Customer SatisfactioMna, R Ket Share, and Profitability: Findings From Sweden

Customer Satisfaction, Market Share, and Profitability: Findings from Sweden
Author(s): Eugene W. Anderson, Claes Fornell, Donald R. Lehmann

Reviewed work(s):
Source: Journal of Marketing, Vol. 58, No. 3 (Jul., 1994), pp. 53-66
Published by: American Marketing Association
Stable URL: http://www.jstor.org/stable/1252310 .
Accessed: 30/04/2012 12:39
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 [email protected].
American Marketing Association is collaborating with JSTOR to digitize, preserve and extend access to
Journal of Marketing.
http://www.jstor.org
Eugene W.Anderson,Claes Fornell,& DonaldR. Lehmann
Customer Satisfaction,Market
Share, and Profitability:
Findings From Sweden
Are there economic benefits to improving customer satisfaction? Many firms that are frustrated in their efforts to im-
prove quality and customer satisfaction are beginning to question the link between customer satisfaction and eco-
nomic returns. The authors investigate the nature and strength of this link. They discuss how expectations, quality,
and price should affect customer satisfaction and why customer satisfaction, in turn, should affect profitability;this
results in a set of hypotheses that are tested using a national customer satisfaction index and traditional account-
ing measures of economic returns, such as return on investment. The findings support a positive impact of quality
on customer satisfaction, and, in turn, profitability.The authors demonstrate the economic benefits of increasing cus-
tomer satisfaction using both an empirical forecast and a new analytical model. In addition, they discuss why in-
creasing market share actually might lead to lower customer satisfaction and provide preliminary empirical support
for this hypothesis. Finally, two new findings emerge: First, the market's expectations of the quality of a firm's out-
put positively affects customers' overall satisfaction with the firm; and second, these expectations are largely ra-
tional, albeit with a small adaptive component.
Does higher customer satisfactionlead to superioreco- companiesare wastingtheireffortsin tryingto improvequal-

nomic returns?Widespreadacceptanceof this relation- ity (American Quality Foundation 1992). The consulting
ship is evident in the growing popularliteratureon quality firms of A.T Kearey and ArthurD. Little present equally
and customersatisfaction,the increasingnumberof consult- disappointingfindings in two separatestudies: (1) 80% of
ing and marketingresearchfirms that promise to improve a more than 100 British firms reported "no significant im-
client's ability to satisfy customers,and-perhaps most per- pact as a result of TQM" and (2) almost two-thirdsof 500
suasively from a market-orientedperspective-the number U.S. companies saw "zero competitive gains" (The Econo-
of organizationsactively using some form of customer sat- mist 1992).
isfaction measurement in developing, monitoring, and/or If frustration with attempts to improve quality leads
evaluatingproductand service offerings, as well as for eval- many business firms to abandon the Quality Movement
uating, motivating, and/orcompensatingemployees. (Newsweek 1992), the recent surge of interest in customer
However, at the level of the firm, recent empirical evi- satisfactionis likely to follow the same path-unless it can
dence casts doubt on whether companies' efforts to im- be demonstratedthat there are positive economic returnsto
prove customer satisfactionand quality throughimplemen- improvingcustomer satisfaction.Firms will appropriatere-
tation approachessuch as total quality management(TQM) sources for improving customer satisfaction only if the ef-
actually are having the desired effects. Specifically, several fects are of sufficient size, as measured by traditional ac-
surveys point to the failure of TQM to enhance either eco- counting methods.
nomic returnsor competitiveness.A study by the American In view of these facts, it is not surprisingthat thereis re-
Quality Foundationand Ernst & Young suggests that many surgentinterest in understandingthe links between quality,
customer satisfaction, and firm performance (e.g., eco-
Eugene W.Andersonis anAssistant
Professorof Marketing
andClaesFor- nomic returns).1In a meta-analysis of strategy variables,
nellis theDonaldC.CookProfessor of BusinessAdministration,
National Capon, Farley, and Hoenig (1990) identify 20 studies that
Quality ResearchCenter,Schoolof BusinessAdministration,
TheUniver- find a positive relationshipbetweenqualityand economic re-
sityof Michigan.
Donald R. Lehmann is theGeorgeE.Warren Professor turns. For example, Buzzell and Gale (1987) and Phillips,
of Marketing,
Graduate Schoolof Business,Columbia Theau-
University.
thorsgratefully thedataprovided thefunding Chang,and Buzzell (1983) each reporta significantrelation-
acknowledge through ofthe
SwedishPostandthesupport oftheNationalQualityResearchCenterat ship between relative quality-as perceived by the business
theUniversity
of MichiganBusinessSchool.Funding fortheanalysis
was unit-and returnon investment (ROI) for firms represented
provided bythe MarketingScienceInstitute's
Market-Driven re-
Quality in the PIMS database. In the last few years, researchers
searchcompetition.
Thisworkhas benefitted fromthecom-
substantially have startedto elaborateon the process by which delivering
mentsof RickStaelin,
JohnHauser, BartWeitz,Roland Rust,andpartici-
pantsintheCustomer Satisfaction
Workshop attheUniversity
of Michigan lIn marketing,customer satisfactionlong has been recognized as a cen-
Business School.TheauthorsthankJaesungChaandJaySinhafortheir tral concept, as well as an importantgoal of all business activity. Satisfac-
helpwiththedata. tion is a core concept in the American Marketing Association's official
definition of marketing.
Journal of Marketing
Vol. 58 (July 1994), 53-66 CustomerSatisfaction,MarketShare,and Profitability
/ 53
high-quality goods and services influences profitability tomer satisfaction?In this section, we develop a conceptual
throughcustomersatisfaction.Building from the individual- frameworklinking customer-basedmeasuresof firm perfor-
level model of customer satisfaction proposed by Oliver mance (e.g., customersatisfaction)with traditionalaccount-
(1980), several studies discuss and/or observe a strong link ing measuresof economic returns,such as ROI.
between customer satisfaction and loyalty (Anderson and Before proceeding,it is importantto make clear what is
Sullivan 1993; Bearden and Teel 1983; Boulding et al. meant by "customer satisfaction" in the context of this
1993; Fornell 1992; LaBarberaand Mazursky 1983; Oliver study. At least two differentconceptualizationsof customer
and Swan 1989). Reichheld and Sasser (1990) discuss why satisfactioncan be distinguished:transaction-specificand cu-
increasingcustomerloyalty should lead to higherprofitabil- mulative (Boulding et al. 1993). From a transaction-specific
ity. Rust and Zahorik (1993) empirically demonstratethe perspective, customer satisfaction is viewed as a post-
relationshipbetween customer satisfactionand profitability choice evaluativejudgment of a specific purchaseoccasion
for a health care organization. (Hunt 1977; Oliver 1977, 1980, 1993). Behavioralresearch-
Our purpose is to examine more closely the links be- ers in marketinghave developed a rich body of literaturein-
tween customer-based measures of firm performance- vestigating the antecedents and consequences of this type
such as customer satisfaction-and traditionalaccounting of customersatisfactionat the individuallevel (see Yi 1991
measures of economic returns.Although there have been a for a review). By comparison,cumulativecustomersatisfac-
few firm-specific studies (e.g., Rust and Zahorik1993), this tion is an overall evaluationbased on the total purchaseand
article represents the first large-scale examination of the consumption experience with a good or service over time
relationship. (Fornell 1992; Johnsonand Forell 1991). Whereastransac-
A unique featureof our empiricalwork is the set of cus- tion-specific satisfactionmay providespecific diagnosticin-
tomer-basedperformancemeasures for firms participating formationabouta particularproductor serviceencounter,cu-
in the Swedish Customer Satisfaction Barometer (SCSB) mulative satisfactionis a more fundamentalindicatorof the
(see Fornell 1992 for a description). The SCSB provides firm's past, current,and future performance.It is cumula-
yearly firm-level indices of quality, expectations,and over- tive satisfaction that motivates a firm's investment in cus-
all customer satisfaction for majorcompetitorsin a variety tomersatisfaction.Because the focus here is on the relation-
of product and service industries.Importantly,each firm's ship between customer satisfaction and economic returns,
set of indices is an estimate based on an annual survey of our theoretical framework treats customer satisfaction as
currentcustomers ratherthan a set of unstandardizednum- cumulative.
bers drawn from multiple "independent" sources (e.g., Whatis quality and how is it distinct from customersat-
trade press, consumer advocates) or based on an internal, isfaction? In this study, perceived quality is taken to be a
self-reportedmeasureof quality.The SCSB providesa stan- global judgment of a supplier's current offering
dard set of customer-basedperformancemeasures that can (Steenkamp 1989). This is similar in spirit to the position
be matchedto economic performancemeasures,such as mar- taken by Zeithaml(1988, p. 3) in summarizingan extensive
ket share and ROI. review of the literatureon quality: "Perceived quality can
Predictionof economic returnsis one of the centralpur- be defined as the consumer's judgment about a product's
poses of the SCSB. The index is constructedusing a meth- overall excellence or superiority."However, it is worthnot-
odology that maximizes the relationshipbetween customer ing thatthereare severaldistinctconceptualizationsof qual-
satisfaction and the likelihood of repeat purchase.It is im- ity (Holbrook 1994). In marketingand economics, quality
portant to note that this methodology distinguishes the often has been viewed as dependenton the level of product
SCSB measures from other common approaches used to attributes(e.g., Hauser and Shugan 1983; Rosen 1974). In
combine the facets of customer satisfaction into a single operations management (e.g., Garvin 1988; Juran 1988),
index-unit weighting schemes or some variationof factor qualityis definedas having two primarydimensions:(1) Fit-
analysis (e.g., the J.D. Power Index for automobiles). The ness for use-Does the productor service do what it is sup-
logic behind the SCSB methodology is to derive the posed to do? Does it possess featuresthat meet the needs of
weights with respect to a proxy for economic returns(e.g., customers?and (2) Reliability-To what extent is the prod-
customerloyalty), providinga betterchance of predictingac- uct free from deficiencies? In the services literaturein mar-
tual economic returns(Fomell 1992). keting, qualityis viewed as an overall assessment (e.g., Par-
We begin by defining and discussing the links between asuraman,Zeithaml, and Berry 1985). Service quality in
quality, expectations, customer satisfaction, and profitabil- this context is believed to depend on gaps between deliv-
ity, as well as the relationship between customer satisfac- ered and desired service on certaindimensions.
tion and marketshare. Next, the data and methodology are The theoretical framework presented here views cus-
discussed. Finally, we presentthe findings and discuss their tomer satisfaction as distinct from quality for several rea-
implications. sons. First, customers requireexperience with a productto
determine how satisfied they are with it. Quality, on the
otherhand,can be perceivedwithoutactualconsumptionex-
Customer Satisfaction and perience (Oliver 1993). Second, it has been long recognized
Profitability that customer satisfaction is dependent on value (Howard
How does satisfying currentcustomers affect profitability? and Sheth 1969; Kotler and Levy 1969), where value can
How do market expectations and experiences affect cus- be viewed as the ratio of perceived quality relative to price
54 / Journalof Marketing,July 1994

or benefits received relative to costs incurred (Dodds, TABLE 1
Monroe, and Grewal 1991; Holbrook 1994; Zeithaml General Specification of the Conceptual Model
1988). Hence, customer satisfaction is also dependent on
EXPECTATIONSt= fl (EXPECTATIONSt1,QUALITYt_1,l,t)
price, whereas the quality of a good or service is not gener-
ally considered to be dependent on price. Third, we view = f2(QUALITYt,
SATISFACTIONt PRICEt,EXPECTATIONSt,2t)
quality as it pertains to customer's currentperception of a PROFITABILITYt = f3 (SATISFACTIONt,43t)
good or service, whereas customer satisfaction is based on
not only currentexperience but also all past experiences, as where tit = vector of other factors (e.g., environmental
well as future or anticipated experiences. Finally, there is trends, firm-specificfactors, error)
ample empirical supportfor quality as an antecedentof cus-
tomer satisfaction (e.g., Anderson and Sullivan 1993; Customer satisfaction should reduce price elasticities
Churchilland Suprenant1982; Croninand Taylor1992; For- for current customers (Garvin 1988). Satisfied customers
nell 1992; Oliver and DeSarbo 1988). are more willing to pay for the benefits they receive and are
more likely to be tolerantof increases in price. This implies
Overview of the Theoretical Framework
high margins and customer loyalty (Reichheld and Sasser
The theoretical framework developed in the remainderof 1990). Low customer satisfaction implies greater turnover
this section can be summarizedin the general set of equa- of the customerbase, higher replacementcosts, and, due to
tions presented in Table 1. Profitability at time t is posi- the difficulty of attracting customers who are satisfied
tively affected by customersatisfaction,as well as otherfac- doing business with a rival, higher customer acquisition
tors (e.g., past values of the dependent variable, economic costs. Decreased price elasticities lead to increased profits
conditions, firm-specific factors, luck, error).Customersat- for a firm providing superiorcustomer satisfaction.
isfaction, in turn, is positively affected by marketexpecta- High customer satisfaction should lower the costs of
tions and experiences. Finally, currentmarketexpectations transactionsin the future.If a firm has high customerreten-
are positively relatedto both historicalexpectationsand the tion, it does not need to spend as much to acquirenew cus-
market'sexperiences with quality in the most recentperiod. tomers each period. Satisfied customers are likely to buy
The nature of each of these relationships is discussed more frequently and in greatervolume and purchaseother
subsequently. goods and services offered by the firm (Reichheld and Sas-
ser 1990).
How Does Customer Satisfaction Affect
Profitability? Consistently providing goods and services that satisfy
customers should increase profitabilityby reducing failure
Fornell (1992) enumeratesseveral key benefits of high cus- costs. A firm that consistentlyprovideshigh customersatis-
tomer satisfactionfor the firm. In general,high customersat- faction should have fewer resourcesdevoted to handlingre-
isfaction should indicate increased loyalty for currentcus- turns,reworkingdefective items, and handling and manag-
tomers, reduced price elasticities, insulation of currentcus- ing complaints (Crosby 1979; Garvin 1988; TARP 1979,
tomers from competitive efforts, lower costs of futuretrans-
1981).
actions, reduced failure costs, lower costs of attractingnew The costs of attractingnew customers should be lower
customers, and an enhanced reputation for the firm. In- for firms that achieve a high level of customer satisfaction
creased loyalty of currentcustomersmeans more customers
(Fornell 1992). For example, satisfied customers are reput-
will repurchase (be retained) in the future. If a firm has
edly more likely to engage in positive word of mouth, and
strong customer loyalty, it should be reflected in the firm's less likely to engage in damaging negative word of mouth,
economic returnsbecause it ensures a steady stream of fu- for the firm (Anderson 1994b; Howard and Sheth 1969;
ture cash flow (Reichheld and Sasser 1990). Reichheld and Sasser 1990; TARP 1979, 1981). Media
The more loyal customers become, the longer they are sources are also more likely to convey positive information
likely to continue to purchasefrom the same supplier.The to prospective buyers. Customer satisfaction claims may
cumulative value of a loyal customer to a firm can be quite make advertisingmore effective, and high customersatisfac-
high. For example, consider the lunch habits of three col- tion may allow the firm to offer more attractivewarranties.
leagues that regularly patronize a restaurantclose to their An increase in customer satisfaction also should en-
workplace. If the average price of a meal is $6 and the trio hance the overall reputationof the firm. An enhancedrepu-
visits the restaurantthree times a week, the annualrevenue tation can aid in introducingnew productsby providing in-
received by the establishment is in the neighborhood of stant awareness and lowering the buyer's risk of trial (Ro-
$2,700. One hundred similarly loyal customers would be bertson and Gatignon 1986; Schmalansee 1978). Reputa-
worth$90,000. This groupwould be worthalmost a half mil- tion also can be beneficial in establishing and maintaining
lion dollars over the next five years-even if they all col-
relationshipswith key suppliers, distributors,and potential
luded to keep the restauranta secretfrom otherpotentialcus- allies (Anderson and Weitz 1989; Montgomery 1975).
tomers. The net present value of the expected margin from
Reputationcan provide a halo effect for the firm that posi-
these customersreflects theirasset value to the restaurant.In-
tively influences customerevaluations,providinginsulation
creasing customer satisfaction increases the value of a from short-term shocks in the environment. Customer
firm's customer assets and future profitability. satisfaction should play an importantrole in building other
CustomerSatisfaction,MarketShare,and Profitability
/ 55
importantassets for the firm, such as brand equity (Aaker relative to the other (Fornell 1992). The resulting index
1992; Keller 1993). measuresthe value received by customers.The expected re-
The first hypothesis of the model can be stated as lationship between quality and satisfaction can be summa-
follows: rized as follows:
H1: Customersatisfactionhas a positiveeffect on economic H2: The currentlevel of qualityas perceivedby the market
returns. should have a positive effect on overall customer
satisfaction.
Although there are many compelling reasons to con-
clude that higher customer satisfactionleads to higher prof- Expectations about the quality of goods and services
also should have a positive impact on customersatisfaction.
itability, it is, nevertheless, not always the case. At some At the aggregate level of analysis here, expectations cap-
point there must be diminishing returnsto increasing cus- tures the accumulatedknowledge of the marketconcerning
tomer satisfaction.For example, many companiesseek to in-
a given supplier's quality. Just as current quality is ex-
crease customer satisfactionby investing in quality control.
There are many economic benefits associated with this ac- pected to have a positive influence on overall customersat-
isfaction, so should all past experienceswith quality,as cap-
tivity (e.g., less rework, lower warrantycosts). Yet quality tured by expectations. In addition,expectations contain in-
control is likely to have its greatestimpact when reliability
formationbased on not actual consumptionexperience but
is relatively low, and there may come a point when the
accumulated information about quality from outside
costs associated with reducing the probability of defects
will be greaterthan the benefits to the firm. sources, such as advertising, word of mouth, and general
media. Like past experience, positive information about
There is also evidence that conformity to specifications
is not as importantin determiningoverall customersatisfac- past quality should affect customer satisfactionpositively.
In addition, in forming expectations, consumers use
tion as the design of a product or service in meeting cus-
past experience and nonexperiential information to con-
tomer needs (Anderson and Sullivan 1993). Given that in- structforecasts of the supplier's ability to deliver quality in
creasing quality and customer satisfaction by design (e.g., the future. This role of expectations is importantbecause
adding features,increasingthe quality of raw materialsand/ the natureof the ongoing relationship between a firm and
or level of features,increasingthe level of personalservice, its customerbase is such thatexpected futurequality is crit-
providing greatervariety by differentiatingthe productline ical to customer satisfaction and retention as it relates to
to meet needs) is likely to increase costs at an increasing
long-term relationships with customers (Bateson 1989;
rate (Shugan 1989), it is likely that thereare diminishingre-
Czepiel and Gilmore 1987; Gronroos 1990; Lovelock
turnsto efforts to improveproductor service qualityand cus-
1984; Shostack 1977). In durable goods categories, cus-
tomer satisfaction. tomer satisfaction depends on both whether the currently
Firm-Level Antecedents owned product will continue to meet customer needs and
of Customer Satisfaction
the anticipated quality of future service. In service indus-
As Table 1 indicates, customer satisfaction is affected by tries, client satisfactionwith the vendor depends on the an-
overall quality, price, and expectations. At the individual ticipated quality of future service as well as the ability of
customer level, several studies have shown that perceived the service to providefor futureneeds. This forecastcompo-
quality affects customer satisfaction (Anderson and Sulli- nent of expectationsalso arguesfor a positive impact of ex-
van 1993; Bearden and Teel 1983; Bolton and Drew 1991; pectationson satisfaction.
Cadotte,Woodruff,and Jenkins 1987; Churchilland Supre- The precedingfactors suggest that we should expect the
nant 1982; Fornell 1992; Oliver and DeSarbo 1988; Tse aggregatemeasureof expectationsused here to have a posi-
and Wilton 1988). This relationshipis intuitive and funda- tive impact on overall customersatisfaction.Althoughthere
mental to all economic activity. Aggregated to the firm may be individual differences affecting expectations at the
level, customers' currentexperience with a supplier'soffer- individuallevel, such differencesshouldcancel out in the ag-
ing also should have a positive influence on their overall as- gregate (Katona 1979). In aggregatingexpectations across
sessment of how satisfied they are with that supplier. customers to the level of the firm, expectations should re-
In addition,price plays an importantrole in this relation- flect more accuratelyboth a firm's reputationfor providing
ship. The received value of a supplier's offering-that is, high (or low) quality and its ability to do so in the future.
quality relative to price-has a direct impact on how satis- The U.S. auto industryprovides an interestingexample
fied customers are with that supplier (Anderson and Sulli- of the effects of expectationson customer satisfaction.The
van 1993; Fornell 1992; Sawyer and Dickson 1984; Zei- reputationof Detroit's productssuffered in the 1970s and a
thaml 1988). Anterasian and Phillips (1988) discuss the good portion of the 1980s. Past negative experiences,
role of value in driving overall business performance. In broadly disseminated through word of mouth and media
both our conceptualization and measurementof quality, it sources, contributedto lower overall expectations with the
is important to consider the relationship between quality productsand service that accompaniesthem. It is likely that
and price. In our empiricalwork, in view of the proposition overall customersatisfactionin the late 1980s was therefore
thatprice affects satisfactionand the possibilityof confound- lower due to not only customers' experiences in the 1970s
ing effects of a price-qualityrelationship-as well the need and 1980s, but also anticipated lower quality. A case in
to compare the hedonic value of quality across categories point is the MercuryTracer and Mazda 323, two virtually
(Lancaster1979; Rosen 1974)-each constructis measured identical cars. Mazda customers were more satisfied over-

all, ceteris paribus,because Mazdacustomershad higherex- isfaction at time t also depends on a forecast of what qual-
pectations than Mercurycustomers (e.g., continuedreliabil- ity will be like in t + 1, t + 2, ..., as well as the impact of all
ity, durability,positive service encounters).This, of course, past quality experiences from t - 1, t - 2, .... Overall cus-
is contradictoryto the pervasive belief that firms that ex- tomer satisfactionwith a particularfirm is a function of all
ceed their customers' expectations will enjoy an immediate past, current,and futureexperience:
increase in customer satisfaction, but it is consistent with
the cumulative notion of satisfaction. H3: Themarket'sexpectationof thequalityof a supplier'sof-
feringshouldhave a positiveeffecton overallcustomer
The argumentsadvanced here differ from those associ- satisfaction.
ated with the transaction-specificconceptualizationof cus-
tomer satisfaction. In a transaction-specific situation, we Customers' Expectations of the Firm's Quality Are
might expect an increase (decrease) in a consumer's expec- Adaptive
tations to lead to a short-termfall (rise) in that consumer's The experiences of customers in a previous period t - 1
satisfactionwith a specific transaction.In the context of cu- should have a positive influence on buyer's expectationsof
mulative customer satisfaction, the long-run effects of in-
quality in the currentperiod t. Customersare likely to up-
creased(decreased)expectationsshould outweigh this short- date expectations on the basis of both past experience and
term effect and lead to a rise (fall) in overall customersatis- other types of nonexperiential information. This updating
faction. Overall customer satisfaction aggregates customer
process is consistent with the notion of adaptive expecta-
experiences over time, and we expect the effect of any tem- tions foundin both psychologicalandeconomic research.Ol-
porarydisconfirmationof expectations to be marginal(An- iver and Winer (1987) provide a comprehensivereview of
derson and Sullivan 1993). Our firm-level measuresof cus- different approachesto modeling the updating of expecta-
tomer satisfactionalso aggregateacross customers,and, un- tions. Johnson,Anderson,and Fornell (1994) comparealter-
less disconfirmationis systematic and widespread,positive native approachesfor modeling expectations and find that
and negative experiences should cancel out and their effect
expectations are very nearly rational in characterbut that
on overall satisfactionshouldbe marginal.Due to this aggre-
they are adjusted over time in an adaptive manner (Lucas
gation process, we expect overall satisfaction to be reflec- 1973; Muth 1961; Taylor 1979). That is, the marketconsid-
tive of actual past levels of perceived quality or delivered ers all available informationconcerningquality and contin-
value and forecasted future quality, ratherthan dominated
ually updates expectations in an efficient manner save for
by the effects of any perceived gap between currentquality "imperfections" (e.g., uncertainty,costs) that impede the
and expectations. This argumentis persuasivefrom a com- flow of information and result in a small updating effect
petitive perspective as well, because expectations and per- that gives the appearanceof being adaptive.
ceived quality cannot remain out of sync for very long in a The relativesize of the adaptiveupdatingeffect is impor-
mature,competitive market.If expectationsare too low, the tant and depends on both productionand consumptionfac-
firm will not attractcustomersand, consequently,new sales tors (Anderson 1994a; Anderson and Sullivan 1993). On
will not develop. If expectations are too high, customers the production side, greater temporal variation in quality
will buy, become dissatisfied, and switch to competitive should imply a greaterupdatingeffect. For example, a high
products, and, again, the firm will have deficient sales. At rate of innovation or technological change may provide
any given time, therefore,the differencebetweenactualqual- shocks that requirethe marketto revise expectations.Qual-
ity and expectations at the aggregatelevel should be small. ity also may change because of period-to-period fluctua-
Although the present aggregate-level study does not tions in materials, production, or the service delivery sys-
allow us to evaluate the efficacy of these argumentscom- tem (e.g., business cycles). Conversely, there should be less
pletely-such as comparing the relative importanceof the of an updatingor learningeffect when thereis greaterstabil-
negative influence of expectations on satisfactionby a per- ity. In this case, the market'sexpectations(based on similar
ceived gap between quality and expectationsversus the im- past experiences) should mirrorthe level of quality experi-
portanceof the positive directimpactof expectationson cus- enced in the currentperiod.
tomer satisfaction-we nonetheless expect to find that the On the consumptionside, the market'sdegree of uncer-
lattereffect is strongerand the impactof expectationson cu-
tainty regarding a particularproduct or service encounter
mulative customer satisfaction is positive. At the same can influence the size of the updatingcoefficient. For exam-
time, it is worth noting that the conclusions reachedprevi- ple, where there is little familiarityor expertise among cur-
ously are not without support. The preceding argument rent customers,it is more likely that the updatingeffect will
leads to the same conclusion reachedby Boulding and col- be large. The mix of newly acquiredversus repeatcustom-
leagues (1993) in an individual-level study of the effects of ers consequently can influence the size of the updatingco-
expectations on overall judgments. Their argumentfor how efficient, as can frequencyof purchase,the stage of market
"will expectations" (expectations of what quality service evolution, or shifting sociodemographic factors. For some
will be like, as distinct from what quality should be like) af- products,marketinformationconcerning the quality of the
fect quality perceptions is based on an adaptationmecha- good or service may be costly or difficult to obtain without
nism (Helson 1964; Oliver 1980), in a manneranalogous to experience (Darby and Kari 1973; Nelson 1970; Zeithaml
an assimilation effect (Anderson and Sullivan 1993). The 1981). In attractingnew customers, advertising itself also
marketlevel argumentpresentedhere is differentin that the can influencethe size of the updatingeffect. Althoughadver-
effect of the market'sexpectationson customers'overall sat- tising may not necessarily distort expectations (e.g., puff-
/ 57
ery), it is unlikely to providecomplete information.Custom- ments advancedin the previous section also provide a com-
ers may be attractedby a limited set of particularbenefits pelling argumentfor a relatively small updatingcoefficient.
stressed in advertising,but they must experience the prod- The difference between the market's expectations and ac-
uct or service to learn more fully about quality-and then tual experiences with quality cannot be great for long peri-
may revise expectations accordingly. A similar argument ods of time or the firm will not survive.
might be constructedfor the efficiency of word of mouth in The preceding arguments can be summarized as
conveying informationabout quality. follows:
Uncertaintyalso can arise if it is difficult for customers H4: The marketplacehas adaptiveexpectationsconcerning
to predict what their consumption experiences will be like the qualityof a supplier'soffering.The size of the adap-
over time. This may be the case if a productor service has tive updatingeffectshouldbe small.
importantexperienceattributes(attributesthatmust be expe-
rienced to be evaluated) or credence attributes(those that Relative Importance of Quality and Expectations
are very difficult to evaluate and force the customerto rely If both currentquality and expectationshave a positive im-
on the product's reputation to evaluate them) (Darby and pact on customer satisfaction,then which effect should we
Karni 1973; Nelson 1970; Zeithaml 1981). If certain as- expect to be stronger?If expectations primarily represent
pects of quality are unobservableor difficult to anticipate,it past qualityexperiencesand/ornonexperientialqualityinfor-
may be problematicfor the marketto predictfuturequality. mation, we would expect currentquality to have a greater
Expectations of quality for a particularfirm would be up- impact for several reasons. First, current quality experi-
dated as information about actual quality becomes availa- ences should be more salient and take precedenceover past
ble. For example, automobilecustomerslearnaboutdurabil- qualityexperiencesin determiningcustomersatisfaction.Ac-
ity, reliability, and quality of service over time. Personal tual experience with a good or service should outweigh
computerusers are likely to encounterunanticipatedbene- other information,especially in the aggregate. In addition,
fits and difficulties as new applications are identified and perceivedqualityis measuredin our study as perceivedqual-
complementaryproducts develop. Customers may have to ity relative to price and contains additionalinformationthat
adaptas the natureof an offering becomes apparent.In con- expectations do not contain. Finally, Oliver (1989) argues
trast, if informationis relatively complete and easy to ob- that transaction-specificsatisfaction for ongoing consump-
tain, the period-to-periodupdatingeffect at the marketlevel tion activities (durablegoods, services, and repeatedlypur-
should be small. Similarly,theremay be less updatingif var- chasing packaged goods) should be primarilya function of
iation in production or consumption is indistinguishable perceived performance.Expectationsshould be passive and
from white noise. This might be the case if a productor ser- have a minimal effect on satisfactionunderthese conditions
vice is difficult to standardizeor quality is difficult for buy- (Bolton and Drew 1991; Oliver 1989). In such situations,
ers to evaluate (Anderson 1994a; Anderson and Sullivan the level of and even degree of variationin quality is well
1993; Deighton 1984; Hoch and Ha 1984). known to customers. This same argumenthas even greater
It can be surmisedfrom these statementsthatthe size of force when the focus is on cumulative customer satisfac-
the updatingeffect dependslargelyon the rateat which qual- tion. Cumulativecustomersatisfactionis based on many ex-
ity changes over time and the market learns. The rate of periences. Customer knowledge, particularlyin relatively
learning or adjustmentby the marketis not likely to be in- matureand stable markets,should be such that expectations
stantaneous-as it might be if the marketwere perfectly ef- should accurately mirrorcurrentquality. The contribution
ficient-due to the cost of acquiringinformationand the ef- of expectationsto customersatisfactionshould be mainly in
fects of uncertaintydiscussed previously. Another implica- the form of predictingfuturequality. Unless there is uncer-
tion is that the updating effect should be small relative to tainty with regardto futurequality, the contributionof ex-
the cumulative effect of all past information.In Sweden, as pectations to overall customer satisfaction should be mini-
in other industrializednations, most industries are mature. mal (Anderson 1994a). In the extreme, expectations pro-
In more mature markets, production-side factors are such vide no additionalinformation.
that quality is relatively stable-even though the most Sweden's economy is well developed. The selected cat-
highly evolved (or complacent) competitorsin these indus- egories are mature,even thoughthese categoriesare compet-
tries certainlyhave been forced to change duringthe period itive and subjectto change-as well as perceivedwith a lim-
of the study. Customers in mature markets may have ited degree of uncertainty-and information flows rela-
greaterexperience with and knowledge of quality (Johnson tively freely. Accordingly,just as we expect the updatingof
and Fornell 1991). This implies that the updating coeffi- expectations from period to period to be small, we argue
cient, representingthe relative weight given by the market the following:
to the most recent information about quality, should be
small relative to the size of the coefficient of lagged expec- H5: Theimpactof perceivedqualityon overallcustomersat-
isfactionshouldbe relativelygreaterthanthe impactof ex-
tations, that is, the relative weight given by the marketto all pectationsabout quality.
past informationabout quality.
We argue that the processes described previously Customer Satisfaction and Market Share
should lead to a similar finding at the firm level, just as Intuitively,customersatisfactionand marketsharemight be
Boulding and colleagues (1993) find evidence for a small up- expected to go hand in hand. Buzzell and Wiersema
dating effect at the individual level. The competitive argu- (198 la, b) find relative quality and marketshare to be posi-

tively relatedfor firms in the PIMS database(though recent In summary,the relationshipbetween customersatisfac-
work by Szymanski, Bharadwaj, and Varadarajan[1993] tion and market share is an emerging issue in need of
suggests this may be the case only for PIMS data or when greaterunderstanding.Achieving success in one may lower
the employed methodology does not control for "unobserv- performancein the other. Marketsharecan be gained by at-
ables"). The same type of relationship might be expected tracting customers with preferences more distant from the
for customer satisfaction.For example, high customersatis- targetmarket.Service capabilities also can be overextended
faction should help in attracting as well as retaining as volume grows. Market share effects on profitabilityare
customers. equally problematic(see Szymanski, Bharadwaj,and Vara-
However, it is not clear that high customer satisfaction darajan1993 for a review of the market share-profitability
and high market share are always compatible. Fornell relationship). Clearly, there can be situations in which in-
(1992) and Griffin and Hauser (1993) discuss the possibil- creasing one and/or the other is not profitablefor the firm.
For example, an extreme approach for maximizing cus-
ity of a negative relationshipbetween customer satisfaction
and marketshare. They argue that whereas a small market- tomer satisfaction would be to eliminate all but one cus-
sharefirm may serve a niche marketquite well, a large mar- tomer and direct all resourcesto that individual.Obviously,
it would be a rare set of circumstances under which it
ket-sharefirm must serve a more diverse and heterogeneous
would be profitable to do so. Conversely, a high market
set of customers.At least two primaryforces are at work in
shareor "one size fits all" strategyis likely to be profitable
determiningwhetherthe relationshipbetween customersat-
isfaction and marketshare is positive or negative. First, in- only if enough customershave similarpreferences.It is also
possible that differentiationmay fail to provide higher sat-
creasing market share, at least up to a point, can produce isfaction due to the difficulty of serving multiple customers
economies of scale. This, for example, may allow the firm
within each segment and the dilution of effort that comes
to charge lower prices, thus increasing the value of the
from serving multiple segments. A firm that manages both
firm's offering and consequently increasingcustomersatis- to provide high customer satisfactionby customizing its of-
faction. By contrast, there may be a dilution of effort that
fering to each customer and maintaina large market share
goes with trying to serve an increasing numberof custom- would have to enjoy very high economies of scope and
ers and/orsegments. This dilution could lead to low-quality scale. Another way to think about this issue is to consider
service and is likely to occur in industries in which cus- what the small niche firm has to do to be successful. Provid-
tomer preferences are heterogeneous and/or personal ser- ing superiorcustomersatisfactionis critical for its survival.
vice is important.In undifferentiatedindustrieswith homo-
geneous customer preferences, it is more likely that cus-
tomer satisfactionand marketshareare positively related,es- Data and Methodology
pecially in the long run.
It is instructive to examine these arguments for the Description of the Data
cases of firms pursuing different "generic" strategies- Annual indices of firm-level expectations,quality, and cus-
differentiation, niche, and low-cost leadership-as origi- tomer satisfaction are made available by the SCSB. Initi-
ated in 1989, the SCSB is an ongoing project managed by
nally categorized by Porter (1980). Firms following pure
niche strategies are likely to be more successful at satisfy- the NationalQualityResearchCenter(NQRC) at the Univer-
ing customers than those pursuing other strategies. Al- sity of MichiganBusiness School and the InternationalCen-
ter for Studies of Quality and Productivity (ICQP) at the
though it is true that firms can differentiatetheir offerings
to meet the needs of multiple segments, it may become dif- Stockholm School of Economics. The 77 firms included in
ficult or costly to do so without diluting the quality of what our NQRC study are all majorcompetitorsin a wide variety
is provided (e.g., personal service). As a firm grows by of industries:airlines, automobiles,banking (consumerand
business), chartertravel, clothing retail, departmentstores,
bringing in customers with preferences furtheraway from
the firm's targetmarket,the overall level of customer satis- furniturestores, gas stations,insurance(life, auto, and busi-
faction is likely to fall. ness), mainframe computers (business), PCs (business),
It is worth noting that this situationis complex because newspapers, shipping (business), and supermarkets. The
of the dual impact of quality and price on satisfaction.For companies surveyed in each industry are the largest share
firms such thatcumulativeshareis approximately70%. Sev-
example, in a market in which there is a relatively large eral state-owned monopolies are also measured by the
price-sensitive segment with homogeneous needs, a low- SCSB but are not included in this study.
cost leader may provide a level of value that creates a rela-
The measurementsin the SCSB begin with a computer-
tively high level of customer satisfaction.There is clearly a aided telephone survey designed to obtain a representative
need for understanding the trade-offs in such situations
sample of customers for each firm. Potential respondents
(e.g., price elasticity versus quality elasticity of returns),if are selected on the basis of recently having purchasedand
there are conditions under which customer satisfaction and used a company's product.To participate,each respondent
marketshare are negatively related.If lowering price can at- is requiredto pass a batteryof screeningquestions.The ques-
tract customers that become less satisfied while increasing tionnaireemploys 10-point scales to collect multiple meas-
the satisfaction of the currentcustomer base, then what are ures for each construct. For example, for the quality con-
the marginal effects of the additionalcustomers on overall struct,respondentsare asked to evaluate quality given price
satisfaction and profitability? and price given quality in two separatequestions. This pro-
1 59
cess results in approximately25,000 observationsper vari- TABLE 2
able (for each year) from which indices are constructed.For- System of Equations Underlying
nell (1992) describes a latent variable approachto estimat- the Conceptual Framework
ing the indices. Lagged Dependent Specification
The SCSB measures are combined with economic re-
turns data for the publicly held firms. Specifically, ROI for EXPt = al + P11EXPt-1+ P12QUALt1 + P13TREND + Elt
each firm (that is, return on assets located in Sweden) is SATt= a2 + P21SATt_1
+ 22QUALt+ J23EXPt+ P24TREND+ e2t
used as a measure of economic returns.Unusual or extraor-
ROIt = a3 + p31ROltI1+ P32SATt+ P33TREND + E3t
dinaryreturnsare treatedas outliers.To make the fullest pos-
sible use of the available data, missing values are treatedas
First Differences Specification
having the same correlationas the values presentin the data
set. In other words, the distributions of the variables are AEXPt = P12AQUALt_1+ P13TREND + elt
treated as censored and a covariance matrix is created as a
basis for estimation. ASATt = P22AQUALt+ P23AEXPt+ P24TREND + E2t
Clearly, there are difficulties in combining the data sets. AROIt= p32ASATt+ P33trend + E3t
For example, the ROI data are for a business as a whole
ratherthan a specific product measured by the SCSB. Al-
though this is not a serious issue for the retail and service reflects the expected persistenceof the benefits of customer
sectors, it is a concern for firms with more diverse product satisfactionfor the firm (consistentwith the overallor cumu-
lines, such as the automobiles. Although ROI is commonly lative natureof satisfactionfocused on in this study). This
used in studies of the impact of strategicvariables,it is not specification is also consistent with the argumentthat the
an ideal measure of economic returns.Capital marketdata marketplacehas adaptive expectations. Finally, it fits with
(stock prices) would have been anotherinterestingmeasure the intuitive notion of RicardianRents resulting from high
if a large portion of the SCSB firms were actively tradedin customer satisfaction (Montgomery and Wernerfelt1988).
Sweden or transactedmost of their business there. Accordingly,the endogenousvariablein each equationis re-
gressed on its lagged quality and a set of independentvaria-
Testing the Hypotheses bles capturingthe appropriateeffects.2 In view of the exis-
The system of equations to be estimated is presented in tence of simultaneityand expected correlationbetween the
Table 2. In keeping with the argumentsadvancedin the pre- errorsof the equations, three-stageleast squaresis used to
vious section, expectations are influenced by past quality, estimate the model.
customer satisfaction is influenced by both quality and ex- It is worth noting that a common-and conservative-
pectations, and economic returnsare influencedby satisfac- correctionfor controlling for heterogeneityand unobserva-
tion. Obviously, there are other variables besides customer bles in short cross-sectionaltime-series data is to transform
satisfaction that affect economic returns. The effects of the datathroughfirstdifferences(Maddala1977). (This spec-
these variables are captured in the lag structure,the error ification restricts 1ll = 321 = 331 = -1.) It should be
term, and a trend term. If the marketplacehas adaptiveex- pointed out that this specification is more consistent with a
pectations,then we should expect the coefficient for the im- transaction-specificconceptualizationof customer satisfac-
pact of past quality QUALn_1on expectations EXPt to be tion. It implies thatshort-termchanges in qualityand expec-
positive 1 > P12 > 0. (To test the adaptive expectations tations have immediateratherthan long-termconsequences
model, we restrictthe coefficients such that P312= 1 - P311.) for customer satisfaction and ultimately profitability. We
For customer satisfaction SATt we expect the impact of therefore expect expectations to have a negative effect on
both currentquality QUALt and EXPt to be positive, 122 > customer satisfactionin this specification.
0 and 123 > 0. The effect of currentqualityon customersatis-
faction should be greater than that of expectations, 322>
323. The impact of SATton profitabilityas measuredby re-
Results
turnon assets ROItis expected to be positive, 132 > 0. This Table 3 presents three-stageleast squaresestimates for the
latterrelationshipis predictive in that the survey measuring two specifications. The findings generally confirm the pat-
customersatisfactionis conductedin the first half of the fis- tern of effects as hypothesized.Let us first discuss the find-
cal year and economic returnsare based on year-endfinan- ings relating quality and expectations to satisfaction and
cial reporting.As logarithmsare taken of each variable,the then turn our attention to the effect on economic returns.
estimated coefficients are interpretableas elasticities. With regardto the first equationof each specification,the co-
efficients supportthe idea of adaptiveexpectations.The rel-
Specification ative size of the coefficients for the impact of past expecta-
To account for heterogeneity in the cross-section of indus- tions EXPt_ and past quality QUALti on currentexpecta-
tries (e.g., differences in accountingpractices,industrialor- tions EXPt is consistent with how one would expect a
ganization considerations) and possible unobservable ef- firm's reputationfor quality to change over time. Although
fects (e.g., firm strategy, pioneering advantage),the system
is formulatedas state dependent(Amemiya 1985; Boulding 2For longer time-series, potential methods of controlling for unob-
servables are the family of error-componentmodels (Amemiya 1985) and
1990; Jacobsen 1990a, b; Maddala 1977). This formulation latent-classpooling methods (Ramasway,Anderson,and DeSarbo 1994).

TABLE 3 A sizeable carryover effect supports the notion that cus-
Empirical Findings tomer satisfaction is indeed cumulative. The implicationis
All coefficients are three-stage least squares estimates. thathigh customersatisfactioninsulatesthe firm from short-
term changes in quality. The strong carryovereffect of past
Lagged Dependent Specification- customer satisfaction also means that it is time-consuming
Weighted R-square is .82 for firms with low customer satisfaction to improve their
EXPt= .01* + .91* EXPt1 + .09* QUALt_,- .003* TREND
standingin the market.
SATt= -.12 + .44* SATt_ + .49* QUALt+ .10* The effect of expectationsof quality on customer satis-
EXPt- .003* TREND faction is positive and significant, as well as relatively
ROlt= -1.10* + .75* ROIt_1+ .40* SATt+ .002 TREND small. For every percentage point change in expectations,
customer satisfaction changes by .10%. This is supportive
First Differences Specification- of the argument for adaptive expectations. Expectations
Weighted R-square is .35 adapt slowly and provide incremental information to that
providedby quality.In particular,in modeling customersat-
AEXPt= .11* AQUALt_,- .011* TREND isfaction as a long-term, dynamic phenomenon, the carry-
ASATt = .72* AQUALt- .50* AEXPt + .000 TREND over effect of past satisfaction naturallycapturesinforma-
AROIt = .76* ASATt - .001 TREND tion about past experience with quality, leaving expecta-
tions with a relatively marginal effect that can be inter-
*indicatesthe coefficientis significantat the .01 level.
preted as the effect of the market'sforecast of futurequal-
ity on currentsatisfaction.
expectations are fundamentallystable, changes in the level It is importantto note that the sign of the impact of ex-
of qualityprovidedby a firm will enhanceor erode the com-
pectations on customer satisfactionis reversed in the first-
pany's reputationfor quality over time. The estimates pre- differences specification (i.e., negative), which implies that
sented in Table 3 suggest that though the marketeventually a short-termincrease in expectationsactually may lead to a
will revise its expectations completely (the long-runelastic- decrease in customer satisfaction. That is, increasing cus-
ity of expectations with respect to changes in past quality is tomer expectations by overpromising is likely to be detri-
restrictedto be one), this will be a slow process for the typ- mental to the firm in the short run, whereas increasingcus-
ical firm. Conversely, there appearsto be considerablemo- tomer expectations throughimproving quality benefits the
mentum for the currentlevel of expectations. The stability firm in the long run.
of expectations suggests that a firm's reputationfor provid- Return on investment, a long-term measure of eco-
ing quality will not change quickly. nomic health, is strongly affected by customer satisfaction.
As seen in the second equation of Table 3, both quality This is true for both specifications.However, the interpreta-
and expectations have a positive impact on customer satis- tion of the two specifications is different. The lagged-
faction.3For the state-dependentor persistenceformulation, dependentvariablespecificationimplies thata changein cus-
these effects are not only in the predicteddirection,but also tomer satisfaction is not reflected all at once in returns.
of the predicted relative size. In fact, the estimates suggest Rather,a percentagepoint change in customer satisfaction
that currentcustomer satisfaction is primarilya function of in one period carriesover to futureperiods, consistent with
(1) currentquality and (2) past satisfaction.Quality has the the cumulative nature of customer satisfaction. The first-
greatest impact on customer satisfaction, accordingto both differences specification, on the other hand, implies that
specifications.The importanceof currentqualityin determin- there is a larger immediate effect from a change in cus-
ing customer satisfaction is consistent with the notion that tomer satisfaction,but that this advantageis short-livedand
currentexperience will be weighted more highly than past unsustainable.Nevertheless,both findings suggest thatpro-
or anticipatedexperience. viding high quality and high customer satisfaction is re-
In the first specification, the size of the effect of lagged wardedby economic returns.Moreover,the log-linearformu-
customer satisfaction indicates a strong carryover effect. lation implies that if the costs of providinghigh quality and
For every percentagepoint change in customer satisfaction customer satisfaction are increasing at an increasing rate,
at t-1, customersatisfactionat t changedby .44%.This sug- then there must be an optimal level of satisfaction. Obvi-
gests that high past satisfaction of currentcustomers pro- ously, then, strategiesthat seek to maximize customersatis-
vides a strong indication that currentand consequently fu- faction are inappropriate.
ture customer satisfactionwill be high. Interestingly,the es- How do these figures comparewith other studies exam-
timate of a carryovereffect of .44 is very nearly identical to ining the impact of marketing mix variables on ROI?
the average carryovereffect for sales, .468, as estimated in Buzzell and Gale (1987) reportan impactcoefficient for rel-
the meta-analysis of Assmus, Farley, and Lehmann(1984). ative quality on ROI of . 1. We can transformthis value
into an average elasticity of ROI with respect to quality by
3Itis worth noting that the methodology here produces similar substan- using their mean values of ROI and quality. This calcula-
tive resultsto other methodsof controllingfor fixed effects (e.g., instrumen- tion yields an average short-runelasticity for ROI with re-
tal variables).The exception is the size of the coefficient for the effect of
customer satisfaction on ROI. This coefficient is significantly larger when
spect to quality of .25. The coefficients in Table 3 can be
instrumentalvariable methods are used (Anderson, Forell, and Lehmann used to compareour findings with this figure. To obtain an
1993). estimate of the short-runimpact of a change in quality on
/ 61
FIGURE 1 ample, if the same coefficients apply to a sample of U.S.
Returns due to Increased Satisfaction firms (e.g., the Business Week1000, with average assets of
One Point Increase In Each Year $7.5 billion and averageROI of 11%),the cumulativeincre-
mental returnsfrom a continuousone-point increasein cus-
Q 3-
tomer satisfaction over a five-year span would be $94 mil-
^3-
0
O lion, or 11.4% of currentROI.
"r
~1 ~
?2 The Value of Current Customer Assets
S - ------- - - .
:3 2-
$
V. -~/
i- i: -* o The precedingempiricalpredictionof the value of customer
O
C.) satisfactioncan be supplementedby an analyticalmodel. If
..............., improving customer satisfactionincreases the likelihood of
C)
Il
.o repurchase,then we can illustratethe economic benefits of
0;n such a change by consideringcurrentcustomersas an asset
ti
o
.?s: :.~3:to
_~ ~3. the firm and calculating their net present value to the
*- firm. A straightforward calculation might capture customer
:;o
,^lan,-
Io<,
assets as a functionof the likelihoodor probabilitythata sat-
Year isfied customer will remain loyal, PR(Loyal|Satisfaction),
the averagegross marginper period G, the length of the av-
erage repurchasecycle X, and a discount factor a. The asso-
ROI, we calculate the elasticities in the chain from quality ciated net present value equation can be written:
to ROI. Here, the short-runelasticity of ROI with respect to
T
quality is .49(.40) = .196. Hence, we find an elasticity of NPV = + a))tX.
ROI with respect to quality comparableto, though slightly hXG(Pr{Loyal|Satisfaction}/(l
t=1
smaller than, that found in the PIMS database.
Empirical Prediction of the Value of a One-Point We assume that there is a monotonic relationship be-
Increase in Customer Satisfaction tween customer satisfaction and repurchaseintentions that
What is the value of an increasein the customersatisfaction is linearfor small changesin satisfaction.Andersonand Sul-
index for the typical Swedish firm represented in the livan (1993) estimatethata .0058 increasein repurchaselike-
SCSB? To illustrate this, let us consider the case of a firm lihood (on a scale from 0 to 1) will result from a one-point
with a five-yearplanninghorizon.Supposethe firm must es- increasein customersatisfaction.Hence, if a firm's satisfac-
timate the increase in ROI resultingfrom increasingits cus- tion index is on average 67 and undergoes an increase to
tomer satisfaction index by a single point in each of the 70, the typical firm's repurchase probabilities would
next five years (cumulativeincreaseof five points). Assum- change from the averageof .75 to .7674. Given the average
ing the firm's ROI in the initial year is the same as the av- gross marginfor the firms in the SCSB ($65 million) and as-
erage for the sample (10.83%), the estimates in Table 3 suming customerspurchasean averageof once per year, the
imply incrementalincreases in ROI for the next five years net present value of customer assets would rise $6.4 mil-
of .07%, .18%, .33%, .51%, and .71%, respectively, over lion, or 5.4%, from $118.8 million to $125.2 million.
what the firm's ROI would have been without increasing
customer satisfaction.The fifth-year ROI of 11.54%repre- Customer Satisfaction and Market Share
sents a 6.59% increase over the original ROI of 10.83%. How are customer satisfaction and market share related?
The five-year cumulative increase of 1.8% (1.8 = .07 + .18 We have been able to obtain 1989-90 company-level mar-
+ .33 + .51 + .71) representscumulativeincrementalreturns
ket sharedata to matchthe customersatisfactionindices for
of 16.66% relative to current ROI (16.66 = 100 [1.8/
a subsample of the SCSB firms. Plots of the raw data and
10.83]). The net present value for the incrementalreturns
can be calculated by assuming that our "typical" firm has year-to-yearchanges in marketshareand customersatisfac-
an asset base correspondingto the sample mean ($600 mil- tion are shown in Figure 2. Both plots suggest downward
lion), a policy of paying out all returnsas dividends,and ap- sloping, that is, inverse relationshipsbetween customersat-
isfaction and marketshare.The plot of raw satisfactionver-
plies a discount rate of 10.00%. As illustratedin Figure 1,
the results of this calculation indicate incrementalreturns sus raw marketshare shows that no firm has both high cus-
over the next five years of $.357 million, $.888 million, tomersatisfactionand high marketshare.Moreover,year-to-
$1.487 million, $2.09 million, and $2.66 million, respec- year increases(decreases)in marketshareare likely to be as-
tively. This represents cumulative discounted returns of sociated with decreases (increases)in customersatisfaction.
$7.48 million, or 11.5% of currentROI. The pearson correlationbetween raw marketshare and sat-
Although the preceding calculations may seem some- isfaction is -.25 (p-value of .03 with n = 77) and the corre-
what modest in absolute size, it should be kept in mind that lation between year-to-yearchanges in the variablesis -.37
the predictionis based on a cross-sectionalanalysis and that (p-value of .05). Regressing changes in the customer satis-
the scale of a typical Swedish firm is much smaller than faction index on changes in market share yields a coeffi-
thatfound in an economy such as the United States'. For ex- cient of -.88 (p-value of .05).
62 1 Journalof Marketing,
July 1994
FIGURE 2
Market Share and Satisfaction
1989 and 1990 ChangesFrom 1989 to 1990

100 10-
.o 90
0 " 5-
A A
80 ID A
S
A6 a A.
C, A
| 70
70 0 A A
c,o
AAA
'o -5 -
60
14.
I
50 t -10 + I I i I I I I I
2
0 10 20 30 40 50 -3 -2 -1 0 1 2 3
MarketShare Change in MarketShare
Figure 2 provides a preliminary indication, similar to the benefits indicated by these findings in reaching their
Griffin and Hauser (1993), that increasingmarketshare ac- decisions.
tually may decrease customer satisfaction. This may indi- Ourfindingsalso indicatethateconomic returnsfrom im-
cate that a more differentiatedstrategycan lead to decreases proving customer satisfactionare not immediatelyrealized.
in marketshare. In addition, it may indicate that, at least in Because efforts to increase currentcustomers' satisfaction
short-run cross-sectional analyses, customer satisfaction primarilyaffect futurepurchasingbehavior,the greaterpor-
and market share are not always compatible goals. tion of any economic returnsfrom improvingcustomersat-
isfactionalso will be realizedin subsequentperiods.This im-
plies that a long-runperspectiveis necessary for evaluating
Summary and Conclusions the efficacy of efforts to improve quality and customer
The widespreadbelief in the intuitive relationshipbetween satisfaction.
quality, customer satisfaction, and economic returns, as The long-run nature of the economic returnsfrom im-
well as the growing frustration with attempts to improve
provingcustomersatisfactionalso has broadstrategicimpli-
quality, serve to underscore the importance of analytical cations. If increasingcustomersatisfactionprimarilyaffects
and empirical work increasing our understanding of cus- future cash flows, then resources allocated to improving
tomer satisfaction and how it relates to economic returns.
qualityand customersatisfactionshould be treatedas invest-
The frustrationof many firms engaged in attempts to im- ments ratherthan expenses. Loyal and satisfied customers
prove quality may be due to any number of factors, from are a revenue-generatingasset to the firm thatis not without
poor marketdata to the intransigenceof functional silos or cost to acquire, retain, and develop. This is very different
fixation with short-termresults that may leave firms unable from viewing sales as a set of more or less disjoint and mu-
to wait for the benefits of investing in quality and customer tually exclusive transactions. Implementing a customer-
satisfaction to materialize (Ettlie and Johnson 1994). Al- asset orientation means aligning the firm's processes, re-
though we do not provideguidancefor managersseeking ei- sources, performance measures, and organizational struc-
ther tools for improving quality (e.g., TQM) or guidelines ture for treatingcustomersas an asset. Ourfindings provide
for implementingquality programs,it does provide motiva- a rationalefor firms to move in this direction.Once the po-
tion for continuing their efforts and overcoming any imped- tential of a customer-asset orientation is acknowledged,
iments encountered:Firms that actually achieve high cus- there are two key proceduralquestions for management:(1)
tomer satisfactionalso enjoy superioreconomic returns.An How do we measure the value of this asset? and (2) How
annualone-point increase in customer satisfactionhas a net do we increase its value? Answers to both these questions
present value of $7.48 million over five years for a typical are now being developed (e.g., Fornell 1991a, b; 1994).
firm in Sweden. Given the sample's average net income of Our findings also provide a preliminary indication of
$65 million, this representsa cumulativeincrease of 11.5%. trade-offs between customer satisfaction and market share
If the impactof customersatisfactionon profitabilityis sim- goals. We find that customer satisfaction actually may fall
ilar for firms in the Business Week 1000, then an annual as marketshareincreases.For example,whereasa small mar-
one-point increase in the average firm's satisfaction index ket-sharefirm may serve a niche marketquite well, a large
would be worth $94 million or 11.4% of current ROI. market-sharefirm often must serve a more diverse and het-
Firms considering implementing or, in an increasing num- erogeneous set of customers. Gains in market share may
ber of cases, curtailing quality programs should consider come from attractingcustomers with preferencesmore dis-
/ 63
tant from the targetmarket.The firm may overextendits ser- tion of its customers.In the context of cumulativecustomer
vice capabilities as the number of customers and/or seg- satisfaction,the long-runeffects of increased(decreased)ex-
ments grows. In such a situation, even though the overall pectationsshould outweigh the short-termeffect of any tem-
level of customer satisfaction is falling, a firm's sales and porarygaps and lead to a rise (fall) in overall customer sat-
profits may be increasing. It is worth noting that this may isfaction. This firm-level finding is consistent with individ-
be a short run versus long runphenomenon.In the long run, ual-level researchshowing that disconfirmationof expecta-
it is possible that customersatisfactionand marketsharego tions has a weaker effect on cumulative customer satisfac-
together, but there is growing evidence that this is not al- tion than the direct impact of perceived quality (Anderson
ways the case in the short run or a cross-sectional analysis. and Sullivan 1993).
When quality and expectations increase, there is a posi- Finally, our findings indicate that, in the aggregate,cus-
tive effect on customer satisfaction in the long run, but in- tomers have adaptive but largely rational expectations.
creased expectations may have a negative impact in the Changes in the level of quality providedby a firm enhance
short run. The large, positive impact of quality on customer or erode a firm's reputationfor quality over time. This is an
satisfaction is intuitive. Expectationshave a positive effect importantprocess to manage for the typical firm because
on customer satisfaction in the long run because they cap- subsequent changes in its reputationfor providing quality
ture the accumulatedmemory of the marketconcerning all may not be immediate.The implicationfor a firm trying to
past quality informationand experience, as well as the mar- make a quality "turnaround"or "comeback" is, therefore,
ket's forecastof the firm's ability to deliver qualityin the fu- not to expect immediate returns but coordinate product/
ture.This forward-lookingcomponentof expectationsis im- service improvements with efforts to accelerate the diffu-
portantbecause this, in part, is how a firm's reputationfor sion of informationregardingsuch improvementsthrough
providinghigh or low qualityinfluences the overall satisfac- the marketplace.
REFERENCES
Aaker,DavidA. (1992), "TheValueof BrandEquity,"Journal fects and Business Performance:Do Fixed Effects Really Mat-
of Business Strategy, 13 (July/August),27-32. ter?"' MarketingScience, 9 (Winter),88-91.
Amemiya, Takeshi (1985), Advanced Econometrics. Cambridge, , Richard Staelin, Ajay Kalra, and Valerie Zeithaml (1993),
MA:HarvardUniversityPress. "A Dynamic Process Model of Service Quality:From Expec-
AmericanQualityFoundation(1992), Best PracticesReport.Cleve- tations to Behavioral Intentions," Journal of Marketing Re-
land: Ernst & Young. search, 30 (February),7-27.
Anderson, Erin and Barton Weitz (1989), "Determinantsof Con- Buzzell, Robert D. (1990), "CommentaryOn 'UnobservableEf-
tinuity in ConventionalIndustrialChannelDyads," Marketing fects and Business Performance,"'MarketingScience, 9 (Win-
Science, 8 (Fall), 310-23. ter), 86-87.
Anderson,Eugene W. (1994a), "Cross-CategoryVariationin Cus- and Bradley T. Gale (1987), The PIMS Principles. New
tomer Satisfactionand RepurchaseLikelihood,"MarketingLet- York:The Free Press.
ters, forthcoming. and Frederick Wiersema (1981a), "Modelling Changes In
_ (1994b), "Post-PurchaseEvaluations and Word of Mouth," MarketShare:A Cross-SectionalAnalysis," StrategicManage-
working paper, National Quality Research Center, University mentJournal,2 (January-March),
27-42.
of Michigan Business School. andFrederickWiersema(1981b),"SuccessfulShare-Build-
, Claes Fornell, and Donald R. Lehmann(1993), "Economic ing Strategies," HarvardBusiness Review, 59 (January-Febru-
Consequences of Providing Quality and Customer Satisfac- ary), 135-44.
tion," WorkingPaper#93-112. Cambridge,MA: MarketingSci- Cadotte, Ernest R., Robert B. Woodruff, and Roger L. Jenkins
ence Institute. (1987), "Expectationsand Norms in Models of ConsumerSat-
_ and Mary Sullivan (1993), "The Antecedents and Conse- isfaction," Journal of MarketingResearch, 24 (August), 305-
quences of Customer Satisfaction For Firms," MarketingSci- 14.
ence, 12 (Spring), 125-43. Capon,Noel, John U. Farley, and Scott Hoenig (1990), "Determi-
Anterasian,Cathy and Lynn W. Phillips (1989), "Discontinuities, nants of Financial Performance:A Meta-Analysis," Manage-
Value Delivery, and the Share-Ret," WorkingPaper #92-121. mentScience,36 (October),1143-59.
Cambridge,MA: MarketingScience Institute. Churchill,GilbertA. andCarolSuprenant
(1982), "An Investiga-
Assmus, Gert, John U. Farley, and Donald R. Lehmann (1984), tion Into The DeterminantsOf Customer Satisfaction," Jour-
"How AdvertisingAffects Sales: Meta-Analysisof Economet- nal of MarketingResearch, 19 (November), 491-504.
ric Results," Journal of Marketing Research, 21 (February), Cronin,J. Joseph,Jr. andStevenA. Taylor,"MeasuringService
65-74.
Quality:A Reexamination andExtension,"Journalof Market-
Bateson, John E. (1989), Managing Services Marketing.London: ing, 56 (July),55-68.
DrydenPress. Crosby, P.B. (1979), Quality is Free. New York:McGraw-Hill.
Bearden,William 0. and Jesse E. Teel (1983), "Selected Determi- Czepiel, JohnA. and RobertGilmore (1987), "Exploringthe Con-
nantsof ConsumerSatisfactionand ComplaintReports," Jour-
cept of Loyalty in Services," in The Services Challenge: Inte-
nal of MarketingResearch, 20 (February),21-28.
gratingfor CompetitiveAdvantage,John A. Czepiel, Carole A.
Becker,GaryS. (1965),"A Theoryof theAllocationof Time,"Ec- Congram,and James Shanahan,eds. Chicago: AmericanMar-
onomic Journal, 75 (September),493-517. keting Association, 91-94.
Bolton,RuthN. andJamesH. Drew(1991),"A MultistageModel Darby, Michael R. and Edi Kami (1973), "Free Competitionand
of Customers'Assessmentsof ServiceQualityand Value," the OptimalAmount of Fraud," Journal of Law and Econom-
Journal of ConsumerResearch, 17 (March),375-84. ics, 16 (April),67-88.
Boulding, William (1990), "CommentaryOn 'UnobservableEf- Deighton,John(1984), "TheInteraction
of AdvertisingandEvi-

dence," Journal of ConsumerResearch, 11 (December), 763- search Center,University of Michigan Business School.
70. andClaes Forell (1991), "A FrameworkFor ComparingCus-
Dodds, William B., Kent B. Monroe, and Dhruv Grewal (1991), tomer Satisfaction Across Individuals and Product Catego-
"Effects of Price, Brand, and Store Information on Buyers' ries," Journal of Economic Psychology, 12 (2), 267-86.
ProductEvaluations," Journal of MarketingResearch, 28 (No- Juran, J.M. (1988), Juran's Quality Control Handbook, 4th ed.
vember), 307-19. New York:McGraw Hill.
The Economist (1992), "The Cracks in Quality" (April 18), 67- Katona,George (1979), "Towarda Macropsychology,"American
68. Psychologist, 34, 118-26.
Ettlie, John and Michael D. Johnson(1994), "Impedimentsto Suc- Keller, Kevin L. (1993), "Conceptualizing,Measuring,and Man-
cessful Implementationof Quality Programs,"MarketingLet- aging Brand Equity," Journal of Marketing,57 (January),1-
ters, forthcoming. 22.
Fornell, Claes (1991a), "National and CorporateCustomerSatis- Kotler, Philip and Sidney J. Levy (1969), "Broadening the Con-
faction Indices," paper presented at the World Quality Day, cept of Marketing,"Journal of Marketing,33 (February),10-
World TradeCenter, Amsterdam,The Netherlands. 15.
(1991b), "The Challenges of a CustomerOrientation,"in In- LaBarbera,Priscilla A. and David Mazursky(1983), "A Longitu-
novation,Integration& Strategic Processes: A ManagerialAp- dinal Assessment of Consumer Satisfaction/Dissatisfaction:
proach, S. Barlinn, K. Haanes, P.S. Bronn, and P. Lorange, The Dynamic Aspect of the Cognitive Process," Journal of
eds. Oslo: BedriftsokonomensForlag, 33-44. MarketingResearch, 20 (November), 393-404.
(1992), "A National Customer Satisfaction Barometer:The Lancaster, Kelvin (1979), Variety, Equity, and Efficiency. New
Swedish Experience," Journal of Marketing,55 (January),1- York:Columbia University Press.
21. Lovelock, ChristopherH. (1984), Services Marketing.Englewood
(1994), "Productivity,Quality and CustomerSatisfaction as Cliffs, NJ: Prentice-Hall,Inc.
Strategic Success Indicatorsat Firm and National Levels," in Lucas, RobertE., Jr. (1973), "Some InternationalEvidenceof Out-
Integral Strategy, H.B. Thorelli, ed. forthcoming. put-Inflation Tradeoffs," American Economic Review, 63
and Birger Wernerfelt(1987), "Defensive MarketingStrat- (July), 326-34.
egy by CustomerComplaintManagement:A TheoreticalAnal- Maddala,George S. (1977), Econometrics.New York: McGraw-
ysis," Journal of Marketing Research, 24 (November), 337- Hill.
46. Montgomery,Cynthia A. and Birger Wererfelt (1988), "Diversi-
and (1988), "Model For CustomerComplaintManage- fication,RicardianRents, and Tobin'sq," RANDJournalof Ec-
ment," MarketingScience, 7 (Summer), 271-86. onomics, 19 (Winter),623-32.
Garvin, David A. (1988), Managing Quality: The Strategic and Montgomery, David B. (1975), "New Product Distribution: An
CompetitiveEdge. New York:The Free Press. Analysis of SupermarketBuyerDecisions," Journal of Market-
Griffin, Abbie and John R. Hauser(1993), "The Voice of the Cus- ing Research, 12 (July), 255-64.
tomer," MarketingScience, 12 (Winter), 1-27. Morrison,Donald G. (1979), "Purchase Intentionsand Purchase
Gronroos,Christian(1990), Service Managementand Marketing: Behavior," Journal of Marketing,43 (Spring), 65-74.
Managing the Moments of Truthin Service Competition.Lex- Muth, John F. (1961), "Rational Expectationsand the Theory of
ington, MA: Lexington Books. Price Movements," Econometrica,29 (July), 315-35.
Hauser, John R. and Steven M. Shugan (1983), "Defensive Mar- Nelson, Phillip (1970), "Informationand Consumer Behavior,"
keting Strategies," MarketingScience, 2 (Fall), 319-60. Journal of Political Economy, 82 (July), 311-29.
_ , Duncan I. Simester, and Birger Wererfelt (1993), "Cus- Newsweek (1992), "The Cost of Quality" (September7), 48.
tomer-SatisfactionBased Incentive Systems," working paper, Oliver, RichardL. (1977), "Effects of Expectationsand Disconfir-
InternationalCenter for Researchon the Managementof Tech- mation on PostexposureProductEvaluations," Journal of Ap-
nology, Sloan School of Management,MassachussettsInstitute plied Psychology, 62 (April), 246-50.
of Technology. (1980), "A Cognitive Model of the Antecedentsand Conse-
Helsen, Harry (1964), Adaptation-Level Theory. New York: quences of SatisfactionDecisions," Journal of MarketingRe-
Harper& Row. search, 17 (November), 460-69.
Hoch, Stephen J. and Young-WonHa (1984), "ConsumerLearn- (1989), "Processingof the SatisfactionResponsein Consump-
ing: Advertising and the Ambiguity of Product Experience," tion: A Suggested Framework and Research Propositions,"
Journal of ConsumerResearch, 13 (September),221-33. Journal of ConsumerSatisfaction, Dissatisfaction, and Com-
Holbrook, MorrisB. (1994), "The Natureof CustomerValue: An plaining Behavior, 2, 1-16.
Axiology of Services in the ConsumptionExperience," in Ser- (1993), "A Conceptual Model of Service Quality and Ser-
vice Quality: New Directions in Theoryand Practice, Roland vice Satisfaction,"in Advancesin ServicesMarketingand Man-
T Rust and RichardL. Oliver, eds. ThousandOaks, CA: Sage agement, TeresaA. Swartz, David E. Bowen, and Stephen W.
Publications,Inc., 21-71. Brown, eds. Greenwich,CT: JAI Press, 65-86.
Howard, John A. and Jagdesh N. Sheth (1969), The Theory of and Wayne DeSarbo (1988), "Response Determinants In
Buyer Behavior. New York:John Wiley & Sons, Inc. Satisfaction Judgments," Journal of Consumer Research, 14
Hunt,H. Keith (1977), "CS/D-Overview and FutureResearchDi- (March),495-507.
rections," in Conceptualizationand Measurementof Customer and JohnE. Swan (1989), "ConsumerPerceptionsof Interper-
Satisfaction and Dissatisfaction, H. Keith Hunt, ed. Cambr- sonal Equity and Satisfaction in Transactions:A Field Survey
idge, MA: MarketingScience Institute,455-88. Approach," Journal of Marketing,53 (April), 21-35.
Jacobson,RobertD. (1990a), "UnobservableEffects and Business and Russel S. Winer (1987), "A Frameworkfor the Forma-
Performance,"MarketingScience, 9 (Winter),74-85. tion and Structureof CustomerExpectations:Review and Prop-
(1990b), "UnobservableEffects and Business Performance, ositions," Journal of Economic Psychology, 8, 469-99.
Reply to Comments of Boulding and Buzzell," MarketingSci- Parasuraman, A., Valerie A. Zeithaml, and Leonard L. Berry
ence, 9 (Winter),92-95. (1985), "SERVQUAL: A Multiple-ItemScale for Measuring
Johnson, Michael D., Eugene W. Anderson, and Claes Fornell Consumer Perceptions of Service," Journal of Retailing, 64
(1994), "Rationaland AdaptiveExpectationsin a CustomerSat- (Spring), 12-40.
isfaction Framework," working paper, National Quality Re- Phillips, Lynn W., Dae Chang, and Robert D. Buzzell (1983),
/ 65
"ProductQuality, Cost Position, and Business Performance:A Shostack,G. Lynn (1977), "BreakingFree From ProductMarket-
Test of Some Key Hypotheses," Journal of Marketing, 47 ing," Journal of Marketing,41 (April), 73-80.
(Spring), 26-43. Shugan, Steven M. (1989), "The Haagen-DazsDilemma: Product
Porter, Michael E. (1980), Competitive Strategy. New York: Assortmentin a Triopoly," ManagementScience, 35 (March),
MacMillan/FreePress. 304-20.
Ramaswamy,Venkatram,Eugene W. Anderson,and WayneS. De- Steenkamp, Jan-Benedict E.M. (1989), Product Quality. Assen/
Sarbo (1994), "A Disaggregate Negative Binomial Regression Maastricht,The Netherlands:Van Gorcum.
Procedure for Count Data Analysis," Management Science, Szymanski, David M., SundarG. Bharadwaj,and P. Rajan Vara-
forthcoming. darajan,"An Analysis of the MarketShare-ProfitabilityRela-
Reichheld, FredrickF. and W. Earl Sasser (1990), "Zero Defec- tionship," Journal of Marketing,57 (July), 1-18.
tions: Quality Comes to Services," Harvard Business Review, TARP/TechnicalAssistance Research Program(1979), Consumer
68 (September/October),105-11. ComplaintHandling In America: An Update Study. Washing-
Ricardo, David (1817), Principles of Political Economy, 3rd ed. ton, DC: White House Office of ConsumerAffairs.
London:John Murray. (1981), Measuring the Grapevine: ConsumerResponse and
Robertson,Thomas S. and HubertGatignon(1986), "Competitive Word-of-Mouth.Atlanta, GA: The Coca-Cola Co.
Effects on Technology Diffusion," Journal of Marketing, 50 Taylor, John (1979), "Staggered Wage Setting in A Macro
(July), 233-55. Model," AmericanEconomic Review, 63 (May), 108-18.
Rosen, Sherwin (1974), "Hedonic Prices and Implicit Markets: Tse, David K. and Peter C. Wilton (1988), "Models of Consumer
ProductDifferentiationin PureCompetition,"Journal of Polit- SatisfactionFormation:An Extension," Journal of Marketing,
ical Economy, 82 (January/February), 34-55. 25 (May), 204-12.
Rust, Roland and Anthony Zahorik (1993), "Customer Satisfac- Yi, Youjae (1991), "A Critical Review of Customer Satisfac-
tion, CustomerRetention,and MarketShare,"Journalof Retail- tion," in Review of Marketing1989, Valarie A. Zeithmal, ed.
ing, 69 (Summer), 145-56. Chicago: AmericanMarketingAssociation.
Sawyer, Alan G. and Peter Dickson (1984), "Psychological Per- Zeithaml, Valarie A. (1981), "How Consumer Evaluation Pro-
spectives on ConsumerResponse to Sales Promotion," in Re- cesses Differ Between Goods and Services," in Marketingof
search on Sales Promotion: Collected Papers, KatherineJocz, Services, James H. Donnelly and William R. George, eds. Chi-
ed. Cambridge,MA: MarketingScience Institute. cago: AMA, 186-89.
Schmalansee,Richard(1978), "A Model of Advertisingand Prod- (1988), "ConsumerPerceptionsof Price, Quality, and Value:
uct Quality," Journal of Political Economy, 86 (September), A Means-EndModel and Synthesis of Evidence," Journal of
485-503. Marketing,52 (July), 2-22.
66 1 Journalof Marketing,
July 1994

Customer SatisfactioMna, R Ket Share, and Profitability: Findings From Sweden

Uploaded by

Copyright:

Available Formats

Customer SatisfactioMna, R Ket Share, and Profitability: Findings From Sweden

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Customer SatisfactioMna, R Ket Share, and Profitability: Findings From Sweden

Uploaded by

Copyright:

Available Formats

Customer Satisfaction, Market Share, and Profitability: Findings from Sweden

Author(s): Eugene W. Anderson, Claes Fornell, Donald R. Lehmann

Does higher customer satisfactionlead to superioreco- companiesare wastingtheireffortsin tryingto improvequal-

54 / Journalof Marketing,July 1994

56 / Journalof Marketing,July 1994

58 / Journalof Marketing,July 1994

60 / Journalof Marketing,July 1994

1989 and 1990 ChangesFrom 1989 to 1990

64 / Journalof Marketing,July 1994

You might also like