestatistics with respect to the mean of the laundry detergentcategorys weekly average unit price series of Retailers Aand C before and after Retailer As move to CM are shown

in Table 2. It appears that Retailer As mean weekly averageunit price rose by approximately $.33 after the move to CM.Furthermore, Retailer As mean weekly average unit pricewas $.16 lower than that of Retailer C in the pre-CM regimebut higher by approximately $.06 in the CM regime.

The time series plots of Retailer As weekly average unitprices and the difference between Retailer As and RetailerCs weekly average unit price series are displayed in Figures13 and 14, respectively. Figure 13 indicates that there is agradual upward shift in As prices from the pre-CM periodto the CM period, and Figure 14 shows that Retailer Asprices rose more and were higher on average than those ofRetailer C in the CM period.

Intervention analysis tests of P1 and P2. The results ofthe ARIMA model identification and estimation stages withrespect to the time series data shown in Figures 13 and 14are reported in Table 3. Diagnosis checks of the autocorre-lation functions (ACFs) and the partial autocorrelation func-tions (PACFs) of the residuals of each of the estimated mod-els in Table 3 revealed that they were white noise; that is, themodels were found to be adequate.

Figures 13 and 14 show that the series in each caseunderwent a gradual and permanent change after the adop-

TABLE 1Price Sensitivities of Retailer A in the Pre-CM

Period

ParameterVariables Values

Intercept 79,467.16*Intrastore own-price sensitivity, 13,762.10*Interstore cross-price sensitivities, 2711.65 (n.s.)

*Significant at the p< .01 level.Notes: R2 = .310, adjusted R2 = .284, F = 11.93. n.s. = not signifi-

cant.


10/17


TABLE2

Descrip

tive

Statis

ticso

fKey

Variab

les

Pre-C

MR

eg

ime

CM

Reg

ime

Re

levan

t

Stan

dar

d

Stan

dard

Propos

ition

Variab

le

Mean

Dev

iatio

n

Minimum

Max

imum

Mean

Dev

iation

Minimum

Max

imum

P1

Weeklyav

erage

unitprice

ofA

3.7

7

.21

3.2

1

4.1

3

4.1

1

.29

3.4

3

5.0

5

P2

Weeklyav

erage

pricedifference

betweenA

andC

.1

6

.23

.7

4

.54

.06

.26

.7

3

1.1

6

P3

Weeklyav

erage

unitsales

ofA

38,2

57

5198

30,0

07

55,9

17

32,9

90

3975

27,1

24

49,6

77

P4

Weeklyav

erage

marketsha

reofA

.344

.028

.280

.420

.346

.025

.280

.430

P5

Weeklyav

erage

revenues

ofA

143,5

89

16,4

71

115,1

00

206,8

93

135,2

95

15,5

67

106,6

57

195,6

06

P6

Computedweekly

averagepro

fitofA

22,3

13

7331

1731

35,4

38

30,7

16

9586

10,3

68

65,7

28


11/17


FIGURE 13

Retailer As Weekly Average Unit Prices


Weekly Average Unit Prices

tion of CM by Retailer A. Therefore, the following modelsfor the intervention hypothesis testing stage of the analysiswere specified:

Note that in these equations, p + I (elsewhere, simply I)is the intervention component, or what is commonly referredto as the transfer function. Furthermore, all the series thatwe consider in this article have been differenced once tomake them stationary. The maximum likelihood estimates ofthe parameters of these models are reported in Table 3.

(6a) ;

0 < < 1 and and

;

0 < < 1 and

p p I a a

Ifor t

for t

b p p I a a

Ifor t

for t

A A t t t

t

At

C A C t t t

t

t t

t t t

= + +

=

Documents

The impact of category management on reatiler pricing