1-s2.0-S002243591400027X-main

Journal of Retailing 90 (2, 2014) 141153

Determinants of Store BraRaj Sethuraman a,, Katrijn Gi

a Marilyn and iness,

oll B

Abstract

Private lab ecadof store bran ral qfactors influ velocharacteristi bilityof data from inandiscuss the k arch 2014 New

Keywords: St

Store brands (SBs) have been growing in sales across theglobe over the last two or three decades. In the United States,supermarket sales of SBs increased 5.1 percent in 2011, pushingSB dollar share up half a point to 19.5 percent, a record high(Nielsen/PL(NBs) gainunit share percent in tare also gro2007). Howcategories for the Unitin the U.S.Asia. Withwas three tthe SB sharshares also

The autho Correspon

E-mail adkatrijn gielen

1 Tel.: +1 9

try. n as

aggregate market share of SBs.Sethuraman (1992) and Hoch and Banerji (1993) were among

the first to provide a comprehensive empirical analysis of thedeterminants of SB share. Since that time, a large body of empir-

http://dx.doi.o0022-4359/MA 2012). By comparison, sales of national brandsed only 2 percent over the same period in the U.S. SBin 2011 rose to 23.6 percent, compared to about 15he 1980s. SB shares are even higher in Europe, andwing in Asia and Australia (Kumar and Steenkampever, market shares of SBs are not uniform across

or countries. For example, in 2012, SB market shareed Kingdom was twice that for the U.S., and SB share

was more than twice the share for most countries inin the U.S., average SB share for all packaged foodsimes as much as in household goods and five timese in personal care products (Euromonitor 2012). SB

vary by retailer and across geographic regions within

rs thank William Caylor for research assistance.ding author. Tel.: +1 214 768 3403; fax: +1 1214 768 4099.dresses: [email protected] (R. Sethuraman),[email protected] (K. Gielens).19 962 9089.

ical research has emerged addressing two questions: (i) At theindividual or household level, what factors influence SB prone-ness and choice vis--vis NBs? and (ii) At the aggregate marketlevel, what are the determinants of SB share? In this study,we attempt to draw empirical generalizations from this bodyof literature.

In particular, we identify 54 empirical studies that provideinformation on the antecedents of SB proneness, choice, andmarket share. These studies yield several directional empiri-cal generalizations related to whether a particular factor, onaggregate, positively influences SB choice or share, negativelyinfluences it, or does not have a significant influence. We thendelve deeper into the data and the studies and offer additionalinsights into the strength of the relationship, the moderatorsof the effect, and other aspects of the relationship that can begleaned from the meta-analysis data.

The rest of the paper is divided as follows. We first presenta utility framework and identify potential determinants of SBshare which we investigate in our meta-analysis. Next, we

rg/10.1016/j.jretai.2014.04.002 2014 New York University. Published by Elsevier Inc. All rights reserved. Leo Corrigan Professor and Chair of Marketing, Edwin L. Cox School of BusTX 75205, USA

b University of North Carolina at Chapel Hill, Campus Box 3490, McC

els or store brands have witnessed considerable growth in the last few dd vary considerably across categories, markets, and countries. A natu

ence store brand market shares. Drawing on a utility framework, we decs that can influence store brand share. We test the empirical generaliza

54 individual and aggregate market studies. Twenty of the 21 determey findings, their implications, and directions for future empirical rese

York University. Published by Elsevier Inc. All rights reserved.

ore brands; Private labels; Empirical generalization; Meta-analysis; Retailing

Introduction a counquestiond Share

elens b,1 Southern Methodist University, 6212 Bishop Boulevard, Dallas,

uilding, Chapel Hill, NC 27599-3490, USA

es, especially in grocery products. However, market sharesuestion of interest to academics and practitioners is whatp 21 consumer, manufacturer, retailer, and product-market

of the effect of these determinants through a meta-analysists show significant, empirically generalizable effects. We.

This variation in market share raises an important to what factors influence consumer choice and thus

142 R. Sethuraman, K. Gielens / Journal of Retailing 90 (2, 2014) 141153

describe the procedure for compiling and meta-analyzing thedata from published literature. We then present and discuss theresults of our meta-analysis. We conclude by summarizing thekey results and stating their implications, as well as offeringdirections for future research.

Framework

We draw on a consumer utility maximization framework todevelop the potential determinants of SB share. Consumers willbuy SBs if they perceive the SB to be of better value thanNBs. Perceived value arises from non-price utility for the brand(owing to perceived quality and imagery) and (dis)utility forprice. We ithe drivers Fig. 1.2

Drivers of

Price utdriven by promotionssensitivity.are the low[p(NB)-p(SSB price diprice discorelative SBand the higferential, hdisutility fo

The NBof both retwho markeretailers inSB price pshare will price prommanufacturturn, depenSethuramation or crosand retailerin turn, depresulting in

2 The 21 drpotentially incriteria: (i) avwith the utilitaffects price deffect of that demographic variable in theconsumer utilhighlight poteof the determ

stimulated by the number of NBs. Other things equal, more NBsmeans less quantity sold of each brand and hence a stronger pres-sure to redgap, resultialso lead tothat a few Nand thus mprice contrto influencing in lowetop retailerand gives mwho are ve

ize toiffernce

se SBsum

(i) cng te segiven

for chas

theold netathe prdsoutilitmis

an

thery isfor g, and

ceiveility

var

er theng ining ed rarity

purers

erceifor tong oppiuencmers

basnom

altaser-pdentify 21 potential determinants of SB share fromof price utility and nonprice utility, as represented in

price utility

ility in the context of NB-SB competition is directlythe price of SBs relative to NBs, temporary price

offered for both NBs and SBs, and consumer price Generally, NBs are the higher-priced brands and SBser-priced options, so that the NB-SB price differentialB)] is generally positive. Hence, the higher the NB-fferential [p(NB)-p(SB)], the higher the temporary SBunts and the lower NB price discounts, the higher the

value and the greater the likelihood of SB purchaseher the SB share. Moreover, for a given price dif-

igher consumer price sensitivity implies higher pricer NBs, resulting in larger SB share.-SB price differential is determined by the conductailers, who sell NBs and SBs, and manufacturers,t their NBs through the retailers. In particular, ifcrease the NB-SB price differential by increasingromotions (e.g. temporary price discounts), then SBincrease. If retailers or manufacturers increase NBotions, then SB share will decrease. Retailers anders price decisions with respect to NBs and SBs, ind on competitive and other marketplace factors. Raju,n, and Dhar (1995a) show that when price competi-s-price sensitivity among NBs is high, manufacturerss reduce the price of NBs. The decreased NB price,resses the price differential between NBs and SBs,

smaller SB share. Price competition may also be

ivers do not represent an exhaustive list of all variables that canfluence SB share. These variables were selected based on twoailability of sufficient data for meta-analysis and (ii) consistencyy framework. For example, NB-SB price competition potentiallyisutility and increases SB share; however, we could not test the

variable due to lack of data. Ethnicity is a potentially interestingvariable which may influence SB share; but we did not include the

meta-analysis since there was no clear link between ethnicity andity. In the same vein, due to lack of data and/or theory, we do notntial nonlinearities or (reverse) causal paths relating to the effect

inants in Fig. 1 on SB share.

their sown dand heincrea

Coning onshoppito pricFor a abilityBy purstretchhousehthe mohence (Richa

Dising a (Erdemthat ifcategoprices sitivitySBs.

PervariabqualitywhethresultienhancperceivFamilirisk ofconsum

then pto opt

Amthe shin inflConsuto bothfor eco(e.g. Bthe lowuce NB prices. This in turn closes the NB-SB priceng in smaller SB share. Higher NB concentration can

smaller SB share. Higher NB concentration impliesB manufacturers garner a large share of the market,

arket power, resulting in wider distribution and moreol. NB manufacturers may leverage this price controle the NB-SB price differential in their favor, result-r SB share. Retail concentration (total share held bys), on the other hand, works in the opposite direction

arket power to retailers. If there are a few retailersry strong, these retailers can then use the power of

obtain better terms for NBs as well as develop theirentiated SBs (e.g. Marks and Spencer in the U.K.)manipulate the price differential in their favor and

share.er price sensitivity, in turn, is posited to vary depend-onsumer demographics, (ii) perceived risk, and (iii)rip/behavior. Consumer demographics often relatednsitivity are household income and household size.

household size, lower income implies less afford-the higher-priced NBs and greater price sensitivity.ing lower-priced SBs, lower-income households mayir limited budgets. In a similar vein, for a givenincome, the greater the size of the family, the tighterry resources leading to higher price sensitivity andropensity to purchase the lower-priced store brandsn, Jain, and Dick 1996).y of uncertainty reflects both the likelihood of mak-take and the consequences of making a mistaked Keane 1996). Sinha and Batra (1999) propose

perceived risk of purchasing a brand in a given less, consumers are more motivated to find lowerreater monetary savings, exhibit greater price sen-

are thus more likely to purchase the lower-priced

d risk, in turn, is influenced by perceived qualityand familiarity with store brands. Higher perceivediability in brands creates greater uncertainty as to

generally lower-priced store brand is of good quality, greater perceived likelihood of making a mistake,perceived risk (Batra and Sinha 2000). Increases inisk will deter SB purchases and diminish SB share.

with SBs, on the other hand, reduces the perceivedchasing SBs (Richardson, Jain, and Dick 1996). If

become familiar with SBs through trial or inspection,ved risk will be reduced and they will be more likelyhe store brand (Fitzell 1992).the shopping trip characteristics, the average size ofng basket and shopping trip frequency play a roleing consumer price sensitivity and thus SB share.

with high quantity requirements, which are relatedket size and trip frequency, are more likely to shopical alternatives, which results in significant savings

1997). Such consumers are thus more likely to buyriced SBs.

R. Sethuraman, K. Gielens / Journal of Retailing 90 (2, 2014) 141153 143

SB sh are

Price

(dis)Ut ilit y

Nonp rice

Utility

Q

Sens

E

(+)(+) (+)

Fig. 1. Determ ffect has a negative nalysiand SB share d riskpositively rela sitive

Drivers of

Non-priperceived qThe perceisively disce.g. Bettmaand Banerjperceived qhigher SB be driven bbelieved thclose to, ifof comparainformationwilling andrelated cuemore proneby buying

The tradquality thamore likelyand Gedensumers miin the choiimage, devwill positivdeveloped atively imp

Becausesells the brelated to to trust th(Bonfrer a

Mo a w

and adi,

omp

adoling

sucfor NB

(relative to SB)

NB-SB Price

Differential (+)*Price Sensitivity (+)

Shopping Behavior

Basket Size (+)

Trip Frequency (+)

Competition

NB-NB price competition (-)

Numbe r o f NBs (-)

NB concentration (-)

Retail conc entration (+)

Demographics

Inc ome (-)

Hous ehold size (+)

Qua lity vari abilit y (+)

SB Familiarity (-)

NB Pro motions (-)

SB Promotions (+)

(+)

SB choice

SB proneness

Disutility of Uncertainty

Perceived risk ( -)

inants of store brand (SB) share. *(+) implies the determinant has a positive e effect. However, we do not test these antecedent or causal paths in our meta-a. For example, SB familiarity is negatively related to perceived risk, perceiveted to SB share through disutility for NB price. Therefore, SB familiarity is po

non-price utility

ce utility comprised quality utility (utility fromuality) and nonquality utility (utility from imagery).ved quality of SBs relative to NBs has been exten-ussed as the key determinant of SB share (see,n 1974; Erdem, Zhao, and Valenzuela 2004; Hoch

i 1993; Steenkamp and Geyskens 2014). Higher SBuality leads directly to higher utility for the SB andshare. Higher perceived quality of SB in turn mayy the level of consumer education. It is generally

at the objective quality of SBs has improved and is not equal to or greater than, the objective quality

1996).across

tainty (Ailaw

Data c

Wecompiportalsble NBs. This objective quality is often reflected in provided in the packaging. Educated consumers are

able to read these labels or access other product-s. They therefore act as smart consumers and are

to saving money without compromising on qualitythe SBs.itional view holds that SBs are generally of lowern NBs. Quality sensitive consumers are therefore

to prefer the higher-quality NBs (Ailawadi, Neslin,k 2001). Finally, brand image, the impression in con-nds of a brands total personality, plays a major rolece between NBs and SBs (Sethuraman 2003). SBeloped through advertising and consistent quality,ely affect SB share. On the other hand, NB image,

through NB advertising and differentiation, will neg-act SB share.

SBs are closely associated with the store thatrand, (nonprice) SB utility may be intrinsicallystore loyalty. Store-loyal consumers are believedeir preferred store and hence the SBs it offersnd Chintagunta 2004; Richardson, Jain, and Dick

ABI/Informprivate lacal studies empirical sreviewed aappropriate

We tracformance aOf those stuestimate ofone of the yielded 54

We codidentificatient variabldeterminanationalizatimeasures r

influencingance and/oon SB sharfor SB

(relati ve to NB)

SB Quality

Uti lit ySB Imagery

Utility

Perceived

SB Qua lity ( +)

ua lity

iti vity ( -)

SB

Image ( +)

ducation (+) NB advertising (-) Store loyalty (+)

on the immediately preceding factor; () implies the determinants but use them to derive the relationship between the determinant

is negatively related to price sensitivity, and price sensitivity isly related to SB share, as indicated in Table 1.

reover, the ability to buy a trusted, single brandide range of product categories reduces uncer-enhances the utility of the shopping experience

Pauwels, and Steenkamp 2008).

Procedure

ilation

pt procedures commonly used in meta-analysis forstudies from the literature. First, we searched onlineh as Social Science Citation Index, Google Scholar,

, and Lexis/Nexis using key words store brands,

bels, retailer brands and identified all empiri-published in academic journals. We identified othertudies from references in these publications. Well the identified papers and selected those that are

for our meta-analysis.ked all empirical studies that captured SB sales per-s either SB proneness, SB choice, or SB market share.dies, we selected all those that provided an empirical

the effect of at least one determinant variable on anythree measures of SB performance. This procedurestudies, which are listed in Appendix Table A1.ed the following data from each study: (i) studyon (authors, year published, journal, etc.); (ii) depend-es (SB proneness/choice/share); (iii) independentt variables (e.g. income, price differential); (iv) oper-ons of the dependent and independent variables; (v)elated to magnitude and/or direction of the effect of

factors on SB share; (vi) measures related to vari-r statistical significance of the effect of determinante; (vii) other study characteristics (e.g. type of data


actual or survey, number of observations used for estimatingthe effect, type of product, etc.).

Method for empirical generalization

Ideally, we would like to assess the following for each deter-minant: (i) direction of the relationship or effect; (ii) statisticalsignificance of the effect; (iii) magnitude or strength of the effect;(iv) nonlinearity of the effect; (v) other factors that moderateor influence the effect and, where appropriate, (vi) direct andreverse cau

of data doedeterminanonly a smaexplore moanalyze datprocedure:

1. Compilthe pottions) aFor examwas opebetweenbetweenSB sharvolumethese stanalyzinlist of stvariable

2. Collect tion, mreportedple cateeach essistent wLehman

3. Provideacross a

(Sethuraas the nu(minus)cant negthere arnegativedetermi

3 Some indIncome: 50K2550K$, is, group or incotreated as an o

and the E-ratio or normalized E-score (Raw E-score/totalnumber of effect observations).

4. Provide summary statistics of the effect using Rosenthal z-scores (Gielens 2012). The Rosenthal z-statistic (Rosenthal1991) is commonly used to provide a magnitude of thedirectional effect in the form of z-scores. It is obtained bycomputing the one-tailed z-score (based on the expected orhypothesized direction) for each effect, summing them toobtain anumber

h E-re. Weffecight os froeighmpleight imate. In asure

nge acco

dies ingsies (onm

acco

artiacco

the edictoasure

anys anare thidenm th

t wethe actur

al b

andip beellin

t studfrom cted dnged

whets, wesality of the effect. However, the nature and paucitys not permit an assessment of all six aspects for everyt. In particular, each study provides information onll subset of the hypothesized variables, and very fewderating factors or reverse causality. So we meta-a on each determinant separately, using the following

e all studies that estimated the relationship betweenential determinant (and its various operationaliza-nd SB share (and its various operationalizations).

ple, potential determinant NB-SB price differentialrationalized in different studies as percent difference

SB price and average NB price or percent difference SB price and leading NB price. Dependent variablee was operationalized in these studies as SB unit

share, SB dollar share or SB versus NB choice. Alludies were included together as the basis for meta-g the effect of that determinant on SB share. Theudies used in the meta-analysis for each determinant

is given in Appendix Table A2.all observations from these studies that provide direc-agnitude, or significance of the effect. If a study

sign, magnitude, or significance results for multi-gories or multiple markets using multiple models,timate was considered a separate observation, con-

ith the spirit of meta-analysis (Assmus, Farley, andn 1984).

summary statistics of the direction of the effectll observations using Empirical support or E-scoreman 2009). The Empirical support score is measuredmber of observations with significant positive effects

the number of empirical observations with signifi-ative effects.3 A large positive E-score indicates thate many more observations with positive effect than

effect implying a strong positive influence of thenant on SB share. We compute both the raw E-score

ependent variables were measured in discrete categories (e.g.K$, 2550K$, >50K$, etc.). The coefficients corresponding toincome categories were recorded. The effect was deemed positive

comparing it to the base category. For example, if the base cat-$ and the coefficient corresponding to SB choice probability for

say 0.2, then the choice probability is higher for the lower incomeme has a negative effect on SB choice. Each binary comparison isbservation.

higsco

all we

tiona w

exa

we

5. Establme

cha6. To

stufindgoror n

7. To any

8. To of preme

andThimo

9. To fro

Firsables, manuf

Nation

Demtionshcore s

4 Mosvalues) the expethen chareportedfew case net z-score and dividing by the square root of the of observations used in the aggregation.4 A relativelyscore generally corresponds with a relative large z-e compute the unweighted Fisher z-score (treating

t observations as independent and assigning them af 1) and weighted z-score (treating multiple observa-m same data as nonindependent and assigning themt equal to 1/# of nonindependent observations for, if three effect estimates are from the same data, the

for each observation is 1/3).e the average magnitude of the effect, if data are avail-particular, where possible, we compute the elasticity

as the percent change in SB share for 1 percentin the independent determinant variable.unt for nonlinearity in effects, (a) identify those (few)that explicitly test for nonlinearity and report their

and (b) look at results for models with discrete cate-e.g. low/medium/high income) to see if nonlinearityonotonicity can be detected.unt for reverse causality, where appropriate, see ifcle investigating the effect has addressed this aspect.unt for other factors that might moderate the strengthffect, we regress the Fisher z-scores against somers such as type of data (recency, geographic location),

of dependent variable (share or choice or proneness), other suitable predictors for which data are available.lysis was only performed for those drivers for which

an fifteen observations were available.tify interactions, if any, look for pertinent findingse individual studies.

Results

present the results for price, then the consumer vari-product-market characteristics, and the retailer ander promotions as listed in Tables 1 and 2.

rand store brand price differential

and utility theories strongly support a positive rela-tween the NB-SB price differential, deemed the

g proposition of SBs, and SB share. Overall, the

ies provided t-tests, standard errors, or actual significance levels (p-which t-statistics could be computed. From these t-tests, based onirection of the effect, we computed a one-tailed p-value which wasto a one-tailed z-value for aggregation. In a few cases, authors onlyher the effect was significant at, for instance, p < .05 or not. In those

assumed p = .05 (z = 1.645) if significant or z = 0 if nonsignificant.


Table 1Empirical generalization results.

Determinant variable Sign # obsns. (studies) E-score (+, , 0) E-score ratio z-Score (weighted) Elasticity (# obsns.)1. NB-SB price differential + 58 (9) +11 (23,12,23) .19 6.18 (8.5) .10 (6)Consumer demographic2. Household income 40 (16) 12 (7,19,14) .31 19.8 (13.9) .13 (29)3. Household size + 38 (12) +4 (10,6,22) .11 3.1 (2.7) n.a.4. Household education + 16 (9) +9 (10,1,5) .56 19.2 (10.0) n.a.Consumer perceptual5. Price sensitivity + 39 (13) +31 (31,0,8) .80 21.5 (20.5) .45 (15)6. Perceived risk 69 (13) 38 (3,41,25) .55 24.4 (12.1) .002 (19)7. Perceived quality variability 18 (7) 9 (0,9,9) .50 9 (7.4) .18 (14)8. SB familiarity + 37 (8) +29 (30,1,6) .78 33.0 (19.3) n.a.9. SB quality + 48 (11) +33 (33,0,15) .69 35.2 (12.6) .09 (5)10. Quality sensitivity 4 (3) 4 (0,4,0) 1 6.5 (5.5) n.a.11. SB image + 14 (7) +11 (12,1,3) .79 9.7 (9.6) n.a.Consumer shopping behavior12. Basket expenditure + 15 (2) 6 (0,6,9) .40 4.8 (.67)a n.a.13. Trip frequency + 21 (9) +14 (14,0,7) .67 10.6 (6.9) n.a.14. Store loyalty + 6 (5) +4 (5,1,0) .67 8.0 (8.2) n.a.Product-market characteristics15. Number NBs 45 (6) 39 (0,39,6) .87 10.8 (5.2) .22 (12)16. Category price elasticity 6 (3) 6 (0,6,0) 1 5.24 (4.0) .05 (5)17a. NB concn. share top NBs 5 (2) 4 (0.4.1) .8 5.06 (4.6) .34 (3)17b. NB concn. var. NB shares 34 (1) +16 (21,5,8) .47 2.81 (2.8) n.a.18. Retail concentration + 40 (4) +13 (14,1,25) .35 10.5 (8.0) .37Retailer and manufacturer promotions19. Retail promotion SB + 67 (11) +28 (30,2,35) .42 15.2 (7.4) .10 (4)20. Retail promotion NB 47 (7) 19 (2,21,24) .40 23.9 (29.1) .31 (4)21. Manufacturer advertising 46 (6) 17 (2,19,25) .37 8.67 (6.1) .17 (9)

a All Fisher z-scores are significant at p < .01, except this score (.67). n.a. = not able to compute elasticities due to lack of primary data.

Table 2Moderating effects.

Determinantvariable

Main effect Data: USA (vs.non US)

Data: recent 5years (vs. nonrecent)

Model: priceincluded (vs.excluded)

DV: share (vs.other)

DV: choice(vs. other)

1. NB-SB price differential + More positive n.s. n.a. Less positive n.a.2. Household income n.s. More negative n.s. n.s. n.s.3. Household size + n.s. n.s. More positive More positive n.s.4. Household education + Less positive More positive n.s. n.s. More positive5. Price sensitivity + n.s. n.s. Less positive n.a. n.s.6. Perceived risk n.s. Less negative More negative n.s. More negative7. Perceived quality var. More negative Less negative n.s. n.s. n.s.8. SB familiarity + More positive n.s. n.s. n.a. n.s.9. SB quality + n.s. n.s. More positive n.s. More positive10. Quality sensitivity n.a. n.a. n.a. n.a. n.a.11. SB image + More positive n.s. n.s. n.a. n.a.12. Basket expenditure + n.a. n.a. n.a. n.a. n.a.13. Trip frequency + Less positive Less positive n.s. n.s. More positive14. Store loyalty + n.a. n.a. n.a. n.a. n.a.15. Number NBs n.a. n.a. n.s. n.a. n.a.16. Category price elasticity n.a. n.a. n.a. n.a. n.a.17a. NB concn share top NBs n.a. n.a. n.a. n.a. n.a.17b. NB concn var. NB shares n.a. n.a. n.a. n.a. n.a.18. Retail concentration + Less positive More positive n.a. n.a. n.a.19. Retail promotion SB + Less positive Less positive More positive Less positive n.a.20. Retail promotion NB n.s. n.s. More negative n.a. n.a.21. Manufacturer advertising n.s. n.s. More negative n.a. n.a.Note: n.s. = non significant; n.a. = data not available.


relationship is positive with an elasticity of .10 (a 1 percentincrease inshare by .1

Interestiin our metvations, thliterature anumber of versus cros

category stquality as cdifferentialbrands; (v)(vii) size o

With resDhar (1995ential betwa low pricelibrium, reThis suggecross-categand/or use Even in caferential, rebe weak anTo obtain quality diff(measured quality wacontrol forsion resultsquality difon SB shaies incorpeffect.

One stuthe relation(leading) Nthe five caby empiricto target thSayman, Hshare is moNB than th

In a moential effecindicating price fight(2014) findSB share, bbecause coSome reseamay work 35 percentprice differresearchersdifferential

in determining SB share, and that retailers may be under- their SBs at the expense of profits (Hoch and Lodish

rap

there boue the an

E-rn-lincati

ands do , andd to nd hi

ers

a n

1965ce, ocatio

(all eta-a

ers

ationakinrodunt onend

sol insuain pitio

ect omed e-drieasinh ress aree op

e theata

famAn aeal-ikelyse S

effenifi, when ct thasma

ived percent price differential increases percent market percent).ngly, this determinant has one of the lowest E-ratiosa-analysis (.19). In fact, in a majority of the obser-e effect is non-significant or even negative. Pastnd our moderator analysis (Table 2) suggest that afactors moderate the effect including (i) longitudinals-sectional studies; (ii) within-category versus cross-udies, (iii) accounting for endogeneity/inclusion ofovariate in the estimation of price effect; (iv) price

with leading national brand versus other national country of study; (vi) share versus choice study; andf price differential.pect to the first three factors, Raju, Sethuraman, anda,b) show analytically that when the quality differ-een NB and SB is low in a category, retailers can set

differential and still get high market share in equi-sulting in a negative relationship across categories.sts that researchers estimating price effects usingory analysis must endogenize the price differentialgood measures of quality differential as covariates.refully designed models incorporating quality dif-searchers have found the price differential effect tod category-dependent (e.g. Dhar and Hoch 1997).

an initial assessment of the confounding effect oferential, we regressed the price differential effectsby z-score) on a dummy variable capturing whethers included in the regression in the original study to

the potential confounding effect. While our regres- do not provide statistical support for (inclusion of)

ferential moderating the effect of price differentialre, further research is required as only two stud-orated quality when estimating price differential

dy (Wang, Kalwani, and Tolga Akcura 2007) findsship between the price differential between SB andB to be strongly positively related to SB share in all

tegories it analyzed. Theoretical arguments, backedal evidence, suggest that stores position their brandse leading NB (Scott-Morton and Zettelmeyer 2004;och, and Raju 2002). Therefore, it is likely that SBre sensitive to its price differential with the leadinge average of all the NBs.derator analysis (Table 2), we find the price differ-t is more pronounced in the U.S. than in Europe,that in the U.S., SBs are still perceived more asers. In a global context, Steenkamp and Geyskens

that an increase in the price differential increasesut this effect is smaller in richer countries, perhapsnsumers in richer countries are less price sensitive.rchers have observed that the price differential effectonly in a specified range between 5 percent and

(Sethuraman 1992). While this nonlinearity in theential effect could not be tested due to lack of data,

using price experiments have suggested that prices may be less important than quality differentials

pricing1998).

Demog

AnoSBs arIt is trnegativbut theand no

Edu(1992)sumer

to NBsior lealow- aconsum

testingBoyd evidenof eduimagethis mconsum

implicand mcery pdiscoucould srelyinghealthin cert

Addthe effperforincomis incr

Witvationhave thmay bin the dable asforth. more dmore lpurcha

Theand sigstudiesand whsuggesbe the a percehics

popular myth debunked in our meta-analysis is thatught primarily by low-income and large households.at the overall main effect of income on SB share isd the effect of household size is positive (Table 1),

atios are low, suggesting the presence of moderatorsear effects.

on can lead to non-linear income effects. Fitzell others opine that the low-income, less educated con-not recognize the quality of SBs, attribute imagery

therefore buy less SBs. This perception and behav-a nonlinear income effect with SB share lower forgh-income consumers and higher for middle-income. While we did not have adequate observations foronlinear income effect and past studies (Frank and; Dick, Jain, and Richardson 1995) provide mixedur result lends credibility to the notion that lackn, lack of SB familiarity, and lack of positive SB

of which have a strong positive effect on SB share innalysis) may be the reasons for why lower-income

are less prone to purchasing SBs. This finding hass for retailers and public policy makers in educating

g low-income consumers familiar with SBs in gro-cts. For example, WalMart and others could give a

SBs to food stamp purchasers, or the government educational materials and SB coupons to householdsely on Social Security payments. A case in point isrance companies only accepting payment for genericsrescription drugs.nally, our moderator analysis (Table 2) revealed thatf income becomes more negative if the study wasfrom data in the last five years, suggesting that

ven price sensitivity with respect to purchase of SBsg over time.pect to household size, nearly 60 percent of the obser-

nonsignificant, while 15 percent of the observationsposite sign. One reason for the lack of a strong effect

lack of statistical power of the studies due to noiseor poor operationalization of the household size vari-ilies with children, families with no children, and solternate explanation is that large households may beprone instead of SB-prone, and such consumers are

to purchase NBs when they are promoted than toBs.ct of education on SB choice or share is positive

cant. The effect is significantly lower in U.S.-basedile it is significantly higher in more recent studieshoice is used as a dependent variable. These findingst, in the U.S., store brands are still not considered tort alternative to NBs, thereby potentially indicating

quality issue.


Consumer perceptions

Interestiratios (overeven whenmore SBs ition is lowenot risky. Tcery producon how retThe resultsprice markenhance SBof these tac

A compof the percetable belowthat qualityprice differ

Effect Price differenPerceived qua

Althouga robust ina consistenbined withimage, thisseek value1996). Thacompromis

Our motual variablthe strong across a w

ditions. Sesignificantlthat SBs mThird, the emore negaties. As witperceived Sthan in Eurmore reliabis significathat SBs mdo in otherthe effect spositive whis a covariato include it is to incluential effecon SB sharreinforcingthe U.S.

Shopping trip variables

pping basket expenditure is the only variable among the ana

ighet va

ity -pr

al-pay strastare. ant lplicaquirng e

is shoicover

re isre hls, a

betwretai

notstoreons

nearrem

ct-ma

pet-NBre, n n

andactur, SB

proliic t

(Sctitiopsi) ler to

thanLal

relaas un

ativeer Sens (

is doseduresosedngly, all consumer perceptual variables have high E- .5), confirming the old adage perception is reality,

it comes to SB purchase. Consumers will purchasef they perceive that the quality is higher, quality varia-r, have a positive image of SBs, and think that SBs arehese findings indicate that, especially in mature gro-ts, the future of SBs and NBs will continue to depend

ailers and NB executives manage these perceptions. underscore the need for retailers to implement non-eting tactics (such as advertising and sampling) to

perceptions and for researchers to study the impacttics on SB share.arison of absolute empirical generalization measuresived quality effect and price differential effect in the

validates the assertion of Hoch and Banerji (1993) is a more important determinant of SB share thanential.

E raw E ratio ztial 11 .19 6.18lity 33 .69 35.2

h the price differential between NBs and SBs is notfluencer of SB share, consumer price sensitivity ist major determinant of household purchases. Com-

the strong results found for SB quality, risk, and meta-analysis validates the notion that consumers

when purchasing SBs (Richardson, Jain, and Dickt is, consumers look for good (low) prices withouting quality.derator analysis with respect to consumer percep-es reveals several additional insights (Table 2). First,positive effect of price sensitivity appears to prevailide range of study characteristics and market con-cond, the effect of perceived risk on SB share isy less negative in more recent studies, suggestingay be increasingly perceived as less risky options.ffect of quality variation on SB share is significantlyive in the U.S. and less negative in more recent stud-h perceived risk, this finding seems to suggest thatB quality variation is more of an issue in the U.S.ope but, overall, SBs are perceived to be becomingle over time. Fourth, the effect size of SB familiarityntly more positive in U.S.-based studies, suggestingay not have the same (brand) appeal there as they

parts of the world, most notably in Europe. Fifth,ize of SB quality on SB share is significantly moreen choice is used as a dependent variable and pricete. The latter effect points out that it is as important

price as a covariate when studying quality effects asde quality as a covariate when studying price differ-ts (discussed earlier). Finally, the effect of SB imagee is significantly higher in U.S.-based studies, again

the notion that SBs still hold an inferior position in

Sho21 wewith h is nocausaling lowNB delays mIn conSB shimporttant immay reshoppiquencywhen c

An SB shaliteratuPauwetaneityDutch loyaltyences

these ccurvilimore p

Produ

Comand NBthermobetwee(.63)manufThat isbrand econom

growthcompeand Pea retaiprofitsshare (

Theis not cumulin lowGeyskshares(as oppproced(as opplyzed for which the expectation that householdsr shopping expenditure are more likely to buy SBslidated. This occurrence may be because of reverse

that households may lower their expenditure by buy-iced SBs, other things being equal or because ofroneness households with large expenditure out-eek and purchase more NBs when they are on deal., shopping trip frequency is an important driver ofThe result that purchase frequency may be a moreever than purchase size to build SB volume has impor-tions for retailers, as stimulating shopping frequency

e different approaches and strategies than stimulatingxpenditure. However, the effect of shopping fre-ignificantly lower in more recent studies, but highere was used as dependent variable.all positive relationship between store loyalty and

supported by the empirical evidence (Table 1). Pastas attempted to refine this overall finding. Ailawadi,nd Steenkamp (2008) explicitly account for simul-een store loyalty and SB share. They find that for two

l chains (a premium and a value-oriented chain), store only impacts SB share, but that SB share also influ-

loyalty at the same time. The relationship betweentructs is further complicated as it was found to be

by Ailawadi, Pauwels, and Steenkamp (2008) for theium chain, but not for a more value-oriented chain.

rket characteristics

ition from NBs, both in the form of number of NBs price competition, negatively affects SB share. Fur-

Hoch and Banerji (1993) find that the correlationumber of manufacturers and SB share is negative

stronger than the correlation between number ofers and share of top two NBs (.21 and .03).

shares are disproportionately decreased if there isferation. These findings give strong credibility to theheory that item proliferation can limit private labelhmalensee 1978). The negative effect of NB-NB pricen on SB share implies that when two NBs (e.g. Cokeare competing intensely on price, it may be better for

leverage the NB-NB competition to obtain higher to engage in NB-SB competition to promote SB

1990; Kumar and Steenkamp 2007).tionship between NB concentration and SB shareequivocal. Higher NB concentration in the form of

brand shares of the top three or four NBs resultsB share. However, according to Steenkamp and2014), the negative effect of NB concentration on SBampened when the markets (countries) are efficient

to those with poor infrastructure and administrative), but enhanced for countries that are secular-rational

to traditional-religious).


Dhar and Hoch (1997) view NB concentration in terms of thevariance in NB shares and offer a different perspective. Accord-ing to theswith one orand others shares willSB againsta decent mmarket, eacfor SBs to keting anddampens thfind a stronbased on an

It appeasword for tthe one hanendow NBterms withpromote ththe other hthen it maythese large

Unlike irelationshipboth share SB share hphenomenowhere retaiSB share wlow E-ratioother factoshows thatretail concemay partlythe U.S. Itis higher thboth in deatheir own sand Geyskis moderatethat the postronger inefficiency a

Retailer an

We comcoupon, distudies repincluded mvery few ob

Notablepromotion The occurrmeasures u

aggregation

instruments. Nevertheless, given that the asymmetric price tiereffect theory (Blattberg and Wisniewski 1989) has been inter-

to ise itpromtitivviou

effecal ge m

SB segore eff

the U) is us incts thtplachoulof SBpredvertiip is servats int

ma

adv the

are ire in

he reverti

labeut is

earcce h

gateew ucal g

sharmetroten

cal vtify

shasionfollo

e Neralverae researchers, if there are ten brands on the market, two brands being big players with large market sharessmall players with low shares, then variance of NB

be higher. In this market, the retailer can position its the big players, ignore the minor brands, and still getarket share. Conversely, if there are ten brands on theh with about 10 percent market share, it is difficult

focus on one set of brands, leading to diffused mar- smaller shares. As a result, variance in NB sharese number of brands effect. Dhar and Hoch (1997)g positive effect of NB share variance on SB sharealysis of 34 categories.rs that NB concentration may be a double-edgedhe NB manufacturers trying to reduce SB share. Ond, high shares concentrated among a few NBs may

manufacturers with the ability to negotiate better retailers, forcing retailers to provide shelf space ande dominant brands, resulting in lower SB share. Onand, if there are one or two NBs that are dominant,

be easier for the retailer to position the SBs against NBs and gain substantial market share.n the case of NB concentration, the expected positive

of retail concentration with SB share is validated forof top retail firms and variance in retail market shares.as often been described as a purely supply-drivenn, meaning that SB share will be high in markets

lers are dominant. As soon as retailer power develops,ill increase. However, this driver has a relatively

(.35), demonstrating that the effect may depend onrs (moderators). Our moderator analysis (Table 2)

the overall significant positive relationship betweenntration and SB share is less positive in the U.S. This

explain why SB shares are higher in Europe than in is well known that retail concentration in European in the U.S. This gives the big retailers leverageling with the NB manufacturers and in developingtrong SB program. In a global context, Steenkamp

ens (2014) find that the effect of retail concentrationd by country characteristics. In particular, they findsitive effect of retail concentration on SB share is

countries characterized by a low degree of marketnd secular-rational culture.

d manufacturer marketing mix

bined all types of retail promotions price discount,splay, feature in our analysis because (i) manyorted just one measure of promotion intensity thatore than one type of promotion and (ii) there wereservations to analyze by individual promotion type.

among the variables with modest E-ratios are thevariables. All of them have E-ratios around .40.ence of nonsignificant results may be attributed tosed in the estimation (percent sold on deal) and

of data across categories, retailers and promotional

pretedincrearetail compe

Premotionempirionly thaffect the catthat thtive inchoiceprice isuggesmarkeelers seffect

As NB adtionshthe obinsighthat NBis, theytion. InSB shSB shaafter tNB adprivateU.S., b

ResformaninvestiWe drempirion SBferent to 21 pempirito quanand SB(regresare as

1. Thov

semply that promotions do not generally help a SBs sales, it is useful to note that, in aggregate, SBotions work from a share standpoint, but so do

e NB retail promotions.s research and our moderator analysis qualify the pro-cts. With respect to NB retail promotions, a recenteneralization study by Gielens (2012) shows that

arket leaders price promotions appear to negativelyhare. Promotion efforts by the second and third NB iny hardly affect SB share. Our results (Table 2) showect of SB retail promotion on SB share is less posi-

.S., in recent times, and when share (as opposed tosed as the dependent variable, but more positive whenluded as a covariate. The presence of moderatorsat SB retail promotion can vary significantly acrosse and method variables, and that marketers and mod-d be cognizant of such differences when assessing the

retail promotions.icted, there is a strong significant negative effect ofsing on SB share. However, while the overall rela-negative, the effect is nonsignificant in a majority oftions (25 out of 46). Individual studies offer furthero the above general finding. Lamey et al. (2012) findnufacturers are pro-cyclical in their advertising. Thatertise more during expansion and less during contrac-

contraction periods, when NBs decrease advertising,ncreases in the short run. Some of this temporarycrease may lead to permanent increases in SB sharecession. Steenkamp and Geyskens (2014) find thatsing is an effective marketing weapon for limitingl shares mainly in large (populous) countries like the

not effective in smaller countries.

Conclusion

h in understanding the determinants of SB sales per-as come a long way in terms of types of variables

d, data used, and sophistication of empirical models.pon this rich and varied body of work to test theeneralizability of the effect of several determinantse. In total, we meta-analyzed 54 studies relating dif-ics of SB proneness, brand choice, and market sharetial drivers. We employed three different metrics alidation (E) score, Fisher z-score, and elasticity

the overall relationship between the potential driversre (Table 1). We also performed limited moderator) analysis with available data (Table 2). Key resultsws.

Result highlights

B-SB price differential has a positive (but not strong)l effect on SB share and the effect is moderated byl factors. In particular, the effect is stronger (more


positive) when the focal NB is the leading brand and in theU.S. than in Europe.

2. The cochasedColleclack ofSB imshare) are lescationseducatSBs.

3. The restrongU.S. msmart

4. All coempiriSBs if tion is think Scially iwill comanagof SB

5. Perceivthe NBtion mvalue on qua

6. Shoppshoppiing SBfreque

7. The stgives cation cproduc

8. The ston SBprice cengagiprofits

9. NB cowhen mship wThe finNBs enand ga

10. Retail may paconcen

11. Overalinfluenare mo

but notlikely

NBs, and SB promotions are less effective in the U.S. thanin Europe, perhaps because of perceived quality.

n thrs, wrnt ah. S

rese

re stawad

ter fitabpiricnageaininre ths woct th

el.lude

Whice-rludedableilablucestion.orpos momen

s?). usa

er. Dy in

ma

goryludear anoss 3iatiotrati

the charsus n

ludearchs (e

at typies (eefendy Sly poe-qu

advenventional belief that SBs are predominantly pur- by low-income, large households is not validated.tively, our results lend credibility to the notion that

education, lack of SB familiarity, and lack of positiveage (all of which have a strong positive effect on SBmay be the reasons for why lower-income consumerss prone to purchasing SBs. This finding has impli-

for retailers and public policy makers interested ining and making low-income consumers familiar with

lationship between education and SB share is fairly and positive. It is more positive in Europe than in thearket, suggesting that SBs may not be considered a alternative in U.S., as much as it is in Europe.nsumer perceptual variables we analyzed have highcal generalization scores. Consumers will purchasethey perceive the quality is higher and quality varia-lower, if they have a positive image of SBs, and if theyBs are not risky. These findings indicate that, espe-n mature grocery products, the future of SBs and NBsntinue to depend on how retailers and NB marketerse perceptions, potentially signaling the growing roleadvertising in coming years.ed SB quality has a stronger effect on SB share than-SB price differential across all empirical generaliza-easures, confirming the notion that consumers seekwhen purchasing SBs without compromising muchlity.ing expenditure does not affect SB choice, butng trip frequency does. Retailers interested in boost-

share may want to focus on stimulating shoppingncy.rong negative effect of number of NBs on SB shareredibility to the economic theory that item prolifer-an limit SB growth and highlights the importance oft assortment for retailers.rong negative effect of NB-NB price competition

share implies that retailers should leverage NB-NBompetition for profits in some categories rather thanng in NB-SB competition to promote SB share or.ncentration has a negative relationship with SB share

easured as share of top NBs, but a positive relation-hen measured as variance in share across all NBs.ding suggests that high market share by one or twoable retailers to position their SBs against these NBs

in share.concentration is positively related to SB share, whichrtly explain higher SB share in Europe, where retailtration is higher than in the U.S.l, NB and SB promotions and NB advertising doce SB shares, but the empirical generalization scoresderate, suggesting that they may work in some cases

others. In particular, promotions by leading NBs areto be more effective in curtailing SB share than other

Eve50 yeabe learesearc

Future

1. MoAilandproem

ma

obtshalineaffelev

2. Incies.choincreliava

redma

incSBmo

tionNBothplaandcate

3. IncDhacr

var

cen

howuctver

4. Increse

tionwhgoror d

5. StuingpricSBough, empirical research on SBs has spanned nearlye have only scratched the surface of what needs to

bout SBs. There are numerous avenues for furtherome important areas are highlighted below.

arch suggestions

udies needed on prot. Only a few studies such asi and Harlam (2004), Pauwels and Srinivasan (2004),Braak, Dekimpe, and Geyskens (2013) address SBility by studying SB gross margins. This lack ofal work does not reflect a lack of academic interest orrial importance, but rather highlights the difficulty ing good profit data, as most retailers are reluctant tois type of information. Still, more research along theseuld allow for deeper insights into how SB strategiese retailers bottom line at both the SB and category

more consumer variables in Choice and Share stud-ereas demographic information is often ignored inelated studies, attitudinal metrics are almost never

in share-based studies. Obviously, the lack of consumer information merged with more readilye scanner panel data or store-based information

the ease of incorporating consumer-related infor- As a matter of fact, few behavioral metrics arerated in the extant literature (e.g. usage situation arere likely to be purchased for specific consumptionts, whereas NBs are more preferred in other situa-Also, little is known about how consumers SB andge may complement rather than substitute for eacheeper insights in the role that both SBs and NBs

consumers consumption patterns may help retailersnufacturers co-promote SBs and NBs and grow the.

more product characteristics in empirical studies.d Hochs (1997) comprehensive empirical study

4 product categories found significant cross-categoryn in the effect of determinants such as price and con-on on SB share. However, few studies have estimated

effects of SB drivers systematically vary with prod-acteristics such as functional versus hedonic, durableondurable.

more promotion variables. There is a dearth of dealing with understanding what type of promo-.g. discounts, coupons, sampling) are effective fore of SBs (regular or premium), in what type of cate-.g. functional or hedonic) and when (e.g. proactivelysively).B advertising and innovation. As retailers increas-sition their SBs as brands in their own right in allality tiers (Geyskens, Katrijn, and Gijsbrechts 2010),rtising and product innovation efforts will become


more important and salient. Little is known about how pre-mium SBs should be priced and positioned.

6. Where appropriate, account for nonlinearity, endogeneity,and reverse causality in estimation models. Even in thevariables we meta-analyzed, there are potential issues ofnonlinearities (e.g. income effect), endogeneity (e.g. pricedifferential), and reverse causality (e.g. store loyalty). Unfor-tunately, we could not study these effects due to paucity ofdata. If theory or prior expectations lead the researcher tobelieve that one or more of these effects may exist, it would beuseful to study these effects, even though the estimation pro-cedure may become more complex (e.g. use of instrumentedvariables or simultaneous equation systems).

7. Study variation of effect over time. At present, few insights areavailable on how consumer effects vary over time. As SBsmature in more and more countries, the impact of several

consumer characteristics may diminish (e.g. risk aversion),while the impact of others may increase (e.g. innovative-ness). If retailers and manufacturers want to successfullytarget consumers, they need insights that are updated overtime.

8. Incorporate role of the retailer. Finally, the role of the retailerhas been vastly unexplored. Although several characteristicscapturing the retail structure at a more aggregate level havebeen considered, little is known about what makes an indi-vidual retailer more or less successful than its rivals. Just likeresults have been replicated across categories and countries,future research should aim for replications across retailers aswell as identify characteristics that make some retailers moresuccessful at gaining SB shares than others.

Appendix.

Table A1Meta-analysis studies.

# Study reference

S1 Ailawadi, K.L., S. A. Neslin & K. Gedenk (2001), Pursuing the value-conscious consumer: store brands versus national brandpromotions. Journal of Marketing, 65 (January), 7189

S2 Ailawadi, K.L., Pauwels, K. and Steenkamp, J.B.E.M. (2008). Private-label use and store loyalty. Journal of Marketing, 72 (6), 1930S3 Baltas, George. (1997). Determinants of store brand choice: a behavioral analysis. The Journal of Product and Brand Management, 6

(5), 315324S4 Baltas, G., Doyle, P., & Dyson, P. (1997). A model of consumer choice for national vs. private label brands. Journal of the Operational

Research Society, 48 (10), 988995S5 Baltas, G., & Doyle, P. (1998). A model of consumer choice for national vs. private label brands. Journal of the Operational Research

Society, 49 (8), 790798S6 Baltas, G. (2003). A combined segmentation and demand model for store brands. European Jl of Marketing, 37,14991513S7 Baltas, G., & Argouslidis, P. C. (2007). Consumer characteristics and demand for store brands. International Journal of Retail &

Distribution Management, 35 (5), 328341S8 Batra, R., & Sinha, I. (2000). Consumer-level factors moderating the success of private label brands. Journal of retailing, 76 (2),

175191S9 Beneke, J., Greene, A., Lok, I., & Mallett, K. (2012). The influence of perceived risk on purchase intent the case of premium grocery

private label brands in South Africa,Jl of Product & Brand Management, 21 (1), 414S10 Berges, F., Hassen, D., Monier-Dilhan, S., & Raynal, H. (2000). Consumers decision between private labels and national brands in a

retailers chain: a mixed multinomial logit application. Gestion, 3, 4157S11 Bettman, J. R. (1974). Relationship of information-processing attitude structures to private brand purchasing behavior. Journal of

Applied Psychology, 59 (1), 7983S12 Bonfrer, Andre and Pradeep K. Chintagunta, (2004). Store brands: who buys them and what happens to retail prices when they are

introduced? Review of Industrial Organization, 24 (2) 195218S13 Bouhlal, Y., & Capps Jr, O. (2012). The impact of retail promotion on the decision to purchase private label products: the case of US

processed cheese. Agribusiness, 28 (1), 1528S14 Bronnenberg, B., & Wathieu, L. (1996). Asymmetric promotion effecS15 Burton, S., Lichtenstein, D. R., Netemeyer, R. G., & Garretson, J. A. (1

s. JouS16 ive int

S17 he com

S18 e setti

S19 ariesS20 -ima

367S21 ore br

h 41 S22 ountry

S23 groceand an examination of its psychological and behavioral correlateCotterill, R.W., Dhar, R. & Putsis, W.P. (1999). On the competitFood Marketing Policy Center. Research Report No. 33Cotterill, R. W., Putsis, Jr, W. P., & Dhar, R. (2000). Assessing tThe Journal of Business, 73 (1), 109137Cotterill, R. W., & Putsis Jr, W. P. (2000). Market share and pricIndustrial Organization, 17 (1), 1739Dhar, S. K., & Hoch, S. J. (1997). Why store brand penetration vDiallo, M. F. (2012). Effects of store image and store brand pricemarket. Journal of Retailing and Consumer Services, 19 (3), 360Erdem, T., Zhao, Y., & Valenzuela, A. (2004). Performance of stpreferences, perceptions, and risk. Journal of Marketing ResearcErdem, T., & Chang, S. R. (2012). A cross-category and cross-cJournal of the Academy of Marketing Science, 40 (1), 86101Frank, R. E., & Boyd Jr, H. W. (1965). Are private-brand-prone (4), 2735ts&brand positioning. Marketing Science, 15, 379394998). A scale for measuring attitude toward private label productsrnal of the Academy of Marketing Science, 26 (4), 293306eraction between private label and branded grocery products.

petitive interaction between private labels and national brands.

ng behavior for private labels and national brands. Review of

by retailer. Marketing Science, 16 (3), 208227ge on store brand purchase intention: application to an emerging

ands: a cross-country analysis of consumer store-brand(1), 86100

analysis of umbrella branding for national and store brands.

ry customers really different? Journal of Advertising Research, 5


Table A1 (Continued)# Study reference

S24 Garretson, J. A., Fisher, D., & Burton, S. (2002). Antecedents of private label attitude and national brand promotion attitude:similarities and differences. Journal of Retailing, 78 (2), 9199

S25 Geyskens, I., Gielens, K., & Gijsbrechts, E. (2010). Proliferating private-label portfolios: how introducing economy and premiumprivate labels influences brand choice,Journal of Marketing Research,47 (5),791807

S26 Glynn, M. S., & Chen, S. (2009). Consumer-factors moderating private label brand success: further empirical results. InternationalJournal of Retail & Distribution Management, 37 (11), 896914

S27 Hansen, K., Singh, V., & Chintagunta, P. (2006). Understanding store-brand purchase behavior across categories. Marketing Science,25 (1), 7590

S28 Hoch, S. J., & Banerji, S. (1993). When do private labels succeed. Sloan Management Review, 34 (4), 5767S29 Jin, B., & Suh, Y. G. (2005). Integrating effect of consumer perception factors in predicting private brand purchase in a Korean discount

store context. Journal of Consumer Marketing, 22 (2), 6271S30 Kara, A., Rojas-Mndez, J. I., Kucukemiroglu, O., & Harcar, T. (2009). Consumer preferences of store brands: Role of prior

experiences and value consciousness. Journal of Targeting, Measurement&Analysis for Marketing, 17, 12737S31 Lamey, L., Deleersnyder, B., Steenkamp, J. B. E., & Dekimpe, M. G. (2012). The effect of business-cycle fluctuations on private-label

share: what has marketing conduct got to do with it? Journal of Marketing, 76 (1), 119S32 Lemon, K. N., & Nowlis, S. M. (2002), Developing synergies between promotions and brands in different price-quality tiers, Journal

of Marketing Research, 171185S33 Levy, S., & Gendel-Guterman, H. (2012). Does advertising matter to store brand purchase intention? a conceptual framework. Journal

of Product & Brand Management, 21 (2), 8997S34 Martnez, E., & Montaner, T. (2008). Characterization of Spanish store brand consumers. International Journal of Retail &

Distribution Management, 36 (6), 477493S35 Ma, Y., Ailawadi, K. L., Gauri, D. K., & Grewal, D. (2011). An empirical investigation of the impact of gasoline prices on grocery

shopping behavior. Journal of Marketing, 75 (2), 1835S36 Miquel, S., Caplliure, E. M., & Aldas-Manzano, J. (2002). The effect of personal involvement on the decision to buy store brands.

Journal of Product & Brand Management, 11 (1), 618S37 Myer, J.G. (1967), Determinants of private brand attitude. Journal of Marketing Research, 4 (1), 7381S38 Raju, J. S., Sethuraman, R., & Dhar, S. K. (1995a). The introduction and performance of store brands. Management Science, 41 (6),

957978S39 Raju, J. S., Sethuraman, R., & Dhar, S. K. (1995b). National brand-store brand price differential and store brand market share. Pricing

Strategy & Practice, 3 (2), 1724S40 Richardson, P.S., Jain, A. K. & Dick, A. (1996). Household store brand proneness: A framework. Journal of Retailing, 72 (2), 159185S41 Rubio, N., & Yague, M. J. (2009). The determinants of store brand market share: a temporal and cross-sectional analysis. International

Journal of Market Research, 51 (4), 501519S42 Sayman, S., & Raju, J. S. (2004). Investigating the cross-category effects of store brands. Review of Industrial Organization, 24 (2),

129141S43 Scott- Morton, F, & Zettelmeyer, F. (2004). The strategic positioning of store brands in retailermanufacturer negotiations. Review of

Industrial Organization, 24 (2), 161194S44 Sethuraman, R. (1992). The effect of marketplace factors on private label penetration in grocery products. Marketing Science Institute,

Report #92128S45 Sethuraman, R. (2001). What makes consumers pay more for national brands than for store brands-image or quality? Review of

Marketing Science WP no. 318S46 Sethuraman, R., & Mittelstaedt, J. (1992). Coupons and private labels: A cross-category analysis of grocery products. Psychology &

Marketing, 9 (6), 487500S47 Sinha, I. & Batra, R. (1999). The effect of consumer price consciousness on private label purchase. International Journal of Research

in Marketing, 16 (3), 237251S48 Stanton, J. L., & Meloche, M. (2012). Macroeconomic Determinants of Private Label Penetration. Journal of International Food &

Agribusiness Marketing, 24 (1), 110119S49 Steenkamp, J. B. E. M., & Geyskens, I. (2014). Manufacturer and retailer strategies to impact store brand share: global integration,

local adaptation, and worldwide learning. Marketing Science 33 (1), 626S50 Szymanowski, M., & Gijsbrechts, E. (2012). Consumption-based cross-brand learning: are private labels really private? Journal of

Marketing Research, 49 (2), 231246S51 Veloutsou, C., Gioulistanis, E., & Moutinho, L. (2004). Own labels choice criteria and perceived characteristics in Greece and

Scotland: factors influencing the willingness to buy. Journal of Product&Brand Management, 13, 22841S52 Wang, Hui-Ming, Kalwani, M. U., & Akcura, T. (2007). A Bayesian multivariate Poisson regression model of cross-category store

brand purchasing behavior. Journal of Retailing and Consumer Services, 14 (6), 369382S53 Walsh, G., & Mitchell, V. W. (2010). Consumers intention to buy private label brands revisited. Journal of General Management, 35

(3), 324S54 Zielke, S., & Dobbelstein, T. (2007). Customers willingness to purchase new store brands. Journal of Product&Brand Management,

16 (2), 112121


Table A2Studies used in meta-analysis for each determinant.

Determinant v

1. NB-SB pri , S28,2. Household S10, S3. Household S7, S4. Household S10, S5. Cons. price S3, S6. Perceived r S8, S7. Perceived q S11, S8. SB familia S5, S9. SB quality , S19,10. Quality se S3411. SB image , S24,12. Basket siz 213. Trip frequ S7, S14. Store loya S7, S15. Number N 8, S3816. Category 1, S4417a. NB conc 917b. NB conc18. Retail con 1, S4819. Retail pro 4, S1720. Retail pro 7, S1921. Manuf. ad 8, S41

Note: Study n

Ailawadi, KuDeterminaof Market

Ailawadi, KuValue-contions, Jou

Ailawadi, KuPrivate-l

Assmus, GertAffects SaResearch,

Baltas, GeorgAnalysis,

Batra, Rajeevthe Succe

Bettman, JamStructureschology, 5

Blattberg, RoCompetiti

Bonfrer, AndThem andReview of

Dhar, Sanjay Varies by

Dick, Alan, APronenessManagem

Euromonitor Erdem, Tlin

tainty: CaGoods Ma

Erdem, TlinStore Braferences, 86100.

Phillin of uctioonal

cery C2735s, In

ate-laels In807.

, Katariable Studies

ce differential S2, S19 Income S2, S7,

size S2, S6, education S2, S7,

sensitivity S1, S2, isk S3, S5, uality var. S5, S8,

rity S3, S4, S7, S11

nsitivity S1, S2, S7, S15e S27, S5ency S3, S5, lty S1, S2, Bs S19, S2

price elasticity S38, S4 share top NBs S41, S4

var NB shares S19centration S19, S4motion SB S13, S1motion NB S13, S1vertising S19, S2

umbers (S#) correspond to the list in Table A1.

References

sum L. and Bari Harlam (2004), An Empirical Analysis of thents of Retail Margins: The Role of Store-brand Share, Journaling, 68 (1), 14765.sum L., Scott A. Neslin and Karen Gedenk (2001), Pursuing thescious Consumer: Store Brands Versus National Brand Promo-rnal of Marketing, 65 (January), 7189.

sum L., Koen Pauwels and Jan-Benedict E.M. Steenkamp (2008),abel Use and Store Loyalty, Journal of Marketing, 72 (6), 1930., John U. Farley and Don R. Lehmann (1984), How Advertising

Fitzell, lutioProd

Frank, RGro(4),

GeyskenPrivLab791

Gielens

les: Meta-analysis of Econometric Results, Journal of Marketing

21 (1), 6574.e (1997), Determinants of Store Brand Choice: A Behavioral

Journal of Product & Brand Management, 6 (5), 31524. and Indrajit Sinha (2000), Consumer-level Factors Moderatingss of Private Label Brands, Journal of Retailing, 76 (2), 17591.es R. (1974), Relationship of Information-processing Attitude

to Private Brand Purchasing Behavior, Journal of Applied Psy-9 (1), 7983.bert C. and Ken Wisniewski (1989), Price Induced Patterns ofon, Marketing Science, 8 (Fall), 291309.re and Pradeep K. Chintagunta (2004), Store Brands: Who Buys

What Happens to Retail Prices When They are Introduced?, Industrial Organization, 24 (2), 195218.K. and Stephen J. Hoch (1997), Why Store Brand PenetrationRetailer, Marketing Science, 15 (3), 20827.run Jain and Paul Richardson (1995), Correlates of Store Brand: Some Empirical Observations, Journal of Brand and Productent, 4 (4), 1522.(2012), Passport Global Market Information Database,.

and Michael P. Keane (1996), Decision-making Under Uncer-pturing Dynamic Brand Choice Processes in Turbulent Consumerrkets, Marketing Science, 15 (1), 120., Ying Zhao and Ana Valenzuela (2004), Performance ofnds: A Cross-country Analysis of Consumer Store-brand Pre-Perceptions, and Risk, Journal of Marketing Research, 41 (1),

Growth?,Hoch, Stephe

ceed?, SlHoch, Stephe

Managemvania.

Kumar, NirmStrategy:

Lal, Rajiv (19Journal o

Lamey, Lien, G. Dekimlabel SharMarketing

Nielsen/PLMPauwels, Koe

Entry?, MRaju, Jagmoh

duction an95778.

(1995b), Market Sh

Richardson, Store Bra59185.

Rosenthal, Ro S31, S41, S44, S46, S49, S5313, S16, S17, S18, S19, S23, S27, S33, S34, S35, S37, S40, S43

10, S11, S13, S23, S27, S34, S35, S40, S4313, S19, S23, S34, S37, S40

5, S6, S7, S8, S26, S29, S34, S47, S51, S539, S11, S20, S21, S22, S26, S40, S43, S47, S5026, S28, S29, S36

11, S21, S22, S40, S54 S22, S21, S28, S33, S49, S50, S51, S53

S29, S30, S53, S54

41, S44, S46, S49, S51, S5212, S34, S43, S44, S46

, S49, S19, S25, S31, S32, S42, S44, S46, S49, S42, S44, S46, S49, S43, S44, S49

p B. (1992), Private Label Marketing in the 1990s: The Evo-Price Labels into Global Brands, New York: Global Book

ns.d E. and Harper W. Boyd Jr. (1965), Are Private-brand-proneustomers Really Different?, Journal of Advertising Research, 5.ge, Gielens Katrijn and Els Gijsbrechts (2010), Proliferatingbel Portfolios: How Introducing Economy and Premium Privatefluences Brand Choice, Journal of Marketing Research, 47 (5),

rijn (2012), New Products: The Antidote to Private Label

Journal of Marketing Research, 49 (3), 40823.n J. and Shumeet Banerji (1993), When Do Private Labels Suc-oan Management Review, 34, 57.n J. and Leonard M. Lodish (1998), Store Brands and Categoryent, Working Paper, The Wharton School, University of Pennsyl-

alya and Jan-Benedict E.M. Steenkamp (2007), Private LabelHow to Meet the Store Brand Challenge, Harvard Business Press.90), Manufacturer Trade Deals and Retail Price Promotions,

f Marketing Research, 27, 42844.Barbara Deleersnyder, Jan-Benedict E.M. Steenkamp and Marnikpe (2012), The Effect of Business-cycle Fluctuations on Private-e: What Has Marketing Conduct Got To Do With It?, Journal of, 76 (1), 119.

A (2012), Annual International Private Label Yearbook,.n and Shuba Srinivasan (2004), Who Benefits from Store Brandarketing Science, 23 (3), 36490.

an S., Raj Sethuraman and Sanjay K. Dhar (1995a), The Intro-d Performance of Store Brands, Management Science, 41 (6),

, andNational Brand-store Brand Price Differential and Store Brandare, Pricing Strategy & Practice, 3 (2), 1724.

Paul S., Arun K. Jain and Alan Dick (1996), Householdnd Proneness: A Framework, Journal of Retailing, 72 (2),

bert (1991), , Sage Publications. Incorporated.


Sayman, Serdar, Stephen J. Hoch and Jagmohan S. Raju (2002), Positioningof Store Brands, Marketing Science, 21 (4), 37897.

Schmalensee, Richard (1978), Entry Deterrence in the Ready-to-rat BreakfastCereal Industry, Bell Journal of Economics, 9 (2), 30527.

Scott-Morton, Fiona and Florian Zettelmeyer (2004), The Strategic Positioningof Store Brands in RetailerManufacturer Negotiations, Review of Indus-trial Organization, 24 (2), 16194.

Sethuraman, Raj (1992), The effect of marketplace factors on private label pen-etration in grocery products, Marketing Science Institute. Report #92-128.

(2003), Measuring National Brands Equity over StoreBrands, Review of Marketing Science, 1 (1), 122.

(2009), Assessing the External Validity of AnalyticalResults from National Brand and Store Brand Competition Models, Mar-keting Science, 28 (4), 75981.

Sinha, Indrajit and Rajeev Batra (1999), The Effect of Consumer Price Con-sciousness on Private Label Purchase, International Journal of Research inMarketing, 16 (3), 23751.

Steenkamp, Jan-Benedict E.M. and Inge Geyskens (2014), Manufacturerand Retailer Strategies to Impact Store Brand Share: Global Integration,Local Integration, and Worldwide Learning, Marketing Science, 33 (1),626.

ter Braak, Anne, Marnik G. Dekimpe and Inge Geyskens (2013), RetailerPrivate-label Margins: The Role of Supplier and QualityTier Differenti-ation, Journal of Marketing, 77 (4), 86103.

Wang, Hui-Ming, Manohar U. Kalwani and M. Tolga Akcura (2007), ABayesian Multivariate Poisson Regression Model of cross-category StoreBrand Purchasing Behavior, Journal of Retailing and Consumer Services,14 (6), 36982.

Determinants of Store Brand ShareIntroductionFrameworkDrivers of price utilityDrivers of non-price utility

ProcedureData compilationMethod for empirical generalization

ResultsNational brand store brand price differentialDemographicsConsumer perceptionsShopping trip variablesProduct-market characteristicsRetailer and manufacturer marketing mix

ConclusionResult highlightsFuture research suggestions

ReferencesReferences

Documents

1-s2.0-S002243591400027X-main