Customer-based brand equity is d efined as the differential effect of brand knowled ge on consumer response to the marketing of the brand. A brand is said to have positive (negative) customer-basedbrand equity when consumers reactmore (less) favorably toan element of the marketing mix.
Customer-based brand equity is d efined as the differential effect of brand knowled ge on consumer response to the marketing of the brand. A brand is said to have positive (negative) customer-basedbrand equity when consumers reactmore (less) favorably toan element of the marketing mix.
Customer-based brand equity is d efined as the differential effect of brand knowled ge on consumer response to the marketing of the brand. A brand is said to have positive (negative) customer-basedbrand equity when consumers reactmore (less) favorably toan element of the marketing mix.
Conceptualizing, Measuring, and Managing Customer-Based Brand Equity
Author(s): Kevin Lane Keller
Source: Journal of Marketing, Vol. 57, No. 1 (Jan., 1993), pp. 1-22 Published by: American Marketing Association Stable URL: http://www.jstor.org/stable/1252054 . Accessed: 30/09/2013 12:03 Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp . JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org. . American Marketing Association is collaborating with JSTOR to digitize, preserve and extend access to Journal of Marketing. http://www.jstor.org This content downloaded from 152.3.102.242 on Mon, 30 Sep 2013 12:03:07 PM All use subject to JSTOR Terms and Conditions Kevin Lane Keller Conceptualizing, Measuring, and Managing Customer-Based Brand Equity The author presents a conceptual mod el of brand equity fromthe perspective of the ind ivid ual consumer. Customer-based brand equity is d efined as the d ifferential effectof brand knowled ge on consumer re- sponse tothe marketing of the brand . A brand issaid tohave positive (negative) customer-based brand equity when consumers reactmore (less) favorably toan elementof the marketing mix for the brand than they d otothe same marketing mix elementwhen itisattributed toa fictitiously named or unnamed version of the prod uct or service. Brand knowled ge is conceptualized accord ing toan associative network memory mod el in terms of two components, brand awareness and brand image (i.e., a setof brand associations). Customer-based brand equity occurs when the consumer isfamiliar with the brand and hold s some favorable,strong, and unique brand associations in memory. Issues in build ing,measuring, and managing customer-based brand equity are d iscussed , aswell as areasfor future research. M UCH attention hasbeen d evoted recently tothe concept of brand equity (Aaker and Biel 1992; Leuthesser 1988; Maltz 1991). Brand equity hasbeen viewed froma variety of perspectives(Aaker 1991; Farquhar 1989; Srivastava and Shocker 1991; Tauber 1988). In a general sense, brand equity isd efined in termsof the marketing effects uniquely attributable to the brand -for example, when certain outcomesre- sultfromthe marketing of a prod uct or service be- cause of itsbrand name thatwould notoccur if the same prod uct or service d id nothave thatname. There have been two general motivationsfor stud ying brand equity. One isa financially based mo- tivation toestimate the value of a brand more pre- cisely for accountingpurposes (in termsof assetval- uation for the balance sheet) or for merger,acquisition, Kevin Lane Keller isAssociate Professor of Marketing and Fletcher Jones Faculty Scholar for 1992-1993, Grad uate School of Business, Stanford Univerity. Thisarticle waswritten while the author was Visiting Profes- sor atthe Australian Grad uate School of Management, University of New South Wales,Syd ney, Australia. He thanksDavid Aaker, Sheri Brid ges, Deborah Macinnis, John Roberts, John Rossiter, Richard Stae- lin,Jennifer Aaker, and the anonymous JMreviewers for d etailed , con- structive comments. Journal of Marketing Vol. 57 (January 1993),1-22 or d ivestiture purposes. Several d ifferentmethod sof brand valuation have been suggested (Barwise etal. 1989; Wentz 1989). For example, Interbrand Group hasused a subjective multiplier of brand profits based on the brand 's performance along seven d imensions (lead ership,stability, market stability, interational- ity,trend ,support, and protection); Grand Metropol- itan hasvalued newly acquired brand s by d etermining the d ifference between the acquisition price and fixed assets. Simon and Sullivan (1990) d efine brand equity in termsof the incremental d iscounted future cash flows thatwould resultfroma prod ucthaving itsbrand name in comparison with the proceed s thatwould accrue if the same prod uct d id nothave thatbrand name. Based on the financial marketvalue of the company, their estimation technique extractsthe value of brand eq- uity fromthe value of a firm'sother assets. A second reason for stud ying brand equity arises froma strategy-based motivation to improve market- ingprod uctivity. Given higher costs,greater compe- tition,and flattening d emand in many markets, firms seek toincrease the efficiency of their marketing ex- penses. As a consequence, marketersneed a more thorough und erstand ing of consumer behavior asa ba- Customer-Based Brand Equity / 1 This content downloaded from 152.3.102.242 on Mon, 30 Sep 2013 12:03:07 PM All use subject to JSTOR Terms and Conditions sisfor making better strategic d ecisionsabout target marketd efinition and prod uctpositioning, aswell as better tactical d ecisionsabout specificmarketing mix actions. Perhaps a firm'smostvaluable assetfor im- provingmarketingprod uctivity isthe knowled ge that hasbeen created aboutthe brand in consumers' mind s fromthe firm'sinvestmentin previousmarketingpro- grams. Financial valuation issueshave little relevance if no und erlying value for the brand hasbeen created or if managers d onotknow how to exploit thatvalue by d evelopingprofitable brand strategies. The goal of thisarticle istoassist managers and researcherswhoare interested in the strategicaspects of brand equity. Specifically, brand equity is concep- tualized fromthe perspective of the ind ivid ual con- sumer and a conceptual framework is provid ed of what consumersknow aboutbrand sand whatsuch knowl- ed ge implies for marketingstrategies. Customer-based brand equity isd efined asthe d ifferential effectof brand knowled ge on consumer response tothe marketing of the brand . That is, customer-based brand equity in- volvesconsumers' reactionstoan elementof the mar- keting mix for the brand in comparison with their re- actionstothe same marketing mix elementattributed toa fictitiously named or unnamed version of the prod uct or service. Customer-based brand equity oc- curswhen the consumer isfamiliar with the brand and hold ssome favorable,strong, and unique brand as- sociationsin memory. Conceptualizing brand equity fromthis perspec- tive isuseful because it suggests both specificguid e- lines for marketingstrategies and tacticsand areas where research can be useful in assistingmanagerial d ecision making. Two importantpointsemerge from this conceptualization. First, marketersshould take a broad view of marketingactivity for a brand and rec- ognize the variouseffectsithason brand knowled ge, aswell ashow changes in brand knowled ge affectmore trad itional outcome measuressuch assales. Second , marketersmustrealize thatthe long-term successof all future marketingprograms for a brand is greatly affected by the knowled ge aboutthe brand in memory thathasbeen established by the firm'sshort-termmar- keting efforts. In short, because the contentand struc- ture of memory for the brand will influence the ef- fectivenessof future brand strategies, itiscritical that managers und erstand how their marketingprograms affectconsumer learning and thus subsequent recall for brand -related information. The nextsection provid es a conceptualization of brand knowled ge by applying some basic memory no- tions. Brand knowled ge is d efined in termsof two components, brand awarenessand brand image. Brand awarenessrelatestobrand recall and recognition per- formance by consumers. Brand image referstothe set of associationslinked tothe brand thatconsumershold in memory. Then the concept of customer-based brand equity isconsid ered in more d etail by d iscussion of how itcan be built,measured ,and managed . After the conceptual framework is summarized ,areasfor future research are id entified . Brand Knowled ge Background A brand can be d efined as"a name,term,sign,sym- bol, or d esign, or combination of themwhich isin- tend ed to id entify the good s and servicesof one seller or group of sellersand tod ifferentiate themfromthose of competitors" (Kotler 1991; p. 442). These ind ivid - ual brand components are here called "brand id enti- ties" and their totality "the brand ." Some basicmem- ory principles can be used tound erstand knowled ge aboutthe brand and how itrelatestobrand equity. The importance of knowled ge in memory toconsumer d ecision making hasbeen well d ocumented (Alba, Hutchinson,and Lynch 1991). Und erstand ing the contentand structure of brand knowled ge is important because they influence whatcomestomind when a consumer thinksabouta brand -for example, in re- sponse to marketingactivity for thatbrand . Most wid ely accepted conceptualizations of mem- ory structure involve some type of associative mod el formulation (And erson 1983; Wyer and Srull 1989). For example, the "associative network memory mod el" viewssemantic memory or knowled ge as consisting of a setof nod esand links. Nod esare stored infor- mation connected by linksthat vary in strength. A "spread ing activation" process fromnod e tonod e d e- terminesthe extentof retrieval in memory (Collins and Loftus 1975; Raaijmakers and Shiffrin 1981; Ratcliff and McKoon 1988). A nod e becomesa po- tential source of activation for other nod eseither when external information is being encod ed or when inter- nal information isretrieved from long-termmemory. Activation can spread fromthisnod e toother linked nod esin memory. When the activation of another nod e exceed s some threshold level, the information con- tained in thatnod e isrecalled . Thus, the strength of association between the activated nod e and all linked nod esd eterminesthe extentof this "spread ing acti- vation" and the particular information thatcan be re- trieved from memory. For example, in consid ering a softd rink purchase, a consumer may think of Pepsi because of its strong association with the prod uct cat- egory. Consumer knowled ge most strongly linked to Pepsi should alsothen come to mind , such as per- ceptions of its taste, sugar and caffeine content, or even recalled images froma recent ad vertising cam- paign or pastprod uctexperiences. Consistentwith an associative network memory 2 / Journal of Marketing, January 1993 This content downloaded from 152.3.102.242 on Mon, 30 Sep 2013 12:03:07 PM All use subject to JSTOR Terms and Conditions mod el, brand knowled ge is conceptualized as con- sisting of a brand nod e in memory towhich a variety of associations are linked . Given this conceptualiza- tion, the key question is, what properties d othe brand nod e and brand associations have? As d eveloped here, the relevantd imensions that d istinguish brand knowl- ed ge and affectconsumer response are the awareness of the brand (in termsof brand recall and recognition) and the favorability, strength, and uniqueness of the brand associations in consumer memory. These d i- mensions are affected by other characteristics of and relationships among the brand associations. For ex- ample, factorsrelated tothe type of brand association (such asitslevel of abstraction and qualitative nature) and the congruity among brand associations, among others, affectthe favorability, strength, and unique- ness of brand associations. To simplify the d iscus- sion, emphasis is placed on the brand name compo- nentof the brand id entities, d efined as "that part of a brand which can be vocalized " (Kotler 1991,p. 442), though other components of the brand id entities (e.g., brand logo or symbol) are consid ered also. Brand Awareness The firstd imension d istinguishing brand knowled ge isbrand awareness. Itisrelated tothe strength of the brand nod e or trace in memory, asreflected by con- sumers' ability to id entify the brand und er d ifferent cond itions (Rossiter and Percy 1987). In other word s, how well d othe brand id entities serve their function? In particular, brand name awareness relates tothe likelihood thata brand name will come tomind and the ease with which itd oes so. Brand awareness con- sists of brand recognition and brand recall perfor- mance. Brand recognition relates toconsumers' abil- ity toconfirm prior exposure tothe brand when given the brand asa cue. In other word s, brand recognition requires thatconsumers correctly d iscriminate the brand as having been seen or heard previously. Brand recall relatestoconsumers' ability toretrieve the brand when given the prod uctcategory, the need s fulfilled by the category, or some other type of probe as a cue. In other word s, brand recall requires thatconsumers cor- rectly generate the brand from memory. The relative importance of brand recall and recognition d epend s on the extenttowhich consumers make d ecisions in the store (where they potentially may be exposed tothe brand ) versus outsid e the store, among other factors (Bettman 1979; Rossiter and Percy 1987). Brand rec- ognition may be more important tothe extentthat prod uct d ecisions are mad e in the store. Brand awareness plays an important role in con- sumer d ecision making for three major reasons. First, itis important thatconsumers think of the brand when they think aboutthe prod uctcategory. Raising brand awareness increases the likelihood thatthe brand will be a member of the consid eration set (Baker etal. 1986; Ned ungad i 1990), the hand ful of brand sthat receive serious consid eration for purchase. Second , brand awareness can affectd ecisions aboutbrand sin the consid eration set, even if there are essentially no other brand associations. For example, consumershave been shown to ad opt a d ecision rule to buy only fa- miliar, well-established brand s (Jacoby,Syzabillo, and Busato-Schach 1977; Roselius 1971). In low involve- mentd ecision settings, a minimumlevel of brand awareness may be sufficientfor prod uct choice, even in the absence of a well-formed attitud e (Bettman and Park 1980; Hoyer and Brown 1990; Park and Lessig 1981). The elaboration likelihood mod el (Petty and Cacioppo1986) suggests thatconsumers may base choices on brand awareness consid erations when they have low involvement, which could resultfromeither a lack of consumer motivation (i.e., consumersd onot care aboutthe prod uct or service) or a lack of con- sumer ability (i.e., consumers d onotknow anything else aboutthe brand s). Finally, brand awareness af- fects consumer d ecision makingby influencing the formation and strength of brand associations in the brand image. A necessary cond ition for the creation of a brand image isthata brand nod e hasbeen estab- lished in memory, and the nature of thatbrand nod e should affecthow easily d ifferentkind s of informa- tion can become attached tothe brand in memory. Brand Image Though brand image long hasbeen recognized as an importantconcept in marketing(e.g., Gard ner and Levy 1955), there is less agreement on its appropriate d ef- inition (Dobni and Zinkhan 1990). Consistentwith d efinitions by Herzog(1963) and Newman (1957), amongothers, and an associative network memory mod el of brand knowled ge, brand image is d efined here as perceptions abouta brand asreflected by the brand associations held in consumer memory. Brand associations are the other informational nod es linked tothe brand nod e in memory and contain the meaning of the brand for consumers. The favorability,strength, and uniqueness of brand associations are the d imen- sions d istinguishing brand knowled ge that play an im- portant role in d etermining the d ifferential response thatmakes up brand equity, especially in high in- volvementd ecision settings. Before consid ering those d imensions, itisuseful toexamine the d ifferent types of brand associations that may be present in consumer memory. Types of brand associations. Brand associations take d ifferentforms. One way to d istinguish among brand associationsis by their level of abstraction (Alba and Hutchinson 1987; Chattopad hyay and Alba 1988; Johnson 1984; Russoand Johnson 1980)-that is,by Customer-Based Brand Equity / 3 This content downloaded from 152.3.102.242 on Mon, 30 Sep 2013 12:03:07 PM All use subject to JSTOR Terms and Conditions how much information issummarized or subsumed in the association. Along thisd imension,brand associ- ationscan be classified intothree major categories of increasingscope: attributes,benefits, and attitud es. Several ad d itional d istinctionscan be mad e within these categoriesaccord ing tothe qualitative nature of the association. Attributesare those d escriptive featuresthatchar- acterize a prod uct or service-whata consumer thinks the prod uct or service isor hasand whatisinvolved with its purchase or consumption. Attributescan be categorized in a variety of ways(Myers and Shocker 1981). Here, attributesare d istinguished accord ing to how d irectly they relate to prod uct or service perfor- mance. Prod uct-related attributes are d efined as the ingred ientsnecessary for performing the prod uct or service function soughtby consumers. Hence,they relate toa prod uct'sphysical composition or a ser- vice's requirements. Prod uct-related attributes vary by prod uct or service category. Non-prod uct-related at- tributesare d efined asexternal aspects of the prod uct or service thatrelate toits purchase or consumption. The four main types of non-prod uct-related attributes are (1) price information,(2) packaging or prod uct appearance information,(3) user imagery (i.e., what type of person usesthe prod uct or service), and (4) usage imagery (i.e., where and in what types of sit- uationsthe prod uct or service is used ). Because prod uct-related attributesare more com- monly acknowled ged ,only non-prod uct-related attri- butesare elaborated here. The price of the prod uct or service is consid ered a non-prod uct-related attribute because it represents a necessary step in the purchase process but typically d oes notrelate d irectly tothe prod uctperformance or service function. Price is a particularly important attribute association because consumersoften have strong beliefsaboutthe price and value of a brand and may organize their prod uct category knowled ge in termsof the price tiersof d if- ferentbrand s (Blattberg and Wisniewski 1989). Sim- ilarly,packaging is consid ered part of the purchase and consumption processbut, in most cases, d oesnot d irectly relate tothe necessary ingred ients for prod uct performance. User and usage imagery attributescan be formed d irectly froma consumer'sown experi- encesand contactwith brand usersor ind irectly through the d epiction of the target marketascommunicated in brand ad vertising or by some other source of infor- mation (e.g., word of mouth). Associationsof a typ- ical brand user may be based on d emographic factors (e.g.,sex,age,race, and income), psychographic fac- tors (e.g., accord ing toattitud estoward career,pos- sessions, the environment, or political institutions), and other factors. Associationsof a typical usage sit- uation may be based on the time of d ay,week, or year, the location (insid e or outsid e the home), or the type of activity (formal or informal),among other as- pects. User and usage image attributescan also pro- d uce brand personality attributes. Plummer (1985) as- serts thatone component of brand image is the personality or character of the brand itself. He sum- marizesresearch d emonstrating thatbrand scan be characterized by personality d escriptors such as "youthful," "colorful," and "gentle." These types of associationsseemtoarise mostoften asa resultof inferencesaboutthe und erlying user or usage situa- tion. Brand personality attributes may alsoreflect emotionsor feelings evoked by the brand . Benefits are the personal value consumersattach tothe prod uct or service attributes-that is,whatcon- sumersthink the prod uct or service can d ofor them. Benefitscan be further d istinguished intothree cate- goriesaccord ing tothe und erlying motivationstowhich they relate (Park,Jaworski, and Maclnnis 1986): (1) functional benefits,(2) experiential benefits,and (3) symbolic benefits. Functional benefits are the more intrinsic ad vantages of prod uct or service consump- tion and usually correspond tothe prod uct-related at- tributes. These benefitsoften are linked to fairly basic motivations, such as physiological and safety need s (Maslow 1970), and involve a d esire for problem re- moval or avoid ance (Fennell 1978; Rossiter and Percy 1987). Experiential benefits relate towhatitfeelslike touse the prod uct or service and also usually corre- spond tothe prod uct-related attributes. These benefits satisfy experiential need ssuch as sensory pleasure, variety, and cognitive stimulation. Symbolicbenefits are the more extrinsic ad vantages of prod uct or ser- vice consumption. They usually correspond tonon- prod uct-related attributesand relate to und erlying need s for social approval or personal expression and outer- d irected self-esteem. Hence, consumers may value the prestige,exclusivity, or fashionability of a brand be- cause of how itrelatestotheir self-concept(Solomon 1983). Symbolic benefitsshould be especially rele- vantfor socially visible,"bad ge" prod ucts. Brand attitud esare d efined asconsumers' overall evaluationsof a brand (Wilkie 1986). Brand attitud es are important because they often formthe basisfor consumer behavior (e.g., brand choice). Though d if- ferentmod elsof brand attitud eshave been proposed , one wid ely accepted approach isbased on a multiat- tribute formulation in which brand attitud esare a function of the associated attributesand benefitsthat are salientfor the brand . Fishbein and Ajzen (1975; Ajzen and Fishbein 1980) proposed whathasbeen probably the mostinfluential multiattribute mod el to marketing(Bettman 1986). This expectancy-value mod el viewsattitud esasa multiplicative function of (1) the salientbeliefsa consumer hasaboutthe prod - uctor service (i.e., the extenttowhich consumersthink the brand hascertain attributesor benefits) and (2) the 4 / Journal of Marketing, January 1993 This content downloaded from 152.3.102.242 on Mon, 30 Sep 2013 12:03:07 PM All use subject to JSTOR Terms and Conditions evaluative jud gment of those beliefs (i.e., how good or bad itisthatthe brand hasthose attributesor ben- efits). Brand attitud escan be related tobeliefs about prod uct-related attributesand the functional and ex- periential benefits, consistentwith work on perceived quality (Zeithaml 1988). Brand attitud escan alsobe related tobeliefsabout non-prod uct-related attributes and symbolic benefits (Rossiter and Percy 1987),con- sistentwith the functional theory of attitud es (Katz 1960; Lutz 1991), which maintainsthatattitud escan serve a "value-expressive" function by allowing in- d ivid ualsto express their self-concepts. Because itis d ifficultto specify correctly all of the relevantattri- butesand benefits, researchers build ing multiattribute mod elsof consumer preference have includ ed a gen- eral component of attitud e toward the brand thatisnot captured by the attribute or benefitvaluesof the brand (Park 1991; Srinivasan 1979). Moreover, as noted previously, research alsohasshown thatattitud escan be formed by less thoughtful d ecision making(Chaiken 1986; Petty and Cacioppo1986)-for example, on the basisof simple heuristicsand d ecision rules. If con- sumerslack either the motivation or ability toevaluate the prod uct or service,they may use signals or "ex- trinsiccues" (Olson and Jacoby 1972) toinfer prod uct or service quality on the basisof what they d oknow aboutthe brand (e.g.,prod uctappearance such ascolor or scent). Thus, the d ifferent types of brand associations makingup the brand image includ e prod uct-related or non-prod uct-related attributes; functional, experien- tial, or symbolicbenefits; and overall brand attitud es. These associationscan vary accord ing totheir favor- ability,strength, and uniqueness. Favorability of brand associations. Associations d iffer accord ing tohow favorably they are evaluated . The successof a marketingprogram isreflected in the creation of favorable brand associations-that is, con- sumersbelieve the brand hasattributesand benefits that satisfy their need sand wantssuch thata positive overall brand attitud e isformed . MacKenzie (1986) summarizesresearch evid ence suggesting thatthe "evaluative jud gment" component of expectancy-value mod els of attitud e (i.e., con- sumer perceptions of the favorability of an attribute) isboth conceptually and empirically related toattri- bute importance. Specifically, attribute importance has been equated with polarity of attribute evaluation (Ajzen and Fishbein 1980; Fishbein and Ajzen 1975). In other word s,consumersare unlikely toview an attribute or benefitas very good or bad if they d onotalsocon- sid er ittobe very important. Hence, itisd ifficultto create a favorable association for an unimportant at- tribute. Notall associationsfor a brand ,however,will be relevantand valued in a purchase or consumption d e- cision. For example, consumersoften have an asso- ciation in memory fromthe brand tothe prod uct or package color. Though thisassociation may facilitate brand recognition or awarenessor lead toinferences about prod uctquality, it may not always be consid - ered a meaningful factor in a purchase d ecision. Moreover, the evaluationsof brand associations may be situationally or context-d epend ent and vary ac- cord ing toconsumers' particular goals in their pur- chase or consumption d ecisions (Day, Shocker,and Srivastava 1979). An association may be valued in one situation butnotanother (Miller and Ginter 1979). For example,speed and efficiency of service may be very important when a consumer isund er time pres- sure but may have little impact when a consumer is lesshurried . Strength of brand associations. Associations can be characterized also by the strength of connection to the brand nod e. The strength of associations d epend s on how the information entersconsumer memory (en- cod ing) and how itismaintained as part of the brand image (storage). Strength is a function of both the amountor quantity of processing the information re- ceives at encod ing(i.e., how much a person thinks aboutthe information) and the nature or quality of the processing the information receivesat encod ing(i.e., the manner in which a person thinksaboutthe infor- mation). For example, the levels-or d epth-of-processing approach (Craik and Lockhart 1972; Craik and Tulv- ing1975; Lockhart,Craik, and Jacoby 1976) main- tainsthatthe more the meaning of information isat- tend ed to d uringencod ing, the stronger the resulting associationsin memory will be. Thus, when a con- sumer actively thinksaboutand "elaborates" on the significance of prod uct or service information,stronger associationsare created in memory. This strength, in turn,increasesboth the likelihood thatinformation will be accessible and the ease with which itcan be re- called by "spread ing activation." Cognitive psychologists believe memory is ex- tremely d urable, sothatonce information becomes stored in memory its strength of association d ecays very slowly (Loftus and Loftus 1980). Though "avail- able" and potentially retrievable in memory, infor- mation may notbe "accessible" and easily retrieved without strongly associated remind ersor retrieval cues (Tulving and Psotka 1971). Thus, the particular as- sociationsfor a brand thatare salientand "come to mind " d epend on the contextin which the brand is consid ered . The larger the number of cueslinked to a piece of information,however, the greater the like- lihood thatthe information can be recalled (Isen 1992). Uniqueness of brand associations. Brand associ- Customer-Based Brand Equity / 5 This content downloaded from 152.3.102.242 on Mon, 30 Sep 2013 12:03:07 PM All use subject to JSTOR Terms and Conditions ations may or may notbe shared with other competing brand s. The essence of brand positioning isthatthe brand has a sustainable competitive ad vantage or "unique sellingproposition" that gives consumersa compelling reason for buying that particular brand (Aaker 1982; Riesand Trout 1979; Wind 1982). These d ifferences may be communicated explicitly by mak- ing d irect comparisons with competitors or may be highlighted implicitly without stating a competitive point of reference. Furthermore,they may be based on prod uct-related or non-prod uct-related attributesor functional,experiential, or image benefits. The presence of strongly held ,favorably evaluated associationsthatare unique tothe brand and imply superiority over other brand siscritical toa brand 's success. Yet, unlessthe brand hasno competitors, the brand will most likely share some associationswith other brand s. Shared associationscan help toestablish category membership(Maclnnis and Nakamoto 1991) and d efine the scope of competition with other prod - uctsand services (Sujan and Bettman 1989). Research on noncomparable alternatives (Bettman and Sujan 1987; Johnson 1984; Park and Smith 1989) suggests thateven if a brand d oesnotface d irect competition in its prod uctcategory, and thusd oesnotshare prod uct- related attributeswith other brand s, itcan still share more abstractassociationsand face ind irect compe- tition in a more broad ly d efined prod uctcategory. Thus, though a railroad may not compete d irectly with an- other railroad , itstill competesind irectly with other formsof transportation, such as airlines,cars, and buses. A prod uct or service category can be characterized also by a setof associationsthatinclud e specific be- liefsabout any member in the category in ad d ition to overall attitud estoward all membersin the category. These beliefsinclud e many of the prod uct-related at- tributesfor the relevant brand s, aswell asmore d e- scriptive attributesthatd onot necessarily relate to prod uct or service performance (e.g., the color of a prod uct, such asred for ketchup). Certain attributes or benefits may be consid ered "prototypical" and es- sential toall brand sin the category, and a specific brand may be consid ered an "exemplar" thatismost representative of the prod uct or service category (Cohen and Basu 1987; Ned ungad i and Hutchinson 1985; Rosch and Mervis 1975; Ward and Loken 1986). For ex- ample, consumers mightexpect a running shoe to pro- vid e support and comfort, be builtwell enough tolast through repeated wearings, and so on, and they may believe thatNike or some other lead ing brand best represents a running shoe. Similarly, consumers might expect a bank tooffer a variety of checking and sav- ingsaccounts,provid e branch and electronic d elivery services, and so on, and they may consid er Bank of America or some other marketlead er tobe the best example of a bank. Because the brand is linked tothe prod uct cate- gory, some prod uctcategory associations may be- come linked tothe brand ,either in termsof specific beliefsor overall attitud es. Prod uct category attitud es can be a particularly important d eterminantof con- sumer response. For example, if a consumer thinks banksare basically "unfriend ly" and "bad ," he or she probably will have similarly unfavorable beliefsabout and attitud e toward any particular bank simply by vir- tue of its membership in the category. Thus,in almost all cases, some prod uctcategory associationsthatare linked tothe brand are shared with other brand sin the category. Note thatthe strength of the brand associ- ations with the prod uctcategory is an important d eterminantof brand awareness (Ned ungad i and Hutchinson 1985; Ward and Loken 1986). Competitive overlap with other brand sassociated with the prod uctcategory d oeshave a d ownsid e,how- ever, in termsof possible consumer confusion. For example, Keller (1987) and Burke and Srull (1988) have shown thatthe number of competing brand sad - vertising in a prod uctcategory can affectconsumers' ability torecall communication effectsfor a brand by creating "interference" in memory. Keller (1991b) also showed that though these interference effectscan pro- d uce lower brand evaluations,they can be overcome through the use of ad retrieval cues-that is, d istinc- tive ad execution information thatis present when a consumer actually makesa brand evaluation (e.g., at the point of purchase). Interaction among characteristics of brand asso- ciations. The level of abstraction and qualitative na- ture of brand associationsshould affecttheir favora- bility,strength, and uniqueness. For example,image- related attributes, such asuser type or usage situation, may easily create unique associations. Abstractas- sociations (e.g., benefitsand especially attitud es), in contrast, tend tobe inherently more evaluative be- cause of the embed d ed meaningthey contain. Be- cause of thisevaluative nature, abstractassociations tend tobe more d urable and accessible in memory than the und erlying attribute information (Chattopad hyay and Alba 1988). In fact, brand attitud es may be stored and retrieved in memory separately fromthe und er- lying attribute information (Lynch,Mamorstein, and Weigold 1988). One important reason for consid ering brand atti- tud estobe a brand association isthat they can vary in strength (Farquhar 1989). Attitud e strength hasbeen measured by the reaction time need ed toevaluative queries aboutthe attitud e object(Fazio etal. 1986). Ind ivid ualswhocan evaluate an attitud e objectquickly are assumed tohave a highly accessible attitud e. Re- 6 / Journal of Marketing, January 1993 This content downloaded from 152.3.102.242 on Mon, 30 Sep 2013 12:03:07 PM All use subject to JSTOR Terms and Conditions search hasshown thatattitud esformed fromd irectbe- havior or experience are more accessible than atti- tud esbased on information or ind irectformsof behavior (Fazio and Zanna 1981). Highly accessible brand at- titud esare more likely tobe activated spontaneously upon exposure tothe brand and guid e subsequent brand choices (Berger and Mitchell 1989; Fazio,Powell, and Williams 1989). Figure 1 summarizesthe d imensionsof brand knowled ge. Congruence of brand associations. The favora- bility and strength of a brand association can be af- fected by other brand associationsin memory. Con- gruence is d efined as the extenttowhich a brand association sharescontentand meaning with another brand association. The congruence of brand associa- tionsshould affect (1) how easily an existing associ- ation can be recalled and (2) how easily ad d itional associationscan become linked tothe brand nod e in memory. In general, information thatisconsistentin meaning with existing brand associationsshould be more easily learned and remembered than unrelated information-though the unexpected ness of infor- mation inconsistentin meaning with the brand some- timescan lead tomore elaborate processing and stronger associationsthan even consistentinformation (Houston, Child ers, and Heckler 1987; Myers-Levy and Tybout 1989; Wyer and Srull 1989). That is, consumers may have expectations astothe likelihood thata prod uct or service hasa particular association given thatithas some other association (Bettman,John, and Scott 1986; Sujan 1985). These expectations should affectcon- sumers' ability tolearn new brand information. For example, if a running shoe hasa brand association with "very d urable and long-lasting," presumably it would be easier toestablish an association with "all weather" than with "stylish." Asnoted subsequently, these expectations also may resultin the formation of inferred brand associations. Thus, the strength of an association should d epend on how itscontentrelates tothe contentof other associationsfor the brand . The congruence among brand associationsd eter- minesthe "cohesiveness" of the brand image-that is, the extenttowhich the brand image ischaracter- ized by associationsor subsetsof associationsthatshare FIGURE 1 Dimensions of Brand Knowled ge Customer-Based Brand Equity / 7 This content downloaded from 152.3.102.242 on Mon, 30 Sep 2013 12:03:07 PM All use subject to JSTOR Terms and Conditions meaning. The cohesivenessof the brand image may d etermine consumers' more holisticor gestalt reac- tionstothe brand . Moreover,a "d iffuse" brand im- age, where there islittle congruence among brand as- sociationsfor consumers, can present several potential problems for marketers. First, consumers may be con- fused astothe meaning of the brand and ,because they d onothave asmuch information towhich new information can be easily related , new associations may be weaker and possibly lessfavorable (Heckler,Keller, and Houston 1992). Moreover, because any one as- sociation shareslittle meaning with other associa- tions, brand associations may be more easily changed by competitive actions. Finally, another problem with a d iffuse brand image is the greater likelihood that consumerswill d iscountor overlook some potentially relevantbrand associationsin making brand d eci- sions. For example, research on "part-listcuing ef- fects" hasshown thatrecall of information can inhibit and lower the recall of other information frommem- ory (Alba and Chattopad hyay 1985a,b,1986; Hoch 1984; Keller 1991a). Hence,only some of the poten- tially retrievable brand associations actually may be recalled when the brand image is notcohesive and consistent. Customer-Based Brand Equity As noted , brand equity hasbeen d efined in a variety of ways,d epend ing on the particular purpose. Be- cause the goal of thisarticle is tofacilitate the d e- velopment of more effective marketingstrategies and tactics, the focusison brand effectson the ind ivid ual consumer. The ad vantage of conceptualizing brand equity fromthis perspective is thatitenablesman- agers toconsid er specifically how their marketing programimproves the value of their brand s. Though the eventual goal of any marketingprogram istoin- crease sales, itisfirst necessary toestablish knowl- ed ge structuresfor the brand sothatconsumersre- spond favorably to marketingactivity for the brand . The preced ing section provid es a d etailed framework of brand knowled ge. In this section, thatframework isused toconsid er in more d etail how knowled ge af- fectsconsumer response tothe marketing of a brand by d efining customer-based brand equity and exam- ining how itis built,measured , and managed . Defining Customer-Based Brand Equity Customer-based brand equity is d efined as the d iffer- ential effectof brand knowled ge on consumer re- sponse tothe marketingof the brand . Three impor- tant concepts are includ ed in the d efinition: "d ifferential effect," "brand knowled ge," and "consumer response to marketing." Differential effect is d etermined by comparing consumer response tothe marketing of a brand with the response tothe same marketingof a fictitiously named or unnamed version of the prod uct or service. Brand knowled ge is d efined in termsof brand awarenessand brand image and is conceptual- ized accord ing tothe characteristicsand relationships of brand associationsd escribed previously. Consumer response to marketing isd efined in termsof consumer perceptions,preferences, and behavior arising from marketing mix activity (e.g., brand choice,compre- hension of copy points froman ad , reactionstoa cou- pon promotion, or evaluationsof a proposed brand extension). Thus,accord ing tothis d efinition, a brand issaid tohave positive (negative) customer-based brand eq- uity if consumers reactmore (less) favorably tothe prod uct,price,promotion, or d istribution of the brand than they d otothe same marketing mix elementwhen itisattributed toa fictitiously named or unnamed ver- sion of the prod uct or service. Favorable consumer response and positive customer-based brand equity, in turn, can lead toenhanced revenue, lower costs, and greater profits. Brand knowled ge iscentral tothisd ef- inition. In particular, the favorability,strength, and uniqueness of the brand associations play a critical role in d etermining the d ifferential response. If the brand isseen by consumerstobe the same asa prototypical version of the prod uct or service in the category, their response should notd iffer fromtheir response toa hypothetical prod uct or service; if the brand hassome salient, unique associations, those responses should d iffer. The actual nature of how the responses d iffer d epend s on consumers' evaluationsof these associa- tions, aswell asthe particular marketing mix element und er consid eration. Thus,establishing brand aware- nessand a "positive brand image" (i.e., favorable, strong, and unique brand associations) in consumer memory createsd ifferent types of customer-based brand equity,d epend ing on what marketing mix elementis und er consid eration. A brief d iscussion highlighting some relevantconsid erationsfor each of these ele- mentsfollows. Fund amentally, high levelsof brand awarenessand a positive brand image should increase the probability of brand choice,aswell as prod uce greater consumer (and retailer) loyalty and d ecrease vulnerability to competitive marketing actions. Thus, the view of brand loyalty ad opted here isthatitoccurswhen favorable beliefsand attitud esfor the brand are manifested in repeatbuying behavior. Some of these beliefs may reflectthe objective reality of the prod uct, in which case no und erlying customer-based brand equity may be present, butin other cases they may reflectfavor- able, strong, and unique associationsthat gobeyond the objective reality of the prod uct(Park 1991). High levelsof brand awarenessand a positive brand image alsohave specificimplications for the pricing, 8 / Journal of Marketing, January 1993 This content downloaded from 152.3.102.242 on Mon, 30 Sep 2013 12:03:07 PM All use subject to JSTOR Terms and Conditions d istribution, and promotion activitiesrelated tothe brand . First, a positive image should enable the brand tocommand larger margins and have more inelastic responses to price increases. The most important as- pect of the brand image thataffectsconsumer re- sponses to prices is probably overall brand attitud e. Consumerswith a strong, favorable brand attitud e should be more willing to pay premiumprices for the brand (Starr and Rubinson 1978). Similarly, a posi- tive image should resultin increased consumer search (Simonson,Huber, and Payne 1988) and a willing- nesstoseek outd istribution channelsfor the prod uct or service. Finally,high levelsof brand awarenessand a positive brand image can increase marketing com- munication effectiveness. All aspects of the brand im- age are relevantin d etermining consumer response to ad vertising and promotion. For example, several au- thorsnote that ad vertisingresponse and d ecay patterns are a function of consumers' attitud esand behavior toward the brand (Ray 1982; Rossiter and Percy 1987). They maintain thatconsumerswhoare positively pre- d isposed toward a brand may require fewer ad ex- posures tomeetcommunication objectives. Similarly, one could argue that strong attribute or benefitasso- ciationsfor the brand require lessreinforcement through marketing communications. In these d ifferent ways, customer-based brand eq- uity is enhanced by creating favorable response to pricing,d istribution, ad vertising, and promotion ac- tivity for the brand . Moreover, a familiar brand with a positive brand image can also yield licensingop- portunities(i.e., the brand name is used by another firmon one of its prod ucts) and support brand exten- sions (i.e., a firmusesan existing brand name toin- trod uce a new prod uct or service), two important growth strategies for firmsin recent years. Licensing can be a valuable source of royalty income, asevid enced by the substantial merchand ising effortsin recent years, and typically hasbeen employed when brand associ- ationshave strong user imagery or brand personality attributes. A more substantial investmentand risk pro- file for the company,however, is required with brand extensions. Because of their potentially lasting effects on consumer knowled ge and the effectivenessof fu- ture marketingactivity, brand extensionsare consid - ered in more d etail in the section on managing customer-based brand equity. Build ing Customer-Based Brand Equity Build ing customer-based brand equity requires the creation of a familiar brand thathas favorable, strong, and unique brand associations. Thiscan be d one both through the initial choice of the brand id entities, such asthe brand name, logo, or symbol, and through the integration of the brand id entitiesintothe supporting marketingprogram. Choosing brand id entities. Tosee how the initial choice of the brand id entitiescan affectbrand equity, consid er the choice of a brand name. A variety of cri- teria have been suggested for the selection of a brand name (e.g., Aaker 1991; Kotler 1991; Robertson 1989). They generally can be classified accord ing towhether they help enhance brand awarenessor facilitate the linkage of brand associations. Alba and Hutchinson (1987) give an extensive d is- cussion of psychological principles thatcan be useful in und erstand ing how the choice of a name affectsbrand recall and recognition processes. Some criteria often noted by other researchersare thatbrand namesshould be simple,familiar, and d istinctive,along the follow- ing lines. Toenhance the likelihood of successful pro- cessing at encod ing, the brand name should be easy to comprehend ,pronounce, and spell. In fact,market researcherssometimesevaluate the "flicker percep- tion" of brand names (i.e., how quickly a brand name can be perceived and und erstood when exposed only for an instant) toassessconsumer learning of cand i- d ate brand names (Dolan 1985). To improve con- sumer learning of the brand , mnemonicfactors (e.g., One-A-Day) and vivid word sare often employed that have rich evaluative or experiential imagery (Robertson 1987; butsee Myers-Levy 1989). Similarly, the use of a familiar word should be ad vantageous because much information is present in memory towhich the name relates. Finally, a d istinctive word isoften sought toattractattention and red uce confusion among com- peting brand s. These d ifferentchoice criteria for a brand name are not necessarily mutually compatible, and it may be d ifficulttochoose namesthatare simple,familiar, and d istinctive. Moreover, factors affecting the ease with which a brand name isrecalled d iffer fromfac- tors affecting the ease with which a brand name is recognized . For example,past research suggests that high frequency word s (accord ing toconventional use in language) are easier torecall than low frequency word s, butlow frequency word s may be easier torec- ognize than high frequency word s (Gregg 1976; Lynch and Srull 1982). This find ingsuggests that choosing a familiar word representing a well-known concept or some other common object or property asa brand name may facilitate brand recall, butthat choosing a more unusual or d istinctive word may facilitate brand rec- ognition. Decid ing whether to emphasize recall or recognition properties in choosing a brand name d e- pend s on managerial prioritiesconcerning the extent of consumers' in-store processing for the prod uct, the nature of the competitive environment, and soon. The choice of a brand name may alsoaffectthe favorability,strength, and uniqueness of brand asso- ciations. The suggestivenessor meaningfulness of the brand name should affecthow easily brand associa- Customer-Based Brand Equity / 9 This content downloaded from 152.3.102.242 on Mon, 30 Sep 2013 12:03:07 PM All use subject to JSTOR Terms and Conditions tionsare created . The brand name can be chosen to suggestsemantically (1) the prod uct or service cate- gory or (2) important attributesor benefitswithin that category. The firstconsid eration should enhance brand name awarenessand the id entification with the prod - uct category. The second consid eration afford stwo important benefits. First, even in the absence of any marketingactivity, the semantic meaning of a sugges- tive brand name may enable consumerstoinfer cer- tain attributesand benefits. For example, consumers could assume on the basisof the namesalone that Daybreak cereal is wholesome and natural,Chief laund ry d etergent removes tough stains, and Diamond toothpaste whitensand brightens teeth. Second ,a suggestive brand name may facilitate marketing ac- tivity d esigned tolink certain associationstothe brand . Id eally, the brand name can be effectively supported through marketing communicationsand a d istinctive slogan thatties together the brand name and its po- sitioning. Similar choice criteria apply tothe other brand id entities, such the brand logo or symbol. Moreover, another importantobjective istochoose the various brand id entitiestobe mutually reinforcing sothat they interact positively to satisfy these criteria. Neverthe- less,although the jud icious choice of brand id entities can contribute significantly tocustomer-based brand equity, the primary input comesfrom supporting mar- keting activitiesfor the brand and the various prod uct, price,ad vertising,promotion, and d istribution d eci- sions, asd iscussed next. Developing supporting marketingprograms. Marketingprograms are d esigned toenhance brand awarenessand establish favorable,strong, and unique brand associationsin memory sothatconsumers pur- chase the prod uct or service. Brand awarenessisre- lated tobrand familiarity. Alba and Hutchinson (1987) d efine brand familiarity asthe number of prod uct- related experiences thathave been accumulated by the consumer (through prod uctusage,ad vertising,etc.). Greater brand familiarity,through repeated exposures toa brand , should lead toincreased consumer ability to recognize and recall the brand . Thus, the appro- priate marketingstrategy toincrease brand awareness and familiarity isclear fromthe d efinition-anything thatcausesthe consumer to "experience" or be ex- posed tothe brand hasthe potential toincrease fa- miliarity and awareness. Frequent and prominent mentionsin ad vertising and promotion vehiclescan intrusively increase consumer exposure tothe brand , ascan eventor sportssponsorship,publicity, and other activities. Favorable, strong, and unique associationscan be created by the marketingprogram in a variety of well- established ways thatare only highlighted here. The prod uct or service specifications themselvesare the primary basisfor the prod uct-related attribute asso- ciationsand d etermine a consumer'sfund amental un- d erstand ing of whatthe prod uct or service means. Similarly, the pricingpolicy for the brand d irectly cre- atesassociationstothe relevant price tier or level for the brand in the prod uctcategory, aswell asitscor- respond ingprice volatility or variance (e.g., in terms of the frequency and magnitud e of d iscounts). The marketing communication efforts by the firm, in contrast, afford a flexible meansof shaping con- sumer perceptions of the prod uct or service. Attimes, marketers may have totranslate attributesintotheir correspond ing benefitsfor consumers through ad ver- tising or other formsof communication. Marketing communicationsalso may be helpful in creating user and usage imagery attributes. The strength of brand associationsfromcommunication effects d epend s on how the brand id entitiesare integrated intothe sup- portingmarketingprogram-for example, the posi- tion and prominence of the brand id entitiesin a tele- vision ad (Keller 1992). Though d elaying brand id entification until the end of a television commercial may increase attention levels d uring commercial ex- posure,resulting in many communication effects being stored in memory (e.g., ad execution and brand claim information, as well as affective and cognitive re- sponses tothat information), it may also prod uce weak linksfromthese effectstothe brand . Finally, word - of-mouth and other social influencesalso play an im- portantrole, especially for user and usage imagery at- tributes. Leveragingsecond ary associations. The d efini- tion of customer-based brand equity d oesnotd istin- guish between the sourcesof brand beliefs (Fishbein and Ajzen 1975)-that is, whether beliefsare created by the marketer or by some other source of influence such asreference groups or publicity. All thatmatters isthe favorability,strength, and uniqueness of brand associations which,combined with brand awareness, can prod uce d ifferential consumer response tothe marketing of a brand . Nevertheless, itisworthwhile toconsid er in greater d epth how belief associations aboutthe attributesand benefitsof the brand arise. One way belief associationsare created ison the basisof d irect experience with the prod uct or service. A second way is by information aboutthe prod uct or service communicated by the company, other com- mercial sources, or word of mouth. Of the two, d irect experience may create stronger associationsin mem- ory given its inherentself-relevance (Hertel 1982). These episod icmemory traces (Tulving1983) may be especially important for user and usage image attri- bute associations. A third importantway thatbelief associationsare created ison the basisof inferences 10 / Journal of Marketing, January 1993 This content downloaded from 152.3.102.242 on Mon, 30 Sep 2013 12:03:07 PM All use subject to JSTOR Terms and Conditions fromsome existing brand associations. Thatis,many associationsare assumed toexistfor the brand be- cause itischaracterized by other associations. The type and strength of inferencing are a function of the cor- relations perceived by consumers among attributesor benefits (Ford and Smith 1987; Huber and McCann 1982). For example, some consumersin certain cat- egoriesmay infer a high level of prod uct or service quality froma high price, aswell asinfer specific attributesor benefitssuch as prestige and social sta- tus. Dick,Chakravarti, and Biehal (1990) refer tothese types of inferencesasbased on "probabilistic consis- tency." They note that"evaluative consistency" in- ferences may also occur, aswhen consumersinfer the favorability of a brand attribute or benefiton the basis of their overall brand attitud e or their evaluation of some other perceived attribute or benefit. Another type of inferred association occurswhen the brand association itself is linked toother infor- mation in memory thatis not d irectly related tothe prod uct or service. Because the brand becomesid en- tified with thisother entity, consumers may infer that the brand sharesassociationswith that entity, thus prod ucing ind irector "second ary" linksfor the brand . These second ary associations may lead toa transfer of global associationssuch asattitud e or cred ibility (e.g., expertise,trustworthiness, and attractiveness) or more specific attributesand benefitsrelated tothe prod uct or service meaning. Second ary associations may arise from primary attribute associationsrelated to (1) the company,(2) the country of origin,(3) the d istribution channels,(4) a celebrity spokesperson or end orsor of the prod uct or service, or (5) an event. The firstthree types of second ary associationsin- volve "factual sources" for the brand (i.e., whomakes it,where itis mad e, and where itis purchased ). This information isalmost alwayspotentially available to consumers, butits strength of association with the brand d epend s on the emphasis itreceives. First, the brand may vary by the extenttowhich itisid entified with a particular company. Establishing a connection with a company may cause existing associationsfor that company tobecome second ary associationsfor the brand (e.g., perceptions of company reputation and cred ibility). The brand ingstrategy ad opted by the company making the prod uct or provid ing the service isthe most important factor affecting the strength of the company's association with the brand . Three main brand ingstrategies are possible (Kotler 1991). First, companiesmay choose ind ivid ual brand namesfor d ifferent prod ucts and serviceswithout any explicit mention of the company (e.g., Procter & Gamble with Tid e,Bold ,Dash,Cheer,Gain, Oxyd ol, and Duz laund ry d etergents). Second , companiesmay choose their name for all of their prod ucts or services (e.g., General Electricand Heinz). Third , companiesmay choose a hybrid or sub-brand strategy whereby they combine their company name with ind ivid ual brand names (e.g.,Kellogg's Corn Flakesand Courtyard by Marriott). The latter two types of brand ingstrategies should facilitate accesstoconsumers' overall attitud es toward the company. The sub-brand strategy offersan ad d itional potential benefitin thatitcan allow for the creation of more specific brand beliefs. Similarly, a brand may be associated with its "country of origin" (i.e., the country in which the company makesthe prod uct or provid es the service) in such a way thatconsumersinfer specific beliefsand evaluations (Erickson,Johansson, and Chao1984; Hong and Wyer 1989,1990). For example, French wines, German automobiles, and Japanese electronics probably all benefitfromsuch inferences. Finally, the d istribution channelsfor a prod uctmay alsocreate second ary associations. Consumerscan form"brand " images of retailers (Jacoby and Mazursky 1984) on the basisof their prod uctassortment, pricing and cred it policy,quality of service, and soon. These store im- ages have associationsthat may be linked tothe prod - ucts they sell (e.g.,prestige and exclusivity vs. bargain- d riven and mass appeal). Similar types of imagesmay be formed for catalogs and other formsof d irectmar- keting. The final two types of second ary associationsoc- cur when the primary brand associationsare for user and usage situation attributes,especially when they are for a particular person or event. Consid er the case in which ad vertising createsan association between a brand and a celebrity end orser (Rossiter and Percy 1987). Asa result, other associationsfor the celebrity may become related tothe brand . Id eally, one such association would be a favorable attitud e toward the celebrity-for example, a well-known person could lend cred ibility to prod uct or service claimsbecause of hisor her expertise,trustworthiness, or attractive- ness. Ad d itionally, more specific beliefs may be in- volved (Kahle and Homer 1985; McCracken 1989). Thus,consumershave images of celebrity end orsors in their mind sasa resultof observing the celebrities in their own field of end eavor or asa resultof med ia coverage. A celebrity invariably hassome personality attribute associations, aswell as possibly some prod uct- related attribute associations, that may become linked tothe brand . Similarly, a brand may alsobecome as- sociated with a particular event. Again, thatevent may be characterized by a setof attribute and attitud e as- sociationsin memory. When the brand becomeslinked with the event, some of these associationswith the event may become ind irectly associated with the brand . Finally, asnoted previously, id entification with the prod uctcategory itself can alsoresultin inferences prod ucingsecond ary associations. Second ary brand associations may be important if Customer-Based Brand Equity / 11 This content downloaded from 152.3.102.242 on Mon, 30 Sep 2013 12:03:07 PM All use subject to JSTOR Terms and Conditions existing brand associationsare d eficientin some way. In other word s,second ary associationscan be lever- aged tocreate favorable,strong, and unique associ- ationsthatotherwise may notbe present. Choosing to emphasize the company or a particular person,place, or eventshould be based on consumers' awarenessof that entity, aswell ashow the beliefsand attitud es aboutthe entity can become linked tothe brand (see chapter 11 of Rossiter and Percy 1987 for an excellent d iscussion). Such a strategy makessense if consumers alread y have associationsfor the company,person, place, or eventthatare congruent with d esired brand associations. For example, consid er a country such as New Zealand , which isknown for having more sheep than people. A New Zealand sweater manufacturer that promotes its prod uct on the basisof itsNew Zealand wool presumably could more easily establish strong and favorable brand associationsbecause New Zea- land may alread y mean "wool" to many people. Sec- ond ary brand associations may be risky,however,be- cause some control of the brand image is given up. The company,person,place, or eventthatmakes up the primary brand association will und oubted ly have a hostof associationsof which only some smaller set will be of interesttothe marketer. Managing the transfer process sothat only the relevant second ary associa- tionsbecome linked tothe brand may be d ifficult. Moreover, these imagesmay change over time ascon- sumerslearn more aboutthe entity, and new associ- ations may or may notbe ad vantageous for the brand . Measuring Customer-Based Brand Equity There are twobasic approaches to measuring customer- based brand equity. The "ind irect" approach attempts toassess potential sourcesof customer-based brand equity by measuring brand knowled ge (i.e., brand awarenessand brand image). The "d irect" approach attempts tomeasure customer-based brand equity more d irectly by assessing the impact of brand knowled ge on consumer response tod ifferentelementsof the firm's marketingprogram. The ind irectand d irect ap- proaches to measuring customer-based brand equity are complementary and should be used together. The ind irect approach isuseful in id entifying what aspects of brand knowled ge cause the d ifferential response that createscustomer-based brand equity; the d irect ap- proach isuseful in d etermining the nature of the d if- ferential response. Though d etailed d escriptions and critiques of the many specifictechniques behind these two approaches are beyond the scope of thisarticle (see Aaker 1991 for ad d itional d iscussion), itis worthwhile to highlight them briefly. Ind irect approach. The first approach tomeasur- ing customer-based brand equity,measuring brand knowled ge,requiresmeasuring brand awarenessand the characteristicsand relationshipsamong brand as- sociations. Because any one measure typically cap- tures only a particular aspect of brand knowled ge, multiple measuresmustbe employed to capture the multid imensional nature of brand knowled ge. Brand awarenesscan be assessed effectively through a variety of aid ed and unaid ed memory measures(see Srull 1984 for a review) thatcan be applied totest brand recall and recognition. For example, brand rec- ognition measures may use the actual brand name or some perceptually d egrad ed version of the brand name (Alba and Hutchinson 1987). Brand recall measures may use d ifferentsetsof cues, such as progressively narrowly d efined prod uctcategory labels. Besid es correctness, the ease of recall and recognition perfor- mance can be assessed with more subtle measuressuch as response latenciesto provid e a fuller picture of memory performance with respect tothe brand (Fazio 1987). Brand recall can alsobe cod ed in termsof the ord er of recall to capture the extenttowhich the name is "top of mind " and thus strongly associated with the prod uctcategory in memory. There are many ways tomeasure the characteris- ticsof brand associations (i.e., their type, favorabil- ity, and strength). Qualitative techniques can be em- ployed to suggestpossible associations. For example, free association taskscan be used whereby consumers d escribe whatthe brand meanstothemin an unstruc- tured format, either ind ivid ually or in small groups. Specifically, consumers might be probed in termsof "who,what,when,where,why, and how" types of questions aboutthe brand . Projective techniques(Levy 1978,1981,1985) such assentence completion,pic- ture interpretation, and brand personality d escriptors may alsobe useful, especially if consumersare un- willing or otherwise unable to express their feelings. These ind irect measures,however,may notad e- quately capture the favorability or strength of asso- ciations, and more d irectmeasuresoften are necessary to provid e ad d itional information. For example,Ajzen and Fishbein (1980) give a d etailed d escription of how beliefsand evaluationsof attributesand benefitscan be scaled and how attitud escan be measured through a structured format, provid ing an illustrative example in a consumer setting. Asnoted previously,response time measuresof attitud eshave been used asa proxy for attitud e strength. Relationshipsamong brand associationscan be measured by two general approaches: (1) comparing the characteristicsof brand associationsin some way and (2) d irectly asking consumersfor information rel- evanttothe congruence,competitive overlap, or le- verage for the brand associations. Congruence is the extenttowhich brand associationsare shared . Con- gruence can be assessed by comparing the pattern of associationsacrossconsumerstod etermine which as- 12 / Journal of Marketing, January 1993 This content downloaded from 152.3.102.242 on Mon, 30 Sep 2013 12:03:07 PM All use subject to JSTOR Terms and Conditions sociationsare common or d istinctive. Ad d itionally, consumerscould be asked d irectly their cond itional expectations for attribute,benefit, or attitud e associ- ations (i.e., the likelihood thata prod uct or service hasone association given thatithas another). Competitive overlap of brand associations is the extenttowhich brand associationsare linked tothe prod uctcategory (i.e., id entification) and are or are notshared with other brand s (i.e.,uniqueness). Id en- tification can be assessed by examining how con- sumers respond tobrand recall taskswith prod uct cat- egory or some other type of cues. Uniqueness of brand associationscan be assessed by comparing the char- acteristicsof associationsof the focal brand (i.e.,their type,favorability, and strength) with the character- isticsof associationsfor competing brand s. Ad d ition- ally, consumerscould be asked d irectly (1) how strongly they id entify the brand with the prod uctcategory and (2) what they consid er tobe the unique and shared aspects of the brand . Multivariate techniques such as multid imensional scaling alsocan be employed (Aaker and Day 1986). Leverage is the extenttowhich other brand as- sociationslinked toa brand association become sec- ond ary associationsfor the brand . Leverage can be assessed by comparing the characteristicsfor the par- ticular company,person,place,event, or prod uct cat- egory with those characteristicsfor the focal brand ac- cord ing totheir type, favorability, and strength. Ad d itionally, consumerscould be asked d irectly what inferencesare mad e aboutthe brand on the basisof knowled ge of the particular person,place, event, company, or prod uctcategory. Direct approach. The second approach tomea- suring customer-based brand equity,d irectly measur- ing the effectsof brand knowled ge on consumer re- sponse to marketing for the brand ,requiresexperiments in which one group of consumers respond s toan ele- mentof the marketingprogram when itisattributed tothe brand and another group of consumers respond s tothatsame elementwhen itisattributed toa ficti- tiously named or unnamed version of the prod uct or service. By attributing the marketing elementtoan unfamiliar or anonymousprod uct, consumersshould interpret itwith respect totheir general knowled ge about the prod uct or service, aswell as prototypical prod uct or service specifications and price,promotion, and d istribution strategies. Comparing the responses of the two groups thus provid es an estimate of the effects d ue tothe specificknowled ge aboutthe brand that goesbeyond basic prod uct or service knowled ge. The classic example of this approach is the so- called "blind " testin which consumersevaluate a prod uct on the basisof a d escription,examination, or actual consumption experience, either with or without brand attribution. Pastresearch of this type hasshown that knowled ge of the brand affectsconsumer percep- tions,preferences, and choicesfor a prod uct(e.g., Allison and Uhl 1964; Jacoby,Olson,and Had d ock 1971). Blind testscould be used toexamine consumer response toother elementsof the marketing mix such as proposed pricing,promotion, and channelsof d is- tribution changes. One important consid eration with the d irect ap- proach isthe experimental realismthatcan be achieved when some aspect of the marketingprogram isattrib- uted toa fictitiously named or unnamed version of the prod uct or service. Detailed concept statementscan be employed in some situationswhen it may be other- wise d ifficultfor consumerstoexamine or experience the marketing mix elementwithout being aware of the brand . Thus,concept statements may be useful in as- sessing customer-based brand equity when consumers make a prod uct choice or evaluate a change in the prod uct or service composition,jud ge a proposed brand extension, or respond toa proposed price or d istri- bution change. Assessing customer-based brand eq- uity with marketing communications presents a bigger challenge with the d irect approach (e.g., consumer re- sponse toa proposed new ad vertisingcampaign). In thiscase, storyboard s and animaticor photomatic ver- sionsof an ad could be used rather than a finished ad toallow for the necessary d isguise of the brand . Though this approach should work well with "informational" ad s, it probably would be less appropriate for "trans- formational" ad s emphasizinguser,usage, or some other type of imagery, in which prod uction valuesare a critical ingred ient in achieving communication goals (Rossiter and Percy 1987). Finally, another potentially useful approach for d i- rectly assessing customer-based brand equity iscon- joint or trad eoff analysis(Green and Srinivasan 1978, 1990; Green and Wind 1975). Conjointanalysis can be used to explore the main effectsof the brand name (i.e.,d ifferencesin preference or choice for the brand ) and interaction effectsbetween the brand name and other marketing mix elementssuch as price,prod uct or service features, and promotion or channel choices (i.e., d ifferencesin perceptions for the brand ). For example,Rangaswamy,Burke, and Oliva (1990) use conjointanalysis to explore how brand namesinteract with physical prod uct featurestoaffectthe extend a- bility of brand namestonew prod uctcategories. Note thatif conjointanalysis is employed , care mustbe taken thatconsumersd onotevaluate unrealistic prod uct profiles or scenariosthatviolate their basic expecta- tionsfor the prod uct or brand (Park 1991; Srinivasan 1979). Table 1 summarizesthe d ifferentmeasuremental- ternativesfor customer-based brand equity. Customer-Based Brand Equity/ 13 This content downloaded from 152.3.102.242 on Mon, 30 Sep 2013 12:03:07 PM All use subject to JSTOR Terms and Conditions TABLE 1 Measurement of Brand Knowled ge Constructs Related toCustomer-Based Brand Equitya Construct Measure(s) Purpose of Measure(s) Brand Awareness Recall Correctid entification of brand given Capture "top-of-mind " accessibility of prod uctcategory or some other type of brand in memory probe ascue Recognition Correctd iscrimination of brand as having Capture potential retrievability or been previously seen or heard availability of brand in memory Brand Image Characteristicsof brand associations Type Free association tasks,projective Provid e insight intonature of brand techniques,d epth interviews associations Favorability Ratings of evaluationsof associations Assess key d imension prod ucing d ifferential consumer response Strength Ratings of beliefsof association Assess key d imension prod ucing d ifferential consumer response Relationshipsamong brand associations Uniqueness Compare characteristicsof associations Provid e insightintothe extenttowhich with those of competitors(ind irect brand associations are notshared with measure) other brand s; assess key d imension Ask consumers what they consid er tobe prod ucing d ifferential consumer response the unique aspects of the brand (d irect measure) Congruence Compare patterns of associations across Provid e insight intothe extenttowhich consumers (ind irectmeasure) brand associations are shared ,affecting Ask consumers cond itional expectations their favorability,strength, or uniqueness aboutassociations (d irectmeasure) Leverage Compare characteristicsof second ary Provid e insight intothe extenttowhich associations with those for a primary brand associations toa particular person, brand association (ind irectmeasure) place,event,company,prod uct class,etc. Ask consumers d irectly whatinferences are linked toother associations, they would make aboutthe brand based prod ucingsecond ary associations for the on the primary brand association (d irect brand measure) 'Thistable d escribes the ind irect approach of assessingpotential sources of customer-based brand equity by measuring brand knowled ge. The d irect approach to measuring customer-based brand equity involves measuring the effects of brand knowled ge on consumer response to marketing-for example,by cond uctingexperiments in which one group of consumers respond toan elementof the marketing mix when itisattributed tothe brand ,and another group of consumers respond tothe same marketing mix elementwhen itisattributed toa fictitiously named or unnamed version of the prod uct or service. Managing Customer-Based Brand Equity Accord ing tothe d efinition of customer-based brand equity, no single number or measure captures brand equity. Rather, brand equity should be thought of as a multid imensional concept that d epend s on (1) what knowled ge structuresare present in the mind sof con- sumersand (2) whatactionsa firmcan take to capi- talize on the potential offered by these knowled ge structures. Differentfirms may be more or lessable tomaximize the potential value of brand accord ing to the type and nature of marketing activitiesthat they are able tound ertake. Nevertheless, six general guid e- linesbased on the preced ingconceptual framework are presented here to help marketersbetter manage customer-based brand equity. First, marketersshould ad opt a broad view of mar- keting d ecisions. Marketingactivity for a brand po- tentially can create value for the brand by improving consumers' ability torecall or recognize the brand and / or by creating, maintaining, or changing the favora- bility,strength, or uniqueness of various types of brand associations. By influencing brand knowled ge in one or more of these d ifferent ways, marketingactivity can potentially affectsales. Second , marketers should d efine the knowled ge structuresthat they would like tocreate in the mind s of consumers-that is,by specifying d esired levels of awareness and favorability, strength, and uniqueness of prod uct- and non-prod uct-related attributes; func- tional,experiential, and symbolicbenefits; and over- all attitud es. In particular, marketersshould d ecid e on the core need s and wantsof consumers tobe satisfied by the brand . Marketersshould alsod ecid e the extent towhich itis necessary to leverage second ary asso- ciations for the brand -that is, link the brand tothe 14 / Journal of Marketing, January 1993 This content downloaded from 152.3.102.242 on Mon, 30 Sep 2013 12:03:07 PM All use subject to JSTOR Terms and Conditions company,prod uct class, or particular person,place, or eventin such a way thatassociationswith those entitiesbecome ind irector "second ary" associations for the brand . Third , marketersshould evaluate the increasingly large number of tactical options available tocreate these knowled ge structures,especially in termsof various marketing communication alternatives. For example, the recent growth of "nontrad itional" med ia,promo- tions, and other marketingactivity (e.g., sports and event sponsorship; in-store ad vertising; "minibill- board s" in transit vehicles, on parkingmeters, and in other locations; and prod uctplacement in moviesand television shows) is appropriate fromthe perspective of customer-based brand equity. Asnoted previously, the manner in which a brand association is created d oes not matter-only the resultingfavorability, strength, and uniqueness. Thus,many of these new alternativescan offer a cost-effective meansof af- fecting brand knowled ge and thus sales, especially to the extentthat they complement more trad itional mar- keting tactics. Regard less of which options are cho- sen, the entire marketingprogram should be coord i- nated tocreate congruent and strong brand associations. Different marketing tacticswith the same strategic goals, if effectively integrated , can create multiple links tocore benefitsor other key associations, helping to prod uce a consistentand cohesive brand image. Mar- ketersshould jud ge the consistency and cohesiveness of the brand image with the businessd efinition in mind (Levitt1960) and how well the specific attributesand benefitsthatthe prod uct or service isintend ed to pro- vid e toconsumers satisfy their core need sand wants (Kotler 1991; Park,Jaworski, and Maclnnis 1986). Fourth, marketersshould take a long-term view of marketing d ecisions. The changes in consumer knowled ge aboutthe brand fromcurrent marketing activity alsowill have an ind irecteffecton the success of future marketing activities. Thus, fromthe per- spective of customer-based brand equity in making marketingd ecisions, itis important toconsid er how resultingchanges in brand awarenessand image may help or hurt subsequentmarketing d ecisions. For ex- ample, the use of sales promotionsinvolvingtempo- rary price d ecreases may create or strengthen a "d is- count" association with the brand , with implications for customer loyalty and responses tofuture price changes or non-price-oriented marketing communi- cation efforts. Fifth,marketersshould employ tracking stud iesto measure consumer knowled ge structuresover time to (1) d etect any changes in the d ifferentd imensionsof brand knowled ge and (2) suggest how these changes might be related tothe effectivenessof d ifferentmar- keting mix actions. Tothe extentthata more precise assessmentof customer-based brand equity is useful, marketersshould alsocond uctcontrolled experi- ments. Consumer knowled ge of competitive brand s should be similarly tracked to provid e information on their sourcesof customer-based brand equity. Exper- imentswith consumer response to marketingactivity for competitive brand scan also provid e a useful benchmark-for example, tod etermine the unique- nessof brand associations. Finally, marketersshould evaluate potential ex- tension cand id atesfor their viability and possible feed back effectson core brand image. Given their po- tential importance to long-term brand value,brand ex- tension d ecisionsare consid ered in d etail in the rest of thissection fromthe perspective of customer-based brand equity and other relevantresearch. Brand extensions capitalize on the brand image for the core prod uct or service to efficiently informcon- sumersand retailersaboutthe new prod uct or service. Brand extensionscan facilitate acceptance of the new prod uct or service by provid ing twobenefits. First, awarenessfor the extension may be higher because the brand nod e is alread y present in memory. Thus, consumersshould need only toestablish a connection in memory between the existing brand nod e and the new prod uct or service extension. Second , inferred associationsfor the attributes,benefits, and overall perceived quality may be created . In other word s, consumers may form expectations for the extension on the basisof what they alread y know aboutthe core brand . These inferencescan lower the costof the in- trod uctory campaign for the extension-for example, by increasingad vertisingefficiency (Smith and Park 1992). Keller and Aaker (1992) review relevantliterature to provid e a conceptual mod el of how consumersuse their knowled ge toevaluate a brand extension. They maintain thatextension evaluationswill d epend on the salience of the core brand associationsin the exten- sion context, how relevantconsumers perceive thisin- formation tobe totheir extension evaluations, and how favorable inferred associationsare in the extension context. In other word s, extension evaluationswill d epend on whatkind of information comestomind aboutthe core brand in the extension context, whether thisinformation isseen as suggestive of the type of prod uct or service thatthe brand extension would be, and whether thisinformation isviewed as good or bad in the extension contextin comparison with compet- itors. The salience or accessibility of the core brand as- sociations d epend s on their strength in memory, as well asthe retrieval cues provid ed by the extension context. Some associations may be salientwhen con- sumersevaluate some extensionsbutnotothers. The relevance of the salientcore brand associationsd e- pend s, in part, on their perceived similarity tothe pro- Customer-Based Brand Equity/ 15 This content downloaded from 152.3.102.242 on Mon, 30 Sep 2013 12:03:07 PM All use subject to JSTOR Terms and Conditions posed extension prod uct or service (Feld man and Lynch 1988). When overall similarity is high, consumersare more likely tobase their extension evaluationson their attitud e toward the core brand (Boush and Loken 1991; Boush etal. 1987; Herr,Farquhar, and Fazio 1990). Overall similarity jud gments could be mad e in d iffer- ent ways(Loken and Ward 1990),though researchers typically assume that they are a function of salient shared associationsbetween the core brand and the extension prod uctcategory. These similarity jud g- mentscould be based on prod uct-related attributes,as well as non-prod uct-related attributessuch asuser type or usage situation (Brid ges1990; Park, Milberg, and Lawson 1991). When overall similarity isnot very high, consumersare more likely toconsid er specific attri- butesand benefitsinvolved . If relevant, the favora- bility of inferred attribute and benefitbeliefswill d e- pend on how they are valued in the extension context. Though these evaluationswill generally correspond to the favorability of the core brand associations, they can d iffer, and in factbe negative, even if the core brand associationsthemselvesare positive (Aaker and Keller 1990). Moreover, even if positive attribute and benefitassociationsfor the core brand lead toinfer- ences of positive brand extension associations, in- ferred negative associations may still emerge (Brid ges 1990). Finally, when overall similarity is very low, consumer evaluationsalsowill be very low. When multiple prod uct or service extensionsare associated with the brand , the congruence among their associationsbecomesan important d eterminantof the consistency and cohesivenessof the brand image. It is often argued thatan extension can help the core brand image by improving the favorability and strength of associationsand clarifying the businessd efinition and core benefitsfor the brand . Aaker (1991) claims thatbrand extensions helped to fortify the brand im- ages of Weight Watchersand Sunkist. Keller and Aaker (1992) found thatthe successful introd uction of a brand extension improved evaluationsof a core brand that originally was perceived tobe of only average quality, although in their research setting consumersd id not have strongly held attitud estoward the core brand and the company ad opted a family brand ingstrategy that raised the salience of itsname (and thus perceptions of its cred ibility). Ithasalsobeen argued thatsuccessful brand ex- tensions may potentially harmthe core brand image if they weaken existing associationsin some way. If a brand becomesassociated with a d isparate setof prod ucts or services,prod uctcategory id entification and the correspond ingprod uct associations may be- come less strong. For example,Pepperid ge Farm, Cad bury, and Scott Paper have been accused of "overextend ing" by introd ucing too d isparate prod - ucts. Dilution effects, with potentially ad verse profit implications,may be especially likely when the ex- isting associationsfor the core brand are alread y fairly weak. For example, the successful introd uction of the Miller Lite beer in the U.S. may have accentuated perceptions of the flagship Miller High Life beer asa "less hearty" beer because that perception had alread y been created in consumers' mind s by itsclear bottle (in contrasttoBud weiser'sd ark bottle). As another example of a potential d ilution effect, successful ex- tensionsfor brand swith an exclusivity and prestige image that effectively broad en the target market may prod uce negative feed back effectson the brand from membersof the original consumer franchise whore- sentthe market expansion. For example, the intro- d uction of the lower priced Cad illacCimaron mod el is thought tohave led tod eclinesin image and sales for the entire Cad illacd ivision (Yovovich 1988). Though these d ifferent types of d ilution effects may occur, multiple prod uct or service extensions may not be asharmful tocertain abstractassociationssuch as brand attitud esand perceived quality. In other word s, although the brand may nothave the same specific prod uct or service meaning because of multiple ex- tensions,consumers may still see the brand as rep- resenting a range of prod ucts or servicesof a certain quality. An unsuccessful brand extension, in contrast, can harmthe core brand image by creating und esirable as- sociations. Such effectsare most likely when there is little d ifference between the original brand and the ex- tension. For example, Sullivan (1990) cond ucted an econometric analysis thatshowed how the perceived "sud d en acceleration" problem of Aud i's5000 mod el "spilled over" and red uced d emand for its4000 and Quattro mod els. Roed d er John and Loken (1990) found that perceptions of quality for a core brand in the health and beauty aid sarea d ecreased with the hypothetical introd uction of a lower quality extension in a similar prod uctcategory (i.e.,shampoo). Quality perceptions of the core brand were unaffected ,however, when the proposed extension wasin a d issimilar prod uct cate- gory (i.e., facial tissue). Similarly, Keller and Aaker (1992) found thatunsuccessful intervening extensions in d issimilar prod uctcategories d id notaffectevalu- ationsof the core brand (also see Romeo 1990). In summary, marketersshould evaluate potential extension cand id atesfor their viability and their feed - back effectson core brand image by (1) id entifying possible extension cand id ateson the basisof core brand associations (especially with respect tobrand posi- tioning and core benefits) and overall similarity of the extension tothe brand ,(2) evaluating extension can- d id ate potential by measuring the salience,relevance, and favorability of core brand associationsin the pro- posed extension contextand the favorability of any inferred associations, and (3) consid ering the exten- 16 / Journal of Marketing, January 1993 This content downloaded from 152.3.102.242 on Mon, 30 Sep 2013 12:03:07 PM All use subject to JSTOR Terms and Conditions sion's potential feed back effectson the core brand im- age and the favorability,strength, and uniqueness of core brand associations. Discussion Summary Thisarticle introd ucesthe concept of customer-based brand equity, d efined asthe d ifferential effectof brand knowled ge on consumer response tothe marketing of the brand . A brand issaid tohave positive (negative) customer-based brand equity if consumersreactmore (less) favorably toan elementof the marketing mix for the brand than -hey d otothe same marketing mix elementwhen itisattributed toa fictitiously named or unnamed version of the prod uct or service. Brand knowled ge is conceptualized accord ing toan associ- ative network memory mod el in termsof twocom- ponents, brand awarenessand brand image (i.e., a set of brand associations). Brand awarenessconsistsof brand recognition and brand recall. Brand associa- tionsare conceptualized in termsof their character- istics by type (level of abstraction and qualitative na- ture),favorability, and strength, and in termsof their relationship with other associations by congruence, competitive overlap(id entification and uniqueness), and leverage. Customer-based brand equity occurswhen the consumer is aware of the brand and hold ssome favorable,strong, and unique brand associationsin memory. The d ifferent types of customer-based brand equity are d iscussed by consid ering the effectsof these d imensionsof brand knowled ge on brand loyalty and consumer response to prod uct,price,promotion, and d istribution strategies. The article also explores some specificaspects of this conceptualization by consid ering how customer- based brand equity is built,measured , and managed . Build ing brand equity requirescreating a familiar brand name and a positive brand image-that is,favorable, strong, and unique brand associations. Strategies to build customer-based brand equity are d iscussed in termsof both the initial choice of the brand id entities (brand name,logo, and symbol) and how the brand id entitiesare supported by and integrated intothe marketingprogram. Twobasic approaches tomea- suring customer-based brand equity are outlined . The ind irect approach measuresbrand knowled ge (brand awarenessand elementsof brand image) toassessthe potential sourcesof brand equity. The d irect approach measuresthe effectsof the brand knowled ge on con- sumer response toelementsof the marketing mix. Ex- amples of both types of approaches are provid ed . Fi- nally, six guid elines for the management of customer- based brand equity are d iscussed . These guid elines emphasize the importance of taking a broad and long- termview of marketing a brand ; specifying the d e- sired consumer knowled ge structuresand core bene- fitsfor a brand ; consid ering a wid e range of trad i- tional and nontrad itional ad vertising,promotion, and other marketingoptions; coord inating the marketing options thatare chosen; cond uctingtracking stud ies and controlled experiments; and evaluatingpotential extension cand id ates. Future Research Directions In the presentation of a conceptual framework of customer-based brand equity, several constructsand relationships are d iscussed . Consequently, ad d itional research is necessary both torefine thisframework and to suggest other implications for marketing strat- egies and tactics. Und oubted ly, much previous re- search may be useful in thiseffort. Because thisre- search was most likely cond ucted with a d ifferent purpose in mind ,however, ad d itional insightsmay be gained by consid ering itfromthe potentially broad er perspective of customer-based brand equity. In clos- ing, some research priorities for build ing,measuring, and managing customer-based brand equity are id en- tified . There are several important research questions about how tobuild customer-based brand equity. First, bet- ter choice criteria should be established for the brand id entities (brand name,logo, and symbol). For ex- ample,remarkably little empirical research has sys- tematically examined brand name consid erationsas they pertain to enhancing brand awarenessand build - ingfavorable, strong, and unique brand associations. Such research should recognize the numeroustrad e- offs in choice criteria by suggesting when certain characteristicsof the brand id entitiesshould be em- phasized . For example,memory retrieval consid er- ationsthatarise fromassociative strength and part-list cueing theoriesin psychology imply thata meaning- ful, "suggestive" brand name may facilitate initial po- sitioning, buta nonsuggestive or neutral brand name may more effectively accommod ate later reposition- ing. Support of this hypothesis would imply thatfirms may be better off ad opting more flexible brand ing strategies,using more neutral brand names, if they anticipate need ing to reposition the brand later. In d e- velopingcontingency-based choice criteria, italsowill be necessary to clarify the rolesof variousbrand id en- tities by consid ering more explicitly how brand names, logos, symbols,slogans, and other trad emarkscan contribute d ifferentially to build ing customer-based brand equity. Thisline of research could consid er vi- sual and verbal properties of these brand id entitiesand how they might affectbrand awarenessand the fa- vorability,strength, and uniqueness of brand associ- ations. In termsof und erstand ing how the supporting mar- Customer-Based Brand Equity / 17 This content downloaded from 152.3.102.242 on Mon, 30 Sep 2013 12:03:07 PM All use subject to JSTOR Terms and Conditions ketingprogram build scustomer-based brand equity, two particular research d irectionscould be pursued . First, factors influencing the favorability,strength, and uniqueness of brand associations,a focusof much past research, should continue tobe explored , but along several d ifferentlines. Are certain types of associa- tions inherently more favorable,stronger, or unique in memory? Which types of associationsare more eas- ily created by a particular marketing or communica- tion mix element? Which types of associationsare more likely toinfluence consumer response with respect to a particular marketing mix element? Finally, whatare the trad eoffsinvolved in creatingfavorable,strong, and unique brand associations? For example, itwas suggested previously thatbenefitscan be more mem- orable than attribute information, butattributes may have tobe communicated to persuad e consumersand create favorable benefitassociations. Itwasalso sug- gested above that non-prod uct-related or image attri- butes, such asuser type or usage situation,may create unique associations, butund er some circumstances they may notbe favorably received or strongly linked to the brand in memory. Second , the costsand benefitsof leveraging sec- ond ary associationsshould be explored . For example, how and und er whatcond itionsshould a firmincrease the salience of source factorsrelated tothe brand (i.e., the company,country of origin, and d istribution chan- nel)? All of these source factorshave their own setof associations. How d oconsumers merge or combine these associationswith other brand associations? In other word s, how d othese source and brand images interact? Another important issue iswhen and how a brand should attempt tobecome associated with a par- ticular person or event. For example, Rossiter and Percy (1987) offer the following criteria for choosing a pre- senter in ad vertising: (1) visibility,(2) cred ibility (ex- pertise and objectivity),(3) attraction (likability and similarity), and (4) power. These criteria could be ad apted toad d resswhen and how a brand should be- come id entified with an event. One important research priority isto d evelop valid benchmarksfor the d irect approach to measuring customer-based brand equity-that is,plausible d e- scriptions of the relevant activity (ad vertising,pro- motion,prod uct,pricing,etc.) with noor fictitious brand id entification. Another useful contribution would be to d esign efficientand effective approaches tocon- d uctingtracking stud ies. Thiswould entail consid er- ing the pros and cons of d ifferent qualitative and quantitative approaches to measuring brand knowl- ed ge of consumers. Several research questions are relevantfor man- aging customer-based brand equity. What strategies are effective in creatingstrong brand associations? How can d ifferent marketing mix elementsbe integrated to create strong and congruent brand associationsand a consistentand cohesive brand image? Thisline of re- search should clearly examine how trad itional and nontrad itional marketingoptions interact. Effective strategies for integratingmarketing communications in termsof ad vertising,promotion,publicity, d irect marketing, and package d esign are especially need ed . For example, how can ad vertising be coord inated across broad castand print med ia toenhance brand awareness and strengthen brand associations? These research stud ies might consid er memory principles and theo- riesof encod ing and retrieval. Also, how should the consistency and cohesive- nessof a brand image be managed over consumer seg- ments (includ inggeographicbound aries) and over time? A d iffuse brand image with weaker and lessfavorable brand associations may be particularly evid entwhen a brand attempts to reposition itself (e.g., switching toa new targetmarket) (Heckler,Keller, and Houston 1992). Are there ways in which a brand image can be "flexible" and appeal tod ifferentconsumer seg- ments? To manage the brand image better over time, more precise guid elines astothe "ind irect" effectsof current marketingactivity on the successof future marketingactivity are need ed -for example, by achieving a better und erstand ing of how brand knowl- ed ge influencesconsumer response. Finally, broad er implications of customer-based brand equity should be explored by consid eringag- gregation issuesassociated with brand knowled ge ef- fectson market segments or the customer franchise as a whole, as opposed toeffectson an ind ivid ual con- sumer. An aggregate analysis alsocould consid er the implication of customer-based brand equity for sales, market share, and profits. Thismore extend ed anal- ysis should consid er aspects of the company (e.g., its strengths and weaknesses) and the competitive nature of itsmarkets. Similarly, it may alsobe useful toin- corporate some of the concepts thatrelate tocustomer- based brand equity toad d ressother management questionspertaining to brand ing-for example, tod e- velop a financially based conceptualization of brand equity. Conclusions The goal of thisarticle is to present a conceptual framework thatwould provid e useful structure for managersd eveloping brand strategies and researchers stud ying brand equity. In particular, the article build s a theoretical found ation based on past research in con- sumer behavior thatshould be helpful in ad d ressing some of the new challenges in d eveloping brand strat- egies thathave arisen because of changes in the mar- keting environment (e.g., fromthe proliferation of brand extensionsand the growth of new,alternative promotional and med ia alternatives). Though many of the id eas expressed in thiscon- 18 / Journal of Marketing, January 1993 This content downloaded from 152.3.102.242 on Mon, 30 Sep 2013 12:03:07 PM All use subject to JSTOR Terms and Conditions ceptual framework may be familiar to managers, its value isin integrating these variousnotions to provid e a more comprehensive picture of how marketerscan create value for a brand . For example, marketers may agree that they should take a broad and long-term view of marketing d ecisionsfor a brand , buthow they should d oso may notbe obvious. By recognizing thatmar- ketingactivity can potentially enhance or maintain consumers' awarenessof the brand or the favorability, strength, and uniqueness of various types of brand as- sociations, the customer-based brand equity frame- work may provid e the perspective thatwill enable marketerstotake better short-termand long-term mar- keting actions. Moreover, this framework may also suggest some consid erations thathave been otherwise overlooked . Thus, thisbroad er contextcan help man- agers can make more insightful and informed brand d ecisions. For researchers, the value of the framework is in suggesting areaswhere managerial guid ance isneed ed butacad emic guid elines are currently lacking. As sug- gested by the large number of suggested future re- search d irections id entified , much work need s tobe d one. REFERENCES Aaker, David A. (1982), "Positioning Your Prod uct," Busi- ness Horizons, 25 (May/June), 56-62. (1991), Managing Brand Equity. New York: The Free Press. and Alexand er Biel, ed s. (1992), Build ingStrong Brand s. Hillsd ale, NJ: Lawrence ErlbaumAssociates. and George S. Day (1986), MarketingResearch, 3rd ed . New York: John Wiley & Sons, Inc. and Kevin Lane Keller (1990), "Consumer Eval- uationsof Brand Extensions," Journal of Marketing, 54 (January), 27-41. Ajzen, Icek and Martin Fishbein (1980), Und erstand ing Atti- tud esand Pred icting Social Behavior. Englewood Cliffs, NJ: Prentice-Hall, Inc. Alba,Joseph W. and Amitava Chattopad hyay (1985a), "The Effectsof Contextand Part-Category Cueson the Recall of CompetingBrand s," Journal of MarketingResearch, 22 (August), 340-9. and (1985b), "The Effects of Part-List Cuing on Attribute Recall: Problem Framing atthe Point of Retrieval," in Ad vancesin Consumer Research,Vol. 12, Elizabeth C. Hirschman and MorrisB. Holbrook,ed s. Provo, UT: Association for Consumer Research,410-13. and (1986), "Salience Effectsin Brand Recall," Journal of MarketingResearch, 23 (November), 363-9. and J. Wesley Hutchinson (1987), "Dimensionsof Consumer Expertise," Journal of Consumer Research, 13 (March),411-53. and John G. Lynch, Jr. (1991), "Memory and Decision Making," in Hand book of Con- sumer Theory and Research, Harold H. Kassarjian and Thomas S. Robertson,ed s. Englewood Cliffs, NJ: Pren- tice-Hall,Inc., 1-49. Allison,Ralph and Kenneth P. Uhl (1964),"Influencesof Beer Brand Id entification on Taste Perception," Journal of Mar- ketingResearch, 2 (August), 36-9. And erson, John R. (1983), The Architecture of Cognition. Cambrid ge, MA: Harvard University Press. Baker,William, J. Wesley Hutchinson, Danny Moore, and Prakash Ned ungad i (1986), "Brand Familiarity and Ad ver- tising: Effectson the Evoked Setand Brand Preferences," in Ad vances in Consumer Research, Vol. 13,Richard J. Lutz,ed . Provo, UT: Association for Consumer Research, 637-42. Barwise,Patrick, Christopher Higson, And rew Likierman,and Paul Marsh (1989),Accountingfor Brand s. Lond on: Lon- d on Business School and the Institute for Chartered Ac- countantsin England and Wales. Berger, Id a E. and And rew A. Mitchell (1989), "The Effect of Ad vertising on Attitud e Accessibility," Journal of Con- sumer Research, 16 (December), 280-8. Bettman, JamesR. (1979), An Information ProcessingTheory of Consumer Choice. Read ing, MA: Ad d ison Wesley Pub- lishingCompany. (1986), "Consumer Psychology," Annual Review of Psychology, 37,257-89. and C. Whan Park (1980), "Effects of Prior Knowled ge and Experience and Phase of the Choice Pro- cess on Consumer Decision Processes: A Protocol Analy- sis," Journal of Consumer Research, 7 (December), 234- 48. ,Deborah Roed d er John, and Carol A. Scott(1986), "Covariation Assessment by Consumers," Journal of Con- sumer Research, 13 (December), 316-26. and Mita Sujan (1987), "Effectsof Framing on Evaluation of Comparable and Noncomparable Alternatives by Expert and Novice Consumers," Journal of Consumer Research, 14 (September), 141-54. Blattberg, RobertC. and Kenneth J. Wisniewski (1989), "Price- Ind uced Patternsof Competition," MarketingScience, 8 (Fall),291-309. Boush, David M. and Barbara Loken (1991), "A Process TracingStud y of Brand Extension Evaluations," Journal of MarketingResearch,28 (February), 16-28. ,Shannon Shipp, Barbara Loken, Ezra Gencturk, Susan Crockett, Ellen Kenned y, Betty Minshall, Dennis Misurell,Lind a Rochford ,and Jon Strobel (1987), "Affect Generalization toSimilar and Dissimilar Brand Exten- sions," Psychology and Marketing, 4 (3), 225-37. Brid ges, Sheri (1990), "A Schema Unification Mod el of Brand Extensions," unpublished d octoral d issertation,Grad uate School of Business, Stanford University. Burke,Raymond R. and ThomasK. Srull (1988), "Compet- itive Interference and Consumer Memory for Ad vertising," Journal of Consumer Research, 15 (June), 55-68. Chaiken, Shelley (1986), "The HeuristicMod el of Persua- sion," in Social Influence: The Ontario Symposium, Vol. 4, Mark P. Zanna, E. Tory Higgins, and C. P. Herman, ed s. Hillsd ale,NJ. Lawrence ErlbaumAssociates. Chattopad hyay, Amitava and Joseph W. Alba (1988), "The Situational Importance of Recall and Inference in Con- Customer-Based Brand Equity/ 19 This content downloaded from 152.3.102.242 on Mon, 30 Sep 2013 12:03:07 PM All use subject to JSTOR Terms and Conditions sumer Decision Making," Journal of Consumer Research, 15 (June), 1-12. Collins, Allan M. and Elizabeth F. Loftus(1975), "A Spread - ing Activation Theory of Semantic Processing," Psycho- logical Review,82,407-28. Craik,Fergus I. M. and RobertS. Lockhart (1972), "Levels of Processing: A Framework for Memory Research," Jour- nal of Verbal Learning and Verbal Behavior, 11,671-84. and End el Tulving(1975), "Depth of Processing and the Retention of Word sin Episod icMemory," Journal of Experimental Psychology, 104 (3), 268-94. Cohen, Joel B. and Kanul Basu (1987), "Alternative Mod els of Categorization: Toward s a Contingent Processing Framework," Journal of Consumer Research, 13 (March), 455-72. Day,George S., Allan D. Shocker, and Rajend ra K. Srivastava (1979), "Customer-Oriented Approaches to Id entifying Prod uct-Markets," Journal of Marketing, 43 (Fall),8-19. Dick,Alan,Dipankar Chakravarti, and Gabriel Biehal (1990), "Memory-Based Inferences During Consumer Choice," Journal of Consumer Research, 17 (June),82-93. Dobni, Dawn and George M. Zinkhan (1990), "In Search of Brand Image: A Found ation Analysis," in Ad vances in Consumer Research,Vol. 17,Marvin E. Gold berg, Gerald Gorn, and Richard W. Pollay, ed s. ProvoUT: Association for Consumer Research, 110-19. Dolan, Carrie (1985), "ConcoctingZingy New Names Starts Turning Intoa Business," Wall StreetJournal. Erickson,Gary M.,Johny K. Johansson, and Paul Chao (1984), "Image Variablesin Multi-Attribute Prod uctEvaluations: Country-of-Origin Effects," Journal of Consumer Re- search, 11 (September), 694-9. Farquhar, Peter H. (1989), "Managing Brand Equity," Mar- ketingResearch, 1 (September), 24-33. Fazio, Russell H. (1987), "Category-Brand Associations and Their Activation From Memory," unpublished report,Ogilvy Center for Research and Development, San Francisco,CA. , Martha C. Powell, and Carol J. Williams (1989), "The Role of Attitud e Accessibility in the Attitud e and Be- havior Process," Journal of Consumer Research, 16 (De- cember), 280-8. , David M. Sanbonmatsu, Martha C. Powell, and Frank R. Kard es (1986), "On the AutomaticActivation of Attitud es," Journal of Personality and Social Psychology, 50 (February), 229-38. and Mark P. Zanna (1981), "Direct Experiences and Attitud e-Behavior Consistency," in Ad vances in Ex- perimental Social Psychology, Vol. 14,Leonard Berkowitz, ed . New York: Acad emic Press,Inc., 161-202. Feld man, Jack M. and John G. Lynch, Jr. (1988), "Self- Generated Valid ity and Other Effects of Measurementon Belief, Attitud e,Intention,and Behavior," Journal of Ap- plied Psychology, 73 (August), 421-35. Fennell, Gerald ine (1978), "Consumer's Perceptions of the Prod uct-Use Situation," Journal of Marketing, 42 (April), 38-47. Fishbein, Martin and Icek Ajzen (1975), Belief,Attitud e,In- tention,and Behavior: An Introd uction to Theory and Re- search. Read ing, MA: Ad d ison-Wesley Publishing Com- pany. Ford ,Gary T. and Ruth Ann Smith (1987), "Inferential Be- liefs in Consumer Evaluations: An Assessmentof Alter- native ProcessingStrategies," Journal of Consumer Re- search, 14 (December),363-71. Gard ner,Burleigh B. and Sid ney J. Levy (1955), "The Prod - uctand the Brand ," Harvard Business Review,33 (March- April), 33-9. Green, Paul E. and V. Srinivasan (1978), "ConjointAnalysis in Consumer Research: Issues and Outlook," Journal of Consumer Research,5 (September), 102-23. and (1990), "ConjointAnalysis in Con- sumer Research: New Developments and Directions," Journal of Marketing, 54 (October),3-19. and YoramWind (1975), "New Ways toMeasure Consumers' Jud gments," Harvard Business Review, 53 (July/August), 107-17. Gregg,Vernon H. (1976), "Word Frequency,Recognition, and Recall," in Recall and Recognition, John Brown,ed . Lon- d on: John Wiley & Sons, Inc. Heckler,Susan,Kevin Lane Keller, and Michael J. Houston (1992), "The Effects of Brand Repositioning on Ad Re- call," workingpaper, Karl Eller School of Management, University of Arizona. Herr,Paul M.,Peter H. Farquhar, and Russell H. Fazio (1990), "Extend ing Brand Equity toNew Categories," workingpa- per, Grad uate School of Business,Ind iana University. Hertel,Paula T. (1982), "Remembering Reactionsand Facts: The Influence of SubsequentInformation," Journal of Ex- perimental Psychology: Learning,Memory, and Cognition, 8 (6), 513-29. Herzog, H. (1963), "Behavioral Science Concepts for Ana- lyzing the Consumer," in Marketing and the Behavioral Sciences, Perry Bliss, ed . Boston: Allyn and Bacon Inc., 76-86. Hoch, Stephen J. (1984), "Availability and Interference in Pred ictive Jud gment," Journal of Experimental Psychol- ogy: Learning,Memory, and Cognition, 10 (October), 629- 62. Hong, Sung-Tai and RobertS. Wyer (1989), "Effects of Country-of-Origin and Prod uctAttribute Information on Prod uctEvaluation: An Information ProcessingPerspec- tive," Journal of Consumer Research, 16 (September), 175- 87. and (1990), "Determinantsof Prod uct Evaluation: Effects of the Time Interval Between Knowl- ed ge of a Prod uct's Country of Origin and Information About Its SpecificAttributes," Journal of Consumer Research, 17 (December), 277-88. Houston,Michael J.,Terry L. Child ers, and Susan E. Heckler (1987), "Picture-Word Consistency and the Elaborative Processing of Ad vertisements," Journal of Marketing Re- search, 24 (November), 359-69. Hoyer,Wayne D. and Steven P. Brown (1990), "Effectsof Brand Awarenesson Choice for a Common,Repeat-Purchase Prod uct," Journal of Consumer Research, 17 (September), 141-8. Huber,Joel and John McCann (1982), "The Impact of Infer- ential Beliefs on Prod uct Evaluations," Journal of Market- ingResearch, 18 (November), 428-41. Isen,Alice M. (1992), "The Influence of Positive Affecton Cognitive Organization: Some Implications for the Influ- ence of Ad vertising on Decisions AboutProd ucts and Brand s," in Ad vertisingExposure,Memory, and Choice, And rew A. Mitchell, ed . Hillsd ale, NJ: Lawrence Erlbaum Associates, in press. Jacoby, Jacob and David Mazursky (1984), "Linking Brand and Retailer Images-Do the Potential Risks Outweigh the Potential Benefits?" Journal of Retailing, 60 (2), 105-22. , Jerry C. Olson, and Rafael A. Had d ock (1971), "Price,Brand Name, and Prod uct Composition Character- istics as Determinantsof Perceived Quality," Journal of Applied Psychology, 55 (December), 570-9. , George J. Syzabillo, and Jacqeline Busato-Schach (1977), "Information Acquisition Behavior in Brand Choice 20 / Journal of Marketing, January 1993 This content downloaded from 152.3.102.242 on Mon, 30 Sep 2013 12:03:07 PM All use subject to JSTOR Terms and Conditions Situations," Journal of Consumer Research,3 (4),209-16. Johnson, Michael D. (1984), "Consumer Choice Strategiesfor ComparingNoncomparable Alternatives," Journal of Con- sumer Research, 11 (December),741-53. Kahle,Lynne R. and Pamela M. Homer (1985), "Physical At- tractivenessof the Celebrity End orsor: A Social Ad aptation Perspective," Journal of Consumer Research, 11 (4),954- 61. Katz, Daniel (1960), "The Functional Approach tothe Stud y of Attitud es," Public Opinion Quarterly, 24 (Summer),163- 204. Keller, Kevin Lane (1987), "Memory Factorsin Ad vertising: The Effectof Ad vertising Retrieval Cues on Brand Eval- uations," Journal of Consumer Research, 14 (December), 316-33. (1991a), "Cue Compatibility and Framing in Ad - vertising," Journal of MarketingResearch, 28 (February), 42-57. (1991b), "Memory and Evaluation Effectsin Com- petitive Ad vertisingEnvironments," Journal of Consumer Research, 16 (March), 436-76. (1992), "Memory Retrieval Factorsand Ad vertis- ingEffectiveness," in Ad vertisingExposure,Memory, and Choice, And rew A. Mitchell, ed . Hillsd ale, NJ: Lawrence Erlbaum Associates, in press. and David A. Aaker (1992), "The Effects of Se- quential Introd uction of Brand Extensions," Journal of MarketingResearch, 29 (February), 35-50. Kotler,Philip H. (1991), MarketingManagement: Analysis, Planning, and Control, 8th ed . Englewood Cliffs, NJ: Prentice-Hall, Inc. Leuthesser, Lance, ed . (1988), "Defining, Measuring and Managing Brand Equity: A Conference Summary," Report #88-104. Cambrid ge, MA: Marketing Science Institute. Levitt,Theod ore (1960), "MarketingMyopia," Harvard Busi- ness Review, 38 (July/August), 45-56. Levy,Sid ney J. (1978),Marketing Behavior: Its Meaningfor Management. New York: AMACOM. (1981), "Interpreting Consumer Mythology: A Structural Approach toConsumer Behavior," Journal of Marketing, 45 (Summer), 49-61. (1985),"Dreams,Fairy Tales,Animals, and Cars," Psychology and Marketing, 2 (2), 67-81. Lockhart,Robert S., Fergus I. M. Craik, and Larry Jacoby (1976), "Depth of Processing, Recognition, and Recall," in Recall and Recognition, John Brown,ed ., New York: John Wiley & Sons,Inc. Loftus, Elizabeth F. and Gregory R. Loftus (1980), "On the Permanence of Stored Information in the Human Brain," American Psychologist, 35 (May), 409-20. Loken, Barbara and James Ward (1990), "Alternative Ap- proaches to Und erstand ing the Determinantsof Typical- ity," Journal of Consumer Research, 17 (September), 111- 26. Lutz, Richard J. (1991), "The Role of Attitud e Theory in Mar- keting," in Perspectives in Consumer Behavior, 4th ed ., Harold H. Kassarjian and ThomasS. Robertson,ed s. Glen- view, IL: Scott,Foresman and Company, 317-39. Lynch, John G.,Jr.,Howard Mamorstein, and Michael Weigold (1988),"ChoicesFromSets Includ ing Remembered Brand s: Use of Recalled Attributesand Prior Overall Evaluations," Journal of Consumer Research, 15 (September), 169-84. and ThomasK. Srull (1982), "Memory and Atten- tional Factorsin Consumer Choice: Concepts and Research Method s," Journal of Consumer Research, 9 (June), 18- 36. Maclnnis, Deborah J. and KentNakamoto (1991), "Factors ThatInfluence Consumers' Evaluationsof Brand Exten- sions," workingpaper, Karl Eller School of Management, University of Arizona. MacKenzie,ScottB. (1986), "The Role of Attention in Me- d iating the Effectof Ad vertising on Attribute Importance," Journal of Consumer Research, 13 (September), 174-95. Maltz,Eliot (1991), "Managing Brand Equity: A Conference Summary," Report #91-110. Cambrid ge, MA: Marketing Science Institute. Maslow,AbrahamH. (1970), Motivation and Personality,2nd ed . New York: Harper & Row Publishers,Inc. McCracken,Grant (1989), "WhoIsthe Celebrity End orsor? Cultural Found ationsof the End orsementProcess," Journal of Consumer Research, 16 (December), 310-21. Miller, K. E. and J. L. Ginter (1979), "An Investigation of Situational Variation in Brand Choice Behavior and Atti- tud e, " Journal of MarketingResearch, 16 (February),111- 23. Myers, JamesH. and Allan D. Shocker (1981), "The Nature of Prod uct-Related Attributes," in Research in Marketing, Vol. 5,Jagd ish Sheth, ed . Greenwich,CT: JAI Press,Inc., 211-36. Myers-Levy, Joan (1989), "The Influence of a Brand Name's Association SetSize and Word Frequency on Brand Mem- ory," Journal of Consumer Research, 14 (September), 197- 207. and Alice M. Tybout(1989), "Schema Congruity asa Basis for Prod uct Evaluation," Journal of Consumer Research, 16 (June), 39-54. Ned ungad i, Prakash (1990), "Recall and Consumer Consid - eration Sets: Influencing Choice Without Altering Brand Evaluations," Journal of Consumer Research, 17 (Decem- ber),263-76. and Wesley Hutchinson (1985), "The Prototypi- cality of Brand s: Relationships With Brand Awareness, Preference and Usage," in Ad vances in Consumer Re- search, Vol. 12,Elizabeth C. Hirschman and MorrisB. Holbrook,ed s. Provo, UT: Association for Consumer Re- search,489-503. Newman,Joseph W. (1957), "New Insight, New Progress for Marketing," Harvard Business Review, 35 (November- December), 95-102. Olson,Jerry C. and Jacob Jacoby (1972), "Cue Utilization in the Quality Perception Process," in The Proceed ingsof the Third Annual Conference of the Association for Consumer Research,M. Venkatesan, ed . Iowa City, IA: Association for Consumer Research,167-79. Park, C. Whan,Bernard J. Jaworski, and Deborah J. Maclnnis (1986), "Strategic Brand Concept-Image Management," Journal of Marketing, 50 (October), 621-35. and V. Parker Lessig(1981), "Familiarity and Its Impact on Consumer Biases and Heuristics," Journal of Consumer Research, 8 (September), 223-30. , Sand ra Milberg, and RobertLawson (1991), "Evaluation of Brand Extensions: The Role of Prod uctLevel Similarity and Brand ConceptConsistency," Journal of Consumer Research, 18 (September), 185-93. and Daniel C. Smith (1989), "Prod uct-Level Choice: A Top-Down or Bottom-Up Process?" Journal of Con- sumer Research, 16 (December),289-99. Park,Chan-Su (1991), "Estimation and Pred iction of Brand EquitiesThrough Survey Measurementof Consumer Pref- erence Structures," unpublished d octoral d issertation pro- posal, Grad uate School of Business, Stanford University. Petty, Richard E. and John T. Cacioppo(1986), Communi- cation and Persuasion. New York: Springer-Verlag. Plummer,Joseph T. (1985), "How Personality Makesa Dif- Customer-Based Brand Equity / 21 This content downloaded from 152.3.102.242 on Mon, 30 Sep 2013 12:03:07 PM All use subject to JSTOR Terms and Conditions ference," Journal of Ad vertisingResearch,24 (6),27-31. Raaijmakers, J. G. W. and Richard M. Shiffrin (1981),"Search of Associative Memory," Psychological Review, 88, 93- 134. Rangaswamy,Arvind ,Raymond Burke,and Terence A. Oliva (1990), "Brand Equity and the Extend ibility of Brand Names," WorkingPaper No. 90-019, The Wharton School, The University of Pennsylvania. Ratcliff,Roger and Gail McKoon (1988), "A Retrieval Theory of Priming in Memory," Psychological Review,95 (3),385- 408. Ray, Michael L. (1982), Ad vertising and Communication Management. Englewood Cliffs, NJ: Prentice-Hall,Inc. Ries, Al and Jack Trout (1979), Positioning: The Battle for Your Mind . New York: McGraw Hill,Inc. Robertson,KimR. (1987), "Recall and Recognition Effects of Brand Name Imagery," Psychology and Marketing, 4 (1), 3-15. (1989), "Strategically Desirable Brand Name Char- acteristics," Journal of Consumer Marketing, 6 (Fall), 61- 71. Roed d er John, Deborah and Barbara Loken (1990), "Diluting Brand Equity: The Impact of Brand Extensions," working paper, Carlson School of Management,University of Min- nesota. Romeo, Jean B. (1990), "The Effectof Negative Information on the Evaluation of Brand Extensions and the Family Brand ," in Ad vances in Consumer Research, Vol. 18, Rebecca H. Holman and Michael R. Solomon, ed s. Provo, UT: Association for Consumer Research,399-406. Rosch, Eleanor and Carolyn B. Mervis (1975), "Family Re- semblances: Stud ies in the Internal Structure of Cate- gories," Cognitive Psychology, 7 (October), 573-605. Roselius, Ted (1971), "Consumer Ranking of Risk Red uction Method s," Journal of Marketing, 35 (January), 56-61. Rossiter, John R. and Larry Percy (1987), Ad vertising and Promotion Management. New York: McGraw-Hill Book Company. Russo,Ed ward J. and EricJ. Johnson (1980), "WhatDoCon- sumersKnow AboutFamiliar Prod ucts?" in Ad vances in Consumer Research, Vol. 7,Jerry C. Olson,ed . Ann Ar- bor, MI: Association for Consumer Research,417-23. Simon, Carol J. and Mary W. Sullivan (1990), "The Mea- surementand Determinantsof Brand Equity: A Financial Approach," workingpaper, Grad uate School of Business, University of Chicago. Simonson,Itamar,Joel Huber, and John Payne (1988), "The Relationship Between Prior Knowled ge and Information Acquisition Ord er," Journal of Consumer Research, 14 (March), 566-78. Smith,Daniel C. and C. Whan Park (1992), "The Effects of Brand Extensions on MarketShare and Ad vertising Effi- ciency," Journal of MarketingResearch,29 (August), 296- 313. Solomon,Michael R. (1983), "The Role of Prod uctsasSocial Stimuli: A Symbolic Interactionism Perspective," Journal of Consumer Research, 10 (December),319-29. Srinivasan, V. (1979), "Network Mod elsfor Estimating Brand - Specific Effects in Multi-Attribute MarketingMod els," ManagementScience, 25 (January), 11-21. Srivastava,Rajend ra and Allan D. Shocker (1991), "Brand Equity: A Perspective on Its Meaning and Measurement," workingpaper, Grad uate School of Business,University of TexasatAustin. Srull,ThomasK. (1984), "Method ological Techniques for the Stud y of Person Memory and Social Cognition," in Hand - book of Social Cognition, Vol. 2, RobertS. Wyer and ThomasK. Srull, ed s. Hillsd ale, NJ: Lawrence Erlbaum Associates, 1-72. Starr, Martin K. and Joel R. Rubinson (1978), "A Loyalty GroupSegmentation Mod el for Brand Purchasing Simula- tion," Journal of MarketingResearch, 15 (August), 378- 83. Sujan, Mita (1985), "Consumer Knowled ge: Effectson Eval- uation StrategiesMed iating Consumer Jud gments," Jour- nal of Consumer Research, 12 (June), 31-46. and JamesR. Bettman (1989), "The Effectsof Brand PositioningStrategies on Consumers' Brand and Category Perceptions: Some Insights FromSchema Research," Jour- nal of MarketingResearch, 26 (November), 454-67. Sullivan,Mary W. (1990), "MeasuringImage Spillovers in Umbrella Brand ed Prod ucts," Journal of Business,63 (July), 309-29. Tauber,Ed ward M. (1988), "Brand Leverage: Strategy for Growth in a CostControlled World ," Journal of Ad vertis- ingResearch, 28 (August/September), 26-30. Tulving, End el (1983), Elements of Episod icMemory. Lon- d on: Oxford University Press. and Joseph Psotka (1971), "Retroactive Inhibition in Free Recall: Inaccessibility of Information Available in Memory Store," Journal of Experimental Psychology, 87 (1), 1-8. Ward ,Jamesand Barbara Loken (1986), "The Quintessential Snack Food : Measurementof Prod uct Prototypes," in Ad - vances in Consumer Research, Vol. 13, Richard J. Lutz, ed . Provo, UT: Association for Consumer Research, 126- 31. Wentz, Laurel (1989), "WPP Consid ers Brand Valuation," Ad vertisingAge, 24. Wilkie,William (1986), Consumer Behavior. New York: John Wiley & Sons,Inc. Wind ,Yoram (1982), Prod uct Policy: Concepts,Method s,and Strategy. Read ing, MA: Ad d ison-Wesley Publishing Com- pany. Wyer, Robert S., Jr. and ThomasK. Srull (1989), "Person Memory and Jud gment," Psychological Review, 96 (1), 58- 83. Yovovich, B. G. (1988), "WhatIsYour Brand Really Worth?" Ad week's Marketing Week (August8), 18-21. Zeithaml,Valarie (1988), "Consumer Perceptions of Price, Quality, and Value: A Means-End Mod el and Synthesis of the Evid ence," Journal of Marketing, 52 (July), 2-22. Reprint No. JM571100 22 / Journal of Marketing, January 1993 This content downloaded from 152.3.102.242 on Mon, 30 Sep 2013 12:03:07 PM All use subject to JSTOR Terms and Conditions