Reversing the Logic: The Path to Profitability through Relationship Marketing

a

A

aeiicaats©

K

tfcmdasss

tdm

i(

0h

Journal of Retailing 89 (3, 2013) 246–262

Revisiting the Satisfaction–Loyalty Relationship: Empirical Generalizationsand Directions for Future Research�

V. Kumar a,∗, Ilaria Dalla Pozza a,b, Jaishankar Ganesh c

Center for Excellence in Brand & Customer Management, J. Mack Robinson School of Business, Georgia State University, Atlanta, GA 30303-3989, United Statesb IPAG Business School, Paris, France

c School of Business, Rutgers University, Camden, NJ, United States

bstract

This extensive literature review highlights the state of the art regarding the relationship between customer satisfaction and loyalty, both attitudinalnd behavioral. In particular, it brings to light several issues that should be carefully considered in analyzing the efficacy of customer satisfaction inxplaining and predicting customer loyalty. In fact, for many years companies all around the world have heavily invested in customer satisfactionn the hope of increasing loyalty, and hence, consequently, profitability. But after having gone through a detailed analysis, it is clear that this linkt is not as strong as it is believed to be and customer satisfaction is not enough to explain loyalty. In fact, the major findings of this review areaptured in the form of a few empirical generalizations. We generalize that, while there is a positive relationship between customer satisfactionnd loyalty, the variance explained by just satisfaction is rather small. Models that encompass other relevant variables as moderators, mediators,
ntecedent variables, or all three are better predictors of loyalty than just customer satisfaction. Further, the satisfaction–loyalty relationship hashe potential to change over time. Similar weaker findings are uncovered and the study offers specific guidelines on who, when, and how much toatisfy. Finally, suggestions for future research to explore this domain are offered.
2013 New York University. Published by Elsevier Inc. All rights reserved.

eywords: Customer satisfaction; Loyalty; Word-of-Mouth; Customer lifetime value; Retention; Generalizations

an2(ttdon

Introduction

While having a satisfied customer base is a laudable goalhat is not to be questioned, its impact on loyalty and per-ormance outcomes is not as obvious. In reality, the questiononcerning the efficacy of the satisfaction–loyalty link isuch more nuanced than if a simple yes, it exists, or no, it

oesn’t. Researchers (Kamakura et al. 2002; Rust, Zahorik,nd Keiningham 1995) have for long suggested that companies
hould not blindly follow the path of only focusing on customeratisfaction in the hope of improving loyalty. Specifically, thesetudies have pointed out the necessity of considering the cost of
� We would like to thank the seminar participants at various universities inhe U.S., France, and Italy and Yashoda Bhagwat for their valuable suggestionsuring the preparation of this manuscript. We thank Renu for copyediting theanuscript.∗ Corresponding author. Tel.: +1 404 413 7590; fax: +1 832 201 8213.

E-mail addresses: [email protected], dr [email protected] (V. Kumar),laria.dalla [email protected] (I.D. Pozza), [email protected]. Ganesh).

t

n(vnmssas(

022-4359/$ – see front matter © 2013 New York University. Published by Elsevier Ittp://dx.doi.org/10.1016/j.jretai.2013.02.001

customer satisfaction improvement when deciding whether orot to make customer satisfaction investments (Kamakura et al.002). A meta-analysis conducted by Szymanski and Henard2001) finds that satisfaction explains less than 25 percent ofhe variance in repeat purchase. More precisely, the associa-ion between customer satisfaction and loyalty is highly variableepending on the industry, customer segment studied, the naturef the dependent and independent variables, and the presence ofumerous factors that serve as mediators, moderators, or botho the relationship.

For instance, while several studies report of a positive sig-ificant relationship between satisfaction and loyalty, Verhoef2003), examining the effect of satisfaction along with otherariables on defection and customer share development, foundo significant direct effect for satisfaction. Only affective com-itment and loyalty program membership were found to have a

ignificant positive direct effect on customer retention. However,
atisfaction comes into play when moderated by relationshipge. Results also vary according to the way loyalty is mea-ured (intentions vs. actual behavior). For instance, Seiders et al.2005) find that customer satisfaction has a strong positive effect
nc. All rights reserved.

dx.doi.org/10.1016/j.jretai.2013.02.001

mailto:[email protected]




dx.doi.org/10.1016/j.jretai.2013.02.001

Retai

oc

iaascififictdi(i

rgoat(eat(msntinecibih2i2erlct

trfutop

gati–ababbr

htpdIwltscttlodocscccic

srbFailpotpfbwctv

V. Kumar et al. / Journal of

n repurchase intentions, but found no direct effects on repur-hase behavior.

Researchers have also indicated the presence of moderatorsn the satisfaction–loyalty relationship. For instance, Homburgnd Giering (2001) in linking customer satisfaction to loy-lty recognized that the link is not universally strong for allegments. The authors find significant moderating effects ofustomer characteristics: among them, age, variety seeking, andncome seem important variables. Mittal and Kamakura (2001)nd that in the automotive industry satisfaction ratings are higheror women than men. Others have pointed out that satisfactions not the main driver of loyalty. Agustin and Singh (2005)onducted their study in the retail clothing and airline indus-ries and found that relational trust and value are the strongesteterminants of loyalty intentions, rather than customer sat-sfaction. Similarly, Ngobo (1999) and Anderson and Mittal2000) found variability of the satisfaction–loyalty link acrossndustries.

Deftly summarizing more than two decades of academicesearch on this issue, Mittal and Frennea (2010) offer strate-ic insights and critical guidelines to managers that, amongther things, identify the differences across customer groupsnd segments and the varying impact of customer satisfac-ion on behavior across industries. While Mittal and Frennea2010) do point out the presence of customer segment differ-nces, they do not systematically address moderators, mediators,nd other predictors of loyalty that could potentially reducehe relevance of customer satisfaction. Luo and Homburg2007) on the other hand explore the moderating impact ofarket concentration on the relationships between customer

atisfaction and future advertising and promotion effective-ess as well as a firm’s human capital performance. Whilehey state that satisfaction increases customer loyalty andnfluences future purchase intentions and behaviors they doot directly examine this relationship. They do not providempirical generalizations regarding the relationship betweenustomer satisfaction and loyalty. Despite a plethora of stud-es examining the impact of satisfaction on a firm’s customerase in multiple contexts using other moderating and mediat-ng variables (Biong 1993 – B2B; Bowen and Chen 2001 –ospitality; Keh and Lee 2006 – services; Vesel and Zabkar009 – DIY programs; Söderlund 2002 – prepurchase famil-arity; Suh and Yi 2006 – product involvement; Yi and La004 – expectations) there still exists a void in terms of gen-ralizable empirical findings (Garbarino and Johnson 1999)elating the various attitudinal and behavioral measures ofoyalty and the role of customer, relational, and marketplaceharacteristics in understanding the satisfaction–loyalty rela-ionship.

Helping to fill this void, Gupta and Zeithaml (2006) iden-ify and develop empirical generalizations on three links: theelationship between unobservable metrics (customer satis-action) and financial performance, the relationship between
nobservable constructs and observable constructs (satisfac-ion and retention), and the impact of observable constructsn financial performance (relationship between retention androfitability). Gupta and Zeithaml (2006) develop empirical
pamf

ling 89 (3, 2013) 246–262 247

eneralizations by considering eleven studies expressing loy-lty as observable actual behavior (retention or repurchase ratherhan repurchase intentions). The focus of our study is on exam-ning the relationship between customer satisfaction and loyalty

using both attitudinal and behavioral measures. In Guptand Zeithaml’s (2006) words, we focus both on relationshipsetween perceptual customer metrics (customer satisfaction andttitudinal loyalty) and on relationships between unobserva-le metrics and behavioral metrics (customer satisfaction andehavioral loyalty) in order to provide a more comprehensiveeview.

The primary objective of this study is to provide a compre-ensive review and draw empirical generalizations addressinghese critical issues that impact the satisfaction–loyalty link. Inarticular, this study examines the following questions: Whato we really know about the customer satisfaction–loyalty link?s customer satisfaction a good predictor of loyalty? Is it reallyorth investing in customer satisfaction in an effort to improve

oyalty? The generalizations are based on studies that span mul-iple retail and service sectors including banking and financialervices, hospitality, insurance, pharmaceuticals, telecommuni-ations, automotive, and retail grocery. Our conclusion is thathe customer satisfaction–loyalty main effect is indeed weak andhat customer satisfaction, by itself, can hardly change customeroyalty in a significant way. In fact, the systematic presencef moderators, mediators, and other predictors of loyalty intro-uce a high variability in the findings, thus reducing the rolef satisfaction. So, does it really make sense for companies toontinue to adopt the conventional paradigm? In a resource con-trained environment, should companies continue to invest inustomer satisfaction in the traditional sense, in the hope thatustomer loyalty and profits will follow? Should companiesontinue to look at the link between satisfaction and loyaltyn isolation or should they examine the relationship in a broaderontext?

The next section presents a literature review on the relation-hip between customer satisfaction and loyalty. The literatureeview and the associated analysis of the empirical findings wille conducted separately for attitudinal and behavioral loyalty.irst, we will look at the direct relationship between satisfactionnd loyalty (direction, shape, variance explained). Then, we willnvestigate the moderators, mediators, and other predictors ofoyalty, after controlling for the effect of satisfaction. Based onast research findings, we draw empirical generalizations thatffer consistent explanations to these complex relationships. Inhe final section we examine research addressing the broaderhenomenon of customer-oriented strategy and customers dif-erences in terms of the value they bring to the firm as measuredy the lifetime value (Gupta et al. 2006) and draw insights onho to satisfy and how much and when to satisfy. For instance,

ompanies should be engaged in proactive strategies that enablehem to target their resources first toward satisfying the highalue customers while minimizing investments targeted at non-
rofitable or less profitable customers, thus bringing profitabilitynd a stronger focus on costs to bear at the outset of the decision-aking process. We conclude by highlighting directions for
uture research.

2 Retai

A

ocaACGH2aa“sip

hmtteofG

(sacc

tHieelfi

wrstepaeit

msmiatmSt(qorim

48 V. Kumar et al. / Journal of

Customer satisfaction–loyalty relationship: literaturereview and generalizations

ttitudinal loyalty measured as intention

Attitudinal loyalty can be expressed as the likelihood to rec-mmend, the likelihood to repurchase, or depending on theontext, the likelihood to visit/repurchase from the retailergain (Agustin and Singh 2005; Anderson and Mittal 2000;nderson and Sullivan 1993; Bloemer and de Ruyter 1998;handrashekaran et al. 2007; Cronin, Michael, and Hult 2000;ustafsson and Johnson 2004; Homburg and Furst 2005;omburg and Giering 2001; Johnson, Herrmann, and Huber006; LaBarbera and Mazursky 1983; Lam et al. 2004; Liangnd Wang 2004; Mittal, Kumar, and Tsiros 1999; Mittal, Ross,nd Baldasare 1998; Ngobo 1999; Seiders et al. 2005). Theselikelihoods” are measured as intentions based on self-reportedurveys. Literature is replete with research addressing the sat-sfaction and attitudinal loyalty relationship. Fig. 1 and Table 1resent a summary of the results.

The studies presented in Table 1 are organized according toow the constructs were measured, that is, using single item orulti-item scales. For each study, Table 1 indicates the direc-

ion of the relationship (positive, negative or not significant),he R2 and the shape of the relationship where reported (lin-ar, concave or convex, asymmetric nonlinear, with increasingr decreasing returns). Based on these results, we identify theollowing generalizations (Bass and Wind 1995):

1: Overall, there is a positive relationship between customer satisfactionand loyalty intentions.

It is important to note that while Szymanski and Henard2001) report their findings based on a meta-analysis of ninetudies on customer satisfaction and repeat purchase, they do
cknowledge that further analysis is necessary because, “feworrelations are available in the literature to report on these asso-iations and so a few studies reporting different effect sizes in
dH2

Fig. 1. Relationship between multiple item custome

ling 89 (3, 2013) 246–262

he future could alter conclusions (24).” Further Szymanski andenard (2001) state that studying the relationship between sat-

sfaction, loyalty, retention, and other variables using researchxcluded from their meta-analysis could be insightful, inter-sting, and valuable. Hence, G1 is grounded on an extensiveiterature base, which studies additional variables, and both con-rms and extends their findings.

Interestingly, there is one study (Homburg and Furst 2005)hich stands apart from G1 by finding a nonsignificant

elationship between satisfaction and loyalty. However, this non-ignificant result is valid only for overall satisfaction but not forransactional satisfaction (positively related to intentions). Thexplanation may rely in the setting investigated. The study takeslace in a complaint management setting and it seems reason-ble to think that satisfaction (as transactional satisfaction isxpressed) recorded after an interaction with customer services dominant in affecting loyalty. Moreover, the sample size ofhe study is relatively small.

Regarding the variance explained in loyalty, it is not nor-ally possible to isolate the unique contribution of customer

atisfaction since other variables are introduced in the study asoderators, mediators or other predictors. The only exception

s presented by Anderson and Sullivan (1993), who indicaten R2 of .19 with the only variable being customer satisfac-ion (Table 1). In general, the R2 always refers to the overall

odel encompassing other variables. For instance, Agustin andingh (2005) report an R2 between .43 and .51 by including

rust and value in the model, while Cronin, Michael, and Hult2000) report an R2 of .94 by including service value and serviceuality (Table 1). Similarly, Seiders et al. (2005) indicate an R2

f .42 by including involvement and convenience. While theesults addressing the shape of the relationship between sat-sfaction and intentions are varied, it is safe to state that the

ajority of the studies report a linear relationship (Bloemer and
e Ruyter 1998; Bolton and Drew 1991; Cronin, Michael, andult 2000; Garbarino and Johnson 1999; Homburg and Furst005; Homburg and Giering 2001; LaBarbera and Mazursky
r satisfaction measures and loyalty intentions.

V. Kumar et al. / Journal of Retailing 89 (3, 2013) 246–262 249

Table 1Summary of satisfaction–loyalty intention findings.

Customer satisfaction

Single item Multiple items

Overall satisfaction Overall satisfaction ACSI/SCSB Transactionalsatisfaction

Attribute satisfaction

Loyaltyintentions

Single itemLinear• LaBarbera andMazursky (1983) (+)• Mittal, Kumar, andTsiros (1999) (+)R2: .37–.50• Shankar, Smith, andRangaswamy (2003) (+)R2: .39–.50• Baumann, Burton, andElliott (2005) (+)R2: .47–.72• Keiningham et al.(2007) (+)• Anderson andSullivan (1993) (+)R2: .19 (only CS)

Linear• Olsen (2002) (+)• Gustafsson andJohnson (2004) (+)R2: .38• Chandrashekaranet al. (2007) (+)

Decreasing returns• Oliva, Oliver, andMacMillan (1992) (+)R2: .33

Concave/convex• Jones and Sasser(1995) (+)

Asymmetric nonlinear• Mittal, Ross, andBaldasare (1998) (+)

Multiple itemLinear• Bloemer and deRuyter (1998) (+)

Linear• Garbarino andJohnson (1999) (+)• Cronin, Michael, andHult (2000) (+)R2: .94• Lam et al. (2004) (+)• Homburg and Furst(2005) (ns)

Increasing returns• Anderson andMittal (2000) (+)

Linear• Homburg andFurst (2005) (+)

Linear• Homburg andGiering (2001) (+)• Liang and Wang(2004) (+)• Seiders et al. (2005)(+)R2: .42

Nonlinear, quadratic• Ngobo (1999) (+)R2: 0.57

Decreasingreturns• Agustin andSingh (2005) (+)

2

( n cust

1Ta

t(MaufweiziPaRS

bhl(

saeaqoOab

c

+), (−) and ns (non significant) refer to the direction of the association betwee

983; Lam et al. 2004; Liang and Wang 2004; Mittal, Kumar, andsiros 1999; Olsen 2002; Seiders et al. 2005; Shankar, Smith,nd Rangaswamy 2003).

The exceptions to the linear relationship findings includehe studies conducted by Jones and Sasser (1995), Ngobo1999), Mittal, Ross, and Baldasare (1998), Oliva, Oliver, and

acMillan (1992), Anderson and Mittal (2000), and Agustinnd Singh (2005). For instance, Anderson and Mittal (2000),sing the ACSI, find that the link between customer satis-action and repurchase intention is asymmetric and nonlinearith increasing returns. The line becomes steeper on each

nd, where the line rises into the delight or extreme dissat-sfaction zone. In the middle there is a flattening zone, aone of apathy where changes in customer satisfaction resultn minor changes in loyalty (Anderson and Mittal 2000; J.D.ower and Associates 2007). As a consequence, when customers
re delighted (Berman 2005; Jones and Sasser 1995; Oliver,ust, and Varki 1997; Reichheld 1996; Rust and Oliver 2000;chneider and Bowen 1999), they tend to ignore competing
wNa

R : .43–.51

omer satisfaction and the dependent variable.

rands, while a decrease in satisfaction below a certain thresholdas a greater impact on repurchase intentions than an equiva-ent increase in the flattening zone (Anderson and Mittal 2000)Fig. 2).

On the contrary, Agustin and Singh (2005) highlight theimultaneity in curvilinear effects of loyalty determinants suchs transactional satisfaction, trust, and relational value, the latterxpressed as an evaluation of price paid. In particular, trans-ctional satisfaction has a positive linear effect but a negativeuadratic effect. That is, as satisfaction increases, its impactn loyalty decreases. Decreasing returns are supported also byliva, Oliver, and MacMillan (1992). Some other authors have

nalyzed the variation of the shape of the relationship on theasis of industry characteristics.

For instance, Jones and Sasser (1995) find that in highlyompetitive industries the shape of the relationship is convex,
hile in less competitive industries it is concave. Similarly,gobo (1999) finds that the nonlinear relationship varies
ccording to the industry (quadratic negative relationship with

250 V. Kumar et al. / Journal of Retai

F(

dmloer1taastG

Ms

i(cvIwvi(

dRfai2nat

mcssattcAiG

tbprstBOacdcOiv(Sttsall

tme1CrNrlfillpasDtimportance in affecting loyalty online than offline. Addressing

ig. 2. Shape of the relationship between satisfaction and repurchase intentionsAnderson and Mittal 2000).

ecreasing returns for an insurance company, two thresholdodel with decreasing returns for the camera and bank industry,

inear relationship for the retailer). Two studies, such as thosef Jones and Sasser (1995) and Ngobo (1999) are considerednough to define an empirical generalization related to nonlinearelationships affected by the type of industry (Bass and Wind995, p. 2). In addition, other authors have argued that industryype impacts the association between customer satisfactionnd behavior (Keiningham et al. 2007; Verhoef 2003). Ittnernd Larcker (1998) find that the value relevance of customeratisfaction measures varies across industries. This leads us tohe following generalization:

2: The type of industry affects the specific shape of the nonlinearrelationship.

oderators in the relationship between customeratisfaction and loyalty intentions

The relationship between customer satisfaction and loyaltyntentions is strongly affected by the presence of moderatorsBaron and Kenny 1986) that can strengthen or weaken the asso-iation. This explains why satisfied customers defect, since otherariables intervene in affecting the strength of the relationship.n fact, Reichheld (1996) notes that 65–85 percent of customersho defect, report before defection, that they were satisfied orery satisfied. Customers can express different levels of loyaltyntentions while holding similar levels of customer satisfactionReichheld 1996).

According to Seiders et al. (2005), moderators have beenivided into customer, relational, and marketplace variables.eferring to customer-related moderators, past research has

ound positive moderator effects with satisfaction strength, andge, and negative effects with variety seeking behavior, andncome (Chandrashekaran et al. 2007; Homburg and Giering001). In particular, Homburg and Giering (2001) find a sig-
ificant moderating effect of customer characteristics such asge, variety seeking behavior, and income. That is, young cus-omers tend to be less loyal, while variety seeking behavior
aaa

ling 89 (3, 2013) 246–262

arkedly weakens the relationship. As a consequence, in highlyompetitive environments that allow for several choices, ifwitching costs are not severe, we can expect a weaker relation-hip due to the natural inclination of the customer to try differentlternatives. Regarding income, past research has found that inhe automotive industry it negatively moderates the relationship:hat is, a greater availability of economic resources broadens theustomer’s range of alternative options, thus reducing loyalty.t similar levels of customer satisfaction, customers with higher

ncomes display less loyalty toward the company (Homburg andiering 2001).Relational moderators are variables that can depict the rela-

ionship between the customer and the company; customers cane variedly interested in forming a relationship with the com-any, thus exhibiting a major or minor propensity in investingesources to strengthen it (Garbarino and Johnson 1999). Inome situations, relational variables can strengthen the associa-ion between satisfaction and loyalty (Agustin and Singh 2005;aumann, Burton, and Elliot 2005; Bloemer and de Ruyter 1998;liva, Oliver, and MacMillan 1992). For instance, Oliva, Oliver,

nd MacMillan (1992) find that when transaction costs are suffi-iently high, a consumer may remain loyal even under moderateissatisfaction. This means that high levels of transaction costsan entangle the customer in a not fully satisfactory relationship.n a similar line, Bloemer and de Ruyter (1998) point out the

mportance of “elaboration”, an indicator of the customer moti-ation to evaluate a store, while Baumann, Burton, and Elliot2005) identify “the length of the relationship” and Agustin andingh (2005) the “value” as elements that strengthen the rela-

ionship. On the contrary, Garbarino and Johnson (1999) findhat for customers reporting high levels of relationship value,atisfaction is less important than “trust” and “commitment” inffecting loyalty. Finally, Chandrashekaran et al. (2007) find “theength of the relationship” as not being influential in determiningoyalty.

Among marketplace moderators we have switching costs,he type of product, the level of competition, and the kind ofedium (online vs. offline) used by customers to have negative

ffects (Baumann, Burton, and Elliot 2005; Jones and Sasser995; Olsen 2002; Shankar, Smith, and Rangaswamy 2003).ustomers perform their economic transactions in different envi-

onments and marketplaces that can affect the relationship.otably, the Internet has radically changed the way customers

elate to a company, ultimately affecting their satisfaction andoyalty. For instance, Shankar, Smith, and Rangaswamy (2003)nd that overall satisfaction has a stronger positive impact on

oyalty online than offline. The Internet has actually createdess loyal customers. In fact, rather than repurchasing the sameroduct over time, Internet consumers are more likely to lookt every purchase as a fresh start, counting on the impres-ive quantity of information and choices coming from the web.ue to the higher competition exacerbated by the Internet and

he customers’ empowerment, satisfaction acquires much more

different marketplace moderator (type of product), Szymanskind Henard (2001) find that the correlation between satisfactionnd repeat purchasing is lower on average when products rather

Retai

ttBaTG

tibpomatbttwdipctletttlilG

R

amacMltpEMtsiif

AeeiiG

A

ooii

pntdtW

iawcieWsTtofmtp

DmomSAJoseWa


han services are the focus of the study. Olsen (2002) also findshat the relationship varies across products, while Baumann,urton, and Elliott (2005) find a very small negative moder-ting effect of switching costs on intentions in the short term.his leads us to:3: The relationship between customer satisfaction and loyalty intentions

is moderated by customer, relational, and marketplace characteristics.Interestingly, these factors show a more mixed effect (positive andnegative) depending on the specific variable used in the analysis.

Also, the impact of customer satisfaction on loyalty inten-ions changes over time. A satisfied customer can state somententions today that may differ from her intentions tomorrow,ecause of the influence of the moderators in the interveningeriod. Customers might discover a new competitor’s productr, more simply, their memory about the positive experienceight decay over time (Mazurski and Geva 1989; Mittal, Kumar,

nd Tsiros 1999). Mazurski and Geva (1989) find that the rela-ionship becomes weaker as time goes by and the time lagetween customer satisfaction and loyalty increases. In addi-ion, the drivers of customer satisfaction can also change overime. For instance, Mittal, Kumar, and Tsiros (1999) find thathen customers buy a car, their initial satisfaction is mainlyriven by the experience with the dealer service. However, dur-ng later consumption periods, when they get to experience theroduct more, satisfaction with the product prevails. To con-lude, customers value different attributes over time, implyinghat different kinds of investment are required over the customerifecycle to improve the overall satisfaction or the total customerxperience. To make matters a bit murkier, in most cross sec-ional studies, customer satisfaction and loyalty are measured athe same time with common method bias potentially influencinghe responses (Agustin and Singh 2005). Unfortunately, a lack ofongitudinal research investigating the impact of customer sat-sfaction on loyalty makes it difficult to judge conclusively theong-term effect of the relationship. Thus, we generalize that:

4: The satisfaction–loyalty relationship has the potential to change overthe customer lifecycle.

ole of mediators and other predictors of loyalty intentions

Past research has shown that customer satisfaction does notlways have a direct effect on loyalty, but often works throughediators. In particular, Agustin and Singh (2005), Garbarino

nd Johnson (1999) and Liang and Wang (2004) identified trust,ommitment, and relational value, to be potential mediators.ost of these studies also introduce other relevant predictors of

oyalty intentions, some of which have shown stronger explana-ory power than satisfaction in determining loyalty. In particular,ast studies (Agustin and Singh 2005; Baumann, Burton, andlliot 2005; Cronin, Michael, and Hult 2000; Lam et al. 2004;ittal, Kumar, and Tsiros 1999) have examined the role of

rust, relational value, switching costs, length of the relation-
hip, affective attitude, service quality, service value, and priorntentions in predicting loyalty intentions. In fact, these stud-es address a critical need in the satisfaction–loyalty literatureor more holistic models explain the outcome variable better.
Rlfih

ling 89 (3, 2013) 246–262 251

gustin and Singh (2005) and Cronin, Michael, and Hult (2000),xpress the need to collectively include more predictors toxplain loyalty, since, from a managerial standpoint, establishingnitiatives to improve only one variable – customer satisfaction,s an incomplete strategy. This leads us to conclude that:

5: Holistic models that encompass other relevant variables as amoderator, mediator, as antecedent variables, or all three are betterpredictors of loyalty than models with just customer satisfaction.

ttitudinal loyalty measured as Word-of-Mouth (WOM)

WOM has received a lot of attention as an alternative measuref loyalty. For instance, Aaker (1991) noted that the real valuef those customers most loyal to an entity stems more from theirmpact on other customers in the marketplace than from theirndividual purchase behavior.

Notably, Reichheld (2003) states that the only number a com-any needs to grow is the net promoter score (NPS), the netumber of customers willing to recommend the company. Evenhough this statement has been largely disproved by recent aca-emic literature, the remarkable impact the NPS has created onhe business environment is proof of the importance imputed to

OM as an alternative measure of loyalty.WOM can be positive or negative. Positive WOM may

nclude making recommendations about a product or service,nd informing others of the quality of an offer. Customersho spread favorable WOM about a company can become the

ompany’s best salespeople. On the contrary, negative WOMncludes expressing disappointment about a negative experi-nce or product or a complaint. Customers spreading negativeOM can poison the company’s reputation and can actively

eek for other more valuable alternatives (Wangenheim 2005).oday, communities of angry customers can easily express

heir complaints about a bad experience by simply postingn the web (examples are consumerreview.com, dpreview.com,ailingenterprise.com). While before companies were prettyuch immune to negative WOM coming from angry consumers,

oday, the Internet has given the customers an unprecedentedower in attacking companies’ reputation.

Memorable is the extreme behavior of a customer, Jeremyorosin, who, in 1995 bought an expensive Starbucks espressoachine for $299 (www.starbucked.com). The machine turned

ut to be defective almost immediately. The replacementachine was also found to be defective. Dorosin complained totarbucks regional offices, but never got a satisfactory answer.s a consequence, he started to purchase ads on Wall Street

ournal to complain about the company. This got the attentionf the national media with appearances on popular televisionhows talking about his bad experience. While this might be anxtreme behavior, it is an example that reiterates the power of

OM, one that companies cannot afford to ignore (J.D. Powernd Associates 2007).

Customer satisfaction is considered an antecedent of WOM.
esearch has shown that positive WOM from satisfied customers
owers the cost of attracting new customers and enhances therm’s overall reputation, while that from dissatisfied customersas the opposite effect (Anderson, Fornell, and Mazvancheryl

http://www.consumerreview.com/

http://www.dpreview.com/

http://www.failingenterprise.com/

http://www.starbucked.com/

2 Retai

2cWemr

svei2nwtfi((cp

eaNtsmtIcstsc

M

iWro2hmFaffAt(o

ttw

sati(scaG

C

braadtutwi

Ba

itaGMMnttctichutnd2c

ba


004; Fornell 1992). The studies on the relationship betweenustomer satisfaction and WOM are mostly cross-sectional, withOM being a self-reported measure of past behavior. Brown

t al. (2005) is an exception where they measure WOM threeonths after reporting customer satisfaction; but it is still a self-

eported measure of past behavior.Table 2 summarizes the results of the relationship between

atisfaction and WOM (after controlling for other possibleariables), identifying the direction, shape, and the variancexplained. While most studies have examined the effect of pos-tive WOM (Brown et al. 2005; Verhoef, Franses, and Hoekstra002; Wangenheim and Bayon 2003), a few consider simulta-eously both positive and negative WOM effects, and investigatehether dissatisfied angry customers have a higher propensity

o report negative experiences to others as compared to satis-ed customers’ propensity to report positive recommendationsAnderson 1998; Bowman and Narayandas 2001). Wangenheim2005) provides an interesting twist, where satisfaction with theurrent provider is related to negative WOM about the previousrovider.

For the most part, the shape of the relationship curve is lin-ar (Brown et al. 2005; Lam et al. 2004; Verhoef, Franses,nd Hoekstra 2002), with Anderson (1998) and Bowman andarayandas (2001) reporting the existence of a U-shaped rela-

ionship. Specifically, Anderson (1998) showed that extremelyatisfied and dissatisfied customers are more vociferous thanerely satisfied customers and that, extremely dissatisfied cus-

omers engage in greater WOM than highly satisfied customers.n the middle lies a big portion of “passive” and complacentustomers, merely satisfied customers, who normally do notpeak about their experiences, good or bad, but are susceptibleo competitive actions. Fig. 3 presents the results of the analy-is, indicating direction, shape, and number of studies in eachategory.

oderators, mediators, and other predictors of WOM

The relationship between customer satisfaction and WOMs characterized by the presence of moderators and mediators.

hile customer satisfaction has a positive effect on customereferral, other variables seem to predict WOM better. Amongther variables, past research (Brown et al. 2005; Lam et al.004; Verhoef, Franses, and Hoekstra 2002; Wangenheim 2005)as found commitment, trust, payment equity, product involve-ent, and market mavenism to be better predictors of WOM.or instance, Verhoef, Franses, and Hoekstra (2002) found thatffective commitment is a better predictor of WOM than satis-action. Similarly, in a meta-analysis, de Matos and Rossi (2008)ound that commitment is the most relevant antecedent of WOM.lso, among moderators, Brown et al. (2005) found that cus-

omer commitment weakens the relationship, while Anderson1998) showed the existence of differences between countriesf origin.

The influence of commitment on the satisfaction–WOM rela-ionship is intriguing. It is interesting to note that a variablehat serves to express the strength of customers’ relationshipith the firm, actually contribute to weakening the effect of

r

sr

ling 89 (3, 2013) 246–262

atisfaction on WOM (Brown et al. 2005). Commitment comescross as a critical variable since it both mediates and moderateshe relationship, while satisfaction assumes a more central rolen explaining the referral activity in low commitment situationsBrown et al. 2005). Similarly, Bowman and Narayandas (2001)how that the more satisfied customers are with the final out-ome of a complaint, the less likely they are to engage in WOMctivity. This leads us to generalize that:

6: While customer satisfaction is positively related to Word-of-Mouth,models with related variables such as commitment, trust, and productinvolvement serve as better predictors of WOM.

ustomer satisfaction and behavioral loyalty

Often, companies are more interested in observing customerehavior, rather than intentions, since it can be directly linked toevenues and profitability (Bemmaor 1995; Chandon, Morwitz,nd Reinartz 2005; Jamieson and Bass 1989). Table 3 presentssummary of the research in this area that have used several

ifferent measures of behavioral loyalty including retention (orhe complementary metric – defection/churn), lifetime duration,sage, share of wallet and cross buying. Retention, lifetime dura-ion and usage reflect the length and the depth of the relationship,hile cross buying and share of wallet provides an indication of

ts breadth (Bolton, Lemon, and Verhoef 2004).

ehavioral loyalty measured in terms of relationship lengthnd depth – customer retention, lifetime duration, and usage

In examining retention, defection, and usage behaviors, it ismportant to note that the behavioral variables are recorded someime after the customer satisfaction survey (Bolton 1998; Boltonnd Lemon 1999; Capraro, Broniarczyk, and Srivastava 2003;ustafsson, Johnson, and Ross 2005; Ittner and Larcker 1998;ittal and Kamakura 2001; Seiders et al. 2005). For instance,ittal and Kamakura (2001), in an automotive setting, record the

ew brand acquired by the customer after a customer satisfac-ion survey, while Bolton, Kannan, and Bramlett (2000) recordhe number of customer transactions and monitor whether theustomer has canceled the service during the year followinghe survey. Different measures of customer behavior are usedn contractual versus noncontractual settings. In particular, forontractual settings (such as financial, telecommunication, andealth insurance), measures of retention or defection/churn aresed since it is relatively straightforward to observe termina-ion of the customer-provider relationship. On the contrary, inoncontractual settings (such as retail and automotive), whereefection cannot be easily detected (Reinartz and Kumar 2000,002), metrics such as repurchase behavior, number of repur-hase visits, and dollar spent are used.

Table 4 and Fig. 4 present a classification of the studiesased on how customer satisfaction and the dependent variablere measured. These exhibits also report the direction of the
elationship and, when possible, the shape of the relationship.
While these studies mostly predict a positive relation-hip between satisfaction and measures of behavioral loyalty,egarding the shape of the relationship, the results are not


Table 2Summary of satisfaction–WOM findings.



Overall Transactional Overall Transactional Attributesatisfaction

ACSI/SCSB

WOMPosit. WOM • Wangenheim and

Bayon (2003) (+)Linear• Lam et al. (2004)(+)

Linear• Brown et al.(2005)R2: .29 (+)• Verhoef, Franses,and Hoekstra (2002)R2: .37 (+)• Wangenheim andBayon (2003) (+)

Negat. WOM • Wangenheim(2005) (+)R2: .26–.59

Posit. andNegat. WOM

U shaped• Bowman and

U shaped• Anderson (1998)

2

( n cust

cr2a2op

Ra

ime

abb2sssytsh

Narayandas (2001)


onclusive. While some studies report a nonlinear and asymmet-ic association (Ittner and Larcker 1998; Mittal and Kamakura001), others (Bolton and Lemon 1999; Gustafsson, Johnson,nd Ross 2005; Perkins-Munn, Lerzan Aksoy, and Keiningham005), find a linear relationship. More interestingly, the kindf setting (contractual vs. noncontractual) does not consistentlyredict a positive or negative relationship.

ole of moderators in the relationship between satisfactionnd behavioral loyalty

Here again, the satisfaction–behavioral loyalty relationships affected by the presence of moderators (customer, relational,

arketplace, or all three). In particular, among customer mod-rators, past research has found positive effects for age, income,

tbh

Fig. 3. Direction and shape of the sa

R : .03–.1 (only CS)


nd gender, and negative effects for level of education, and num-er of children, with marital status, and competitor knowledgeeing not significant (Capraro, Broniarczyk, and Srivastava003; Mittal and Kamakura 2001; Seiders et al. 2005). Morepecifically, Mittal and Kamakura (2001) find that the relation-hip between satisfaction and repurchase behavior for cars to betronger for women than for men, and stronger for older than forounger consumers. Moreover, subjects with more educationend to have lower levels of retention than those with a highchool education. Also, consumers with one or more child in theousehold have lower tolerance than those without any children.

Interestingly, there have been very few studies examining
he moderating role of marketplace variables in the relationshipetween satisfaction and behavioral loyalty. Of these, mostave found little or no moderating effect of these variables.
tisfaction–WOM relationship.

254 V. Kumar et al. / Journal of Retailing 89 (3, 2013) 246–262

Table 3Satisfaction – retention, lifetime, and usage relationship shapes and direction.

Study Dependent variable anddirection of therelationship

Shape of therelationship ortechnique used

Industry Percent of the varianceexplained

Bolton (1998) Duration of theprovider–customerrelationship (+)

Proportionalhazard regression

Cellular telephone industry(contractual setting)

8 percent (only CS)

Bolton and Lemon (1999) Actual usage level (+) Linear Continuous service providers(contractual setting)

12 percent with otherpredictors

Bolton, Kannan, andBramlett (2000)

Retention (+) (renewalof the membership) andnumber of transaction inthe following year (+)

Logisticregression andtobit model

Financial services (creditcard) (contractual setting)

Capraro, Broniarczyk,and Srivastava (2003)

Defection (−) Hierarchicallogistic regression

Choice of health insuranceplan at a large University(contractual setting)

8 percent (only CS)25 percent with otherpredictors

Gustafsson, Johnson, andRoss (2005)

Churn (−) Linear Financial services (credit cardmembership) (contractualsetting)


Ittner and Larcker (1998) Retention rate in thefollowing year (+)

Linear Telecommunication industrywith one year contract

Adjusted R2 from 1.3percent to 4.9 percentwith relationship age

Mittal and Kamakura(2001)

Repurchase behavior ofa new car (+)

Nonlinear Automotive (noncontractual) 11.25 percent with otherpredictors

Perkins-Munn, LerzanAksoy, andKeiningham (2005)

Actual repurchase (+) Linear Truck industry;pharmaceutical(noncontractual)


Seiders et al. (2005) Number of repurchasevisits and repurchasespending in thefollowing 52 weeks (ns)

Linear Retail chain of upscalewomen’s apparel(noncontractual)

From 10 percent to 13percent with otherpredictors

Verhoef (2003) Retention (ns) Probit model Insurance products(contractual)


(+), (−) and ns (non significant) refer to the direction of the association between customer satisfaction and the dependent variable.

Table 4Summary of satisfaction – retention, lifetime, and usage relationship findings.



Overall satisfaction Overall satisfaction Attribute satisfaction Relative satisfaction

Behavioralloyalty

Retention/occurrence of the repurchase Nonlinear increasingreturns• Mittal andKamakura (2001) (+)

Diminishing returns• Ittner and Larcker(1998) (+)

Linear• Perkins-Munn,Lerzan Aksoy, andKeiningham (2005)(+)

• Bolton, Kannan, andBramlett (2000) (+)

Churn Linear• Gustafsson,Johnson, and Ross(2005) (−)

• Capraro,Broniarczyk, andSrivastava (2003) (−)• Verhoef (2003) (ns)

Duration of the relationship • Bolton (1998) (+)

Usage

Minutes ofusage

Linear• Bolton and Lemon(1999) (+)

Number ofrepurchasevisits

Linear• Seiders et al. (2005)(ns)

Amount ofspending

Linear• Seiders et al. (2005)(ns)

Number oftransactions

• Bolton, Kannan, andBramlett (2000) (+)

(+), (−) and ns (non significant) refer to the direction of the association between customer satisfaction and the dependent variable.


tion–r

FevpSorv

K2roibw2TG

R

atIrcRAmma

ASovpmttcawtmmo

iacciKaiHmca

ielst

Fig. 4. Direction and shape of the satisfac

or instance, research (Mittal and Kamakura 2001; Seiderst al. 2005) has found very little effect of other marketplaceariables such as the area of customer’s residence and com-etitive intensity in moderating this relationship. However,eiders et al. (2005) did report a positive moderating effectf convenience in the relationship between satisfaction for aetailer and behavioral loyalty. More studies are needed toerify the moderating role of the marketplace variables.

Among relational moderators, studies (Bolton 1998; Bolton,annan, and Bramlett 2000; Gustafsson, Johnson, and Ross005; Seiders et al. 2005; Verhoef 2003) have found that churn,elationship age, membership in a loyalty program, and levelf involvement to have positive effects. It is worth noting thatn the two studies in which satisfaction has no direct effect onehavior, satisfaction turns out to be significant by interactingith other variables, for instance, with relationship age (Verhoef003), involvement, and household income (Seiders et al. 2005).his leads us to the generalization that:

7: While customer satisfaction is mostly positively related to behavioralloyalty measures, by itself, it does not always result in higherlikelihoods of retention, longer lifetime duration, and higher levels ofusage. Customer, relational, and marketplace variables play asignificant moderating role.

ole of other predictors in explaining behavioral loyalty

Past research studies have shown that other predictors of loy-lty are significant and can have a stronger explanatory powerhan satisfaction (Capraro, Broniarczyk, and Srivastava 2003;ttner and Larcker 1998). Among significant predictors, we haveelationship age (Ittner and Larcker 1998), prior churn or priorustomer tendency to switch provider (Gustafsson, Johnson, andoss 2005), likelihood to repurchase (Perkins-Munn, Lerzan
ksoy, and Keiningham 2005), commitment, loyalty productembership, type of product (Verhoef 2003), level of involve-ent (Seiders et al. 2005), knowledge about competitive offers,
nd switching risk (Capraro, Broniarczyk, and Srivastava 2003).

ttt

etention, lifetime, and usage relationship.

n interesting finding is presented in Capraro, Broniarczyk, andrivastava (2003), where it is shown that customer knowledgef competitive alternatives account for about twice as muchariance in explaining customer defection as satisfaction anderceived switching risk. In fact, it appears that consumers areore likely to stay with a brand, even one that has disappointed

hem in the past, if they have no information of alternatives. Onhe contrary, an in-depth knowledge of alternate offers providesustomers an incentive to switch. Likewise, Bolton, Kannan,nd Bramlett (2000) argue that members of loyalty programseigh re-patronage intentions more heavily than nonmembers,

hus indicating a direct relationship between reward programembership and behavioral loyalty. Further, they argue thatembers of loyalty programs reveal stronger ties to the service

rganization than nonmembers.Of the variables shown by past studies as predictors of behav-

oral loyalty, purchase and ego involvement can be considereds important antecedents to brand loyalty. Purchase involvementan best be understood as the cost, effort or investment in a pur-hase (Mittal and Lee 1989). It is the outcome of an individual’snteraction with a product and the purchase situation (Beatty,ahle, and Homer 1988). Ego involvement has been defined

s the importance of the product to the individual and to thendividual’s self concept, values and ego (Beatty, Kahle, andomer 1988). Ego involvement is similar to enduring involve-ent defined as an ongoing concern for a particular product

lass and relatively independent of purchase situations (Blochnd Richins 1983; Richins and Bloch 1986).

Beatty, Kahle, and Homer (1988) conceptualized and empir-cally tested an involvement–commitment model, showing thatgo involvement leads to purchase involvement, which in turneads to brand commitment. Other research has empiricallyupported the purchase involvement–brand commitment rela-ionship (Mittal and Lee 1989). Dick and Basu (1994) advance
he proposition that higher ego involvement is likely to leado customer loyalty. Other researchers have similarly suggestedhat ego or enduring involvement leads to higher brand loyalty

256 V. Kumar et al. / Journal of Retailing 89 (3, 2013) 246–262

Table 5Summary of satisfaction – share of wallet findings.



Overall satisfaction Change in overallsatisfaction

ACSI/SCSB Attributesatisfaction

Share of wallet

Objective Cubic• Keiningham et al.(2003) (+)R2: .07 (only CS)

Nonlinear• Cooil et al. (2007) (+)

Linear• Mägi (2003) (+)R2: .21–.29

Linear• Perkins-Munnet al. (2005) (+)R2: .07

Self-reportedDecreasing returns• Bowman andNarayandas (2001) (+)

Linear• Perkins-Munn et al.(2005) (+)R2: .14Linear

• Keiningham et al.(2007) (+)

Partially self-reported • Verhoef (2003)(ns)

( n cust

actLtstup

scw(at2tLMefvaG

Oc

tci

RZEii1sdttreFr

sswm(iimif2

tEepsMe


ttitudes or intentions (Zaichkowsky 1985), and in a servicesontext that involvement tends to lead to stronger loyalty tohe service provider (Ganesh, Arnold, and Reynolds 2000;ongfellow and Celuch 1992). Keiningham et al. (2007) ques-

ion that any single attitudinal measure alone, such as customeratisfaction, could best determine future customer behavior. Inheir study of three different industries, the authors argue for these of a multiple indicator instead of a single predictor model toredict customer retention.

Table 4, that summarizes the studies relating customeratisfaction and behavioral loyalty, clearly indicates that the per-entage of variance explained in behavioral loyalty increaseshen adding variables such as switching risk and knowledge

Capraro, Broniarczyk, and Srivastava 2003), previous churnnd commitment (Gustafsson, Johnson, and Ross 2005), affec-ive commitment, participation in a loyalty program (Verhoef003), involvement, relationship age, relationship program par-icipation (Seiders et al. 2005), prior usage and price (Bolton andemon 1999), likelihood to purchase and brand image (Perkins-unn, Lerzan Aksoy, and Keiningham 2005), age, gender and

ducation (Mittal and Kamakura 2001). When customer satis-action is considered the sole predictor of behavioral loyalty, theariance explained is lower (Bolton 1998; Capraro, Broniarczyk,nd Srivastava 2003). Hence:

8: Models that encompass along with satisfaction other relevant predictorvariables such as past customer tendency to switch provider,relationship age, commitment, loyalty program membership, level ofinvolvement, switching risk are better predictors of behavioral loyaltythan models with just customer satisfaction.

ther measures of behavioral loyalty: share of wallet andross buying

Recently, academic and practitioners have started to focusheir attention on share of wallet as a better metric to detectustomer behavior. In fact, research has shown that customersncreasingly hold polygamous loyalty to brands (Bennett and

Nai


undle-Thiele 2005; Cooil et al. 2007; Rust, Lemon, andeithaml 2004b; Uncles, Dowling, and Hammond 2003; Uncles,hrenberg, and Hammond 1995). Customers divide their spend-

ng among different brands in a category and are continuouslynfluenced by competition in their choices (Yim and Kannan999). For instance, some customers may just change theirpending pattern with a company rather than completely stopoing business with it, by shifting some of their share of walleto another brand. Therefore, companies are expending substan-ial effort in understanding the spending patterns of customersather than their defection. Once again, satisfaction is consid-red as a strong antecedent of share of wallet. Table 5 andig. 5 present a classification of the studies addressing thiselationship.

Table 5 classifies past studies based on the way customeratisfaction and share of wallet are measured. While customeratisfaction is measured using traditional methods, share ofallet can be a self-reported measure, a partially self-reportedeasure, or a measure recorded in the company’s database

objective). When share of wallet is a self-reported measure ands recorded in cross-sectional studies, it may be correlated to sat-sfaction as a result of common method bias. The self-reported

easures of share of wallet is similar to the use of repurchasententions questions commonly contained in a customer satis-action questionnaire (Keiningham, Perkins-Munn, and Evans003).

In the only study that allows isolating the single contribu-ion of customer satisfaction (Keiningham, Perkins-Munn, andvans 2003) the variance explained is only 7 percent. In gen-ral, the explained variance is rather small and it ranges from 7ercent to 29 percent, when other variables are introduced. Thehape of the relationship varies from linear (Mägi 2003; Perkins-

unn, Lerzan Aksoy, and Keiningham 2005) to nonlinear (Cooilt al. 2007), nonlinear with decreasing returns (Bowman and
arayandas 2001), and cubic in Keiningham, Perkins-Munn,
nd Evans (2003). This latter study states that the greatest pos-tive impact occurs at the upper extreme levels of satisfaction.


tisfact

Ms

vettailioct“wrwsBm

caocpaaKctG

i

wwioaoacVicht

ccnHiras(ttt(lfwG

Fig. 5. Direction and shape of the sa

oreover, the functional form of the relation varies by customeregments.

Moderators play a major role in the relationship and otherariables can be significant predictors than satisfaction. Here, wencounter customer and relational moderators more frequentlyhan marketplace variables. For instance, Cooil et al. (2007) findhat income, and length of the relationship negatively moder-te the relationship between change in satisfaction and changen share of wallet. That is, customer with high income and aengthy relationship are less likely to reduce their level of spend-ng with the company. Furthermore, Mägi (2003), in his studyf the grocery setting, found that “the interest of the customer inomparing different shopping alternatives on price (price sensi-ivity)” has a negative moderating effect on the relationship. Alsothe interest of a customer in establishing a personal relationshipith service personnel” has a negative moderating effect on the

elation. In other words, shoppers who value a personal relationith store personnel are less likely to decrease their share of

hopping as a consequence of a decrease in satisfaction. Also,owman and Narayandas (2001) find support for the positiveoderating effect of prior loyalty, and volume of purchase.Among significant predictors, research reveals that in a gro-

ery setting, customers who own a card of competing chainsnd are prone to compare price, and tend to reduce their sharef wallet (Mägi 2003). Further, Verhoef (2003) reports thatommitment, direct mailing, and the participation in a loyaltyrogram, positively affect share of wallet. In addition, Bowmannd Narayandas (2001) find that the level of loyalty directlyffect share of wallet, while Perkins-Munn, Lerzan Aksoy, andeiningham (2005) indicate repurchase intentions as a signifi-

ant predictor. The findings presented above lead us to generalizehat:

9: Here again, while customer satisfaction is positively related to share ofwallet, models that include other relevant moderator and predictorvariables explain share of wallet behavior better than models that relyonly on customer satisfaction.

Conventional wisdom states that customer satisfactionmpacts cross buying. In other words, higher the satisfaction t

ion–share of the wallet relationship.

ith a firm’s product, greater is the probability that the customerill buy other products/services from the firm. However interest-

ngly, the empirical studies that exist on the effect of satisfactionn cross buying, report contrasting findings. Verhoef, Franses,nd Hoekstra (2001, 2002) find no significant direct effectf satisfaction on cross-buying. However, Verhoef, Franses,nd Hoekstra (2001) find satisfaction to have an effect onross-buying when moderated by relationship length. Similarly,erhoef, Franses, and Hoekstra (2002) find that a change in sat-

sfaction level between two points in time positively affects thehange in number of services purchased; but, satisfaction itselfas no significant direct effect. The variance explained in thewo studies is 15 percent and 8 percent, respectively.

Loveman (1998), in a retail banking setting, finds that averageustomer satisfaction with the branch is significantly positivelyorrelated with average cross-sell, which expresses the averageumber of services purchased per household. In a bank setting,allowel (1996) reports that overall division satisfaction is pos-

tively related to the division-reported cross sell rates, whichecord the percentage of customer households with multipleccounts (account cross sell) or multiple services (service crossell). In these particular situations, the level of aggregation usedin Loveman’s study the branch level and in Hallowel’s studyhe division level) may have influenced the results. In fact, inheir comparison of two models for the sales–advertising rela-ionship at the individual and aggregate level, Bass and Leone1986) find that a model of the same form estimated at a higherevel of aggregation is characterized by an increased coefficientor the independent variable (advertising, in this situation). Thus,e conclude that:10: The relationship between customer satisfaction and cross buying is

inconclusive, with the level of aggregation used to analyze the datapotentially impacting the strength of the relationship.

So what do we know for sure about the customer
satisfaction–loyalty relationship?
This extensive literature review has highlighted the state ofhe art regarding the relationship between customer satisfaction

2 Retai

ahceyihstta

1

2

3

4

5

6

7

8

sdslNSKwch

rebtsen

(1

Baeosct(feadcaati

fncvrlmicocbmti

ritIooPirp

s


nd loyalty, both attitudinal and behavioral. In particular, itas brought to light several issues that should be carefullyonsidered in analyzing the efficacy of customer satisfaction inxplaining and predicting customer loyalty. In fact, for manyears companies all around the world have heavily investedn customer satisfaction in the hope of increasing loyalty, andence, consequently, profitability. Academics have conductedeveral studies on the satisfaction–loyalty relationship some-imes with contrasting findings. After having gone throughhe above analysis, the major findings of this review and theccompanying empirical generalizations include:

. Overall, there is a positive relationship between customersatisfaction and loyalty.

. However, the variance explained by just satisfaction is rathersmall – around 8 percent.

. Holistic models that encompass other relevant variables asmoderators, mediators, antecedent variables, or all three arebetter predictors of loyalty than models with just customersatisfaction.

. Inclusion of these variables increases the variance explained,on an average, to 34 percent (54 percent for attitudinal loyaltyand 15 percent for behavioral loyalty, respectively).

. The satisfaction–loyalty relationship has the potential tochange over the customer lifecycle.

. While customer satisfaction has a positive relationship withWOM, other related variables such as commitment, trust, andproduct involvement serve as better predictors of WOM.

. Customer satisfaction, by itself, does not always result inretention, lifetime duration and usage. Customer, relationaland marketplace variables often play a significant moderatingrole.

. The relationship between customer satisfaction and crossbuying is characterized by contrasting findings. The level ofaggregation used to analyze the data may impact the strengthof the relationship.

The preceding review and analysis indicate that customeratisfaction is often times a necessary but not a sufficient con-ition to predict loyalty. Our empirical generalizations are alsoupported by the findings on the customer satisfaction–loyaltyink discussed in service-profit-chain research (Bowman andarayandas 2004; Heskett et al. 1994; Heskett, Sasser, andchlesinger 1997; Kamakura et al. 2002; Loveman 1998; Rucci,irn, and Quinn 1998). The service profit chain (SPC) frame-ork states that exceptional customer service results in greater

ustomer satisfaction and retention, which in turn results inigher profitability.

Heskett et al. (1994) theoretically support the notion that theelationship between customer satisfaction and loyalty is non lin-ar with increasing returns. Heskett et al. (1997) find that the linketween customer satisfaction and loyalty, although positive, is
he weakest of all in the service profit chain, and that the relation-hip between them is not constant. The SPC proposed by Heskettt al. (1994) became rather popular as it is demonstrated by theumerous case studies reported by academics and practitioners
sRaC

ling 89 (3, 2013) 246–262

Loveman 1998; Rhian and Cross 2000; Rucci, Kirn, and Quinn998).

In an interesting application of the SPC to business markets,owman and Narayandas (2004) find that the experience of theccount manager and the client satisfaction with a competitornhance the relation between customer satisfaction and the Sharef Customer Wallet (SCW). Customer size decreases the respon-iveness of SCW to satisfaction. SCW is influenced by overallustomer satisfaction and the relation shows increasing returns,hus supporting the notion of customer delight. Kamakura et al.2002) using structural equation models, simultaneously testor all the links of the chain, investigating also for mediatingffects. Customer satisfaction itself is not an unconditional guar-ntee of profitability and some firms may remain unprofitableespite high levels of satisfaction due to a high investment inustomer satisfaction. Moderators are not investigated but theuthors advocate their inclusion in the model. Specifically, theuthors find a positive relationship between customer percep-ions of personnel and equipment with consumers’ behavioralntentions (intentions to recommend).

If customer satisfaction is not enough – what needs to bedone?

A more holistic view of the relationship: Customer satis-action is not enough to fully explain loyalty; other variableseed to be included in the relationship model to depict a moreomplete picture. In particular, it is clear from the review thatariables such as customer perceived value, switching costs, andelational variables such as trust, commitment, relationship age,oyalty program membership, and level of customer involve-

ent, seems to be the most desirable candidates for inclusionn the model. While it is clear that these additional variables areritical in customer satisfaction studies, their specific role in theverall model indeed varies depending on the circumstances andontext. Past research has shown these variables to alternativelye predictors of loyalty, antecedents to satisfaction, and act asoderators, mediators, or both in the satisfaction–loyalty rela-

ionship. The decision to include one or more of these variablesn a holistic model is very much context specific.

Who to satisfy? One of the main paradigms of customerelationship management stresses the fact that customers arendeed heterogeneous. However, companies still invest in cus-omer satisfaction in the same way for the entire customer base.n particular, customers are different in terms of the future value,r profitability, they can bring to a company. A truly customerriented approach optimizes customer selection (Kumar andetersen 2005), that is, allocation of resources to the most prof-

table customers for the company. When allocating financialesources, the most resources should be assigned to the mostrofitable (or potentially profitable) customers.

The future value of a customer can be efficiently mea-ured through the customer lifetime value (CLV) metric, whose
uperiority over other metrics (such as past profitability orFM models) in defining future customer profitability haslready been well demonstrated in the literature (Kumar 2008).LV is generally defined as the present value of all future

Retai

ptaeao(mhcski

aotCafaibeptTi

D

t(edHibt2wmtpwfAai

att2i2

opsspsinr

noncGfttsfatc

standtdstcsagotw

bdilscatos

t


rofits coming from a customer during his/her life or rela-ionship with a firm. It is similar to the discounted cash flowpplied in finance (Gupta and Zeithaml 2006). CLV is gen-rally applied at the individual customer or segment levelnd it is a forward-looking metric since makes a projectionver the future by incorporating sophistication in modelingGupta et al. 2006; Gupta and Zeithaml 2006). Given this, itakes sense to allocate resources first to customers with a

igh CLV. These customers are also the most attractive forompetitors. Clearly, the most profitable customers should beatisfied first in order to strengthen their relationship and toeep them away from the temptations of competitive offer-ngs.

How much and when to satisfy? The use of the CLV canlso provide several important insights on the maximum levelf investment that should be allocated to each customer. In fact,his is set by the future customer profitability as measured byLV. In other words, a company should not invest in a customern amount of resources greater than his/her expected level ofuture profitability. However, in order to decide when to invest,company should also look at the current level of customer sat-

sfaction for each customer and at the shape of the relationshipetween customer satisfaction and loyalty. For instance, in pres-nce of decreasing returns in the relationship, a company shoulday attention before deciding to invest in highly satisfied cus-omers to further secure their loyalty and hope for higher returns.he definition of the shape of the relationship plays a major role

n the cost/benefit relationship.

irections for future research

Our approach depicts a customer satisfaction strategyhat starts with future customer profitability considerationsCLV), with the end goal of undertaking different investments,fforts/expenditures incurred to exceed expectations or causeelight, for customers segments according to their profitability.owever, the satisfaction–loyalty relationship is not generally

nvestigated for different levels of customer profitability bothefore and after a customer satisfaction investment (i.e., effortso improve customer service) (Homburg, Koschate, and Hoyer005). In a recent study, Kumar et al. (2009) elaborate on theeakness of the satisfaction–loyalty link, as it is currently imple-ented by companies, to present an alternate path that reverse

he logic, the profitability–loyalty–satisfaction chain. The newaradigm starts the customer relationship management strategyith customer profitability and the idea that customers with dif-

erent profitability should be rewarded and satisfied differently.systematic analysis of the relationship between satisfaction

nd loyalty for the different levels of profitability is much neededn the literature.

The need to better investigate the link between satisfactionnd profitability as expressed by CLV is also supported byhe consideration that recent research has clearly demonstrated
hat loyalty is not appropriately measured (Reinartz and Kumar002) and that CLV is the best measure for predicting profitabil-ty of the company (Gupta, Lehmann, and Stuart 2004; Rust et al.004a). According to the above premises, a direct investigation
sftt

ling 89 (3, 2013) 246–262 259

f the satisfaction–CLV link that discards loyalty could be aromising avenue for future research. The presence of mediatorsuggests that researchers need to clearly examine how customeratisfaction affects financial performance. For instance, a recentaper (Luo, Homburg, and Wieseke 2010) shows that customeratisfaction led to improved analyst recommendations and thosen turn led to better financial performance. Such insights do notecessarily imply a reduced role for satisfaction, but rather theole of satisfaction needs to be better understood.

Further, we have seen that customer satisfaction itself mayot be enough to explain loyalty. However, relationships amongther relevant variables may change over time. There is an urgenteed for longitudinal studies in customer satisfaction that canapture these changing relationships over time. For instance,arbarino and Johnson (1999) demonstrate that whereas satis-

action mediates the relationship between trust and loyalty forransactional exchanges, the mechanism is different for rela-ional exchanges. In the latter case, trust mediates the effect ofatisfaction on loyalty intentions and therefore the effect of satis-action in affecting loyalty becomes less central. In other words,ntecedents of loyalty for customers with a relational orienta-ion are different from the antecedents of transaction-orientedustomers.

However, this study is cross sectional, so we cannot under-tand the dynamics and the interrelations among variables overime. In fact, did satisfaction contribute to the formation of trustnd commitment over time? What role does satisfaction playot only on loyalty but also on trust and commitment over time,uring the evolution of the relationship? It may be possiblehat, as relationships evolve and go through different phases, theynamics among variables change as well as the role of customeratisfaction on all the other variables. According to the results ofhe literature review, we may expect that early in the relationshipustomer satisfaction is more relevant, while, when the relation-hip gets firm, greater importance is attributed to commitmentnd trust. In this particular situation, the use of models with lon-itudinal data that can capture variation both cross-sectional andver time can be extremely useful. Researchers need to developheory to understand when and under what conditions the linkill be systematically stronger or weaker.The importance of time in customer satisfaction studies has

een highlighted also by other authors, since measures made atifferent points in time may drive to different conclusions. Fornstance, Mazurski and Geva (1989) find that satisfaction andoyalty intentions are highly correlated when measured in theame survey at the same time. However, for the same persons,ustomer satisfaction is not correlated with intentions measuredfter two weeks. In this particular situation, time plays an impor-ant role since the effect of customer satisfaction seems to decayver time. Hence, longitudinal studies are required to answeruch critical questions.

A third important issue is related to the way customer satisfac-ion is measured. While an attribute based measure of customer
atisfaction can be useful for managers to identify areas ofuture intervention and improvement, it does not lends itselfoward the delivery of a holistic experience for the customerhat involves “sense, feel, think, act and relate”. According to the

2 Retai

psstctaisl

aslIitttrgi

AA

A

A

A

A

B

B

B

B

B

B

B

B

B

B

B

B

B

B

B

B

B

B

B

C

C

C

C

C

d

D


rinciples of experiential marketing (Schmitt 1999), marketershould touch upon higher levels of the customer experience andtart thinking of an operationalization of customer satisfactionhat encompasses not only physical product characteristics oroncrete aspects of the service, but also intangible elements ofhe customer experience that can satisfy higher order needs suchs self-esteem, socialization, or both. Future research should alsonvestigate these aspects and delineate more precise measures ofatisfaction that encompass intangible aspects of an experienceeading to satisfaction.

In an interesting study of online markets, Shankar, Smith,nd Rangaswamy (2003) found that overall satisfaction had atronger positive impact on loyalty online than offline and thatoyalty is higher online than offline. As the relevance of thenternet in developing and strengthening customer relationshipsncreases, and customers are more and more used to making theirransactions online, a better understanding of these dynamics inhe online setting would be advocated. Are the dynamics of cus-omer satisfaction and loyalty the same online and offline? Moreesearch in this direction is certainly needed to shed light on arowing phenomenon that is marking the Marketing disciplinen the 21st century.

References

aker, David (1991), Managing Brand Equity, New York: Free Press.gustin, Clara and Jagdip Singh (2005), “Curvilinear Effects of Consumer

Loyalty,” Journal of Marketing Research, 43 (February), 96–108.nderson, Eugene W. (1998), “Customer Satisfaction and Word-of-Mouth,”

Journal of Service Research, 1 (1), 1–14.nderson, Eugene W., Claes Fornell and Sanal K. Mazvancheryl (2004), “Cus-

tomer Satisfaction and Shareholder Value,” Journal of Marketing, 68 (4),172–85.

nderson, Eugene W. and Vikas Mittal (2000), “Strengthening theSatisfaction–Profit Chain,” Journal of Service Research, 3 (November (2)),107–20.

nderson, Eugene W. and Mary Sullivan (1993), “The Antecedents and Con-sequences of Customer Satisfaction for Firms,” Marketing Science, 12(Spring), 125–43.

aron, Reuben and David Kenny (1986), “The Moderator–Mediator VariableDistinction in Social Psychology: Conceptual, Strategic and Statistical Con-siderations,” Journal of Personality and Social Psychology, 51 (6), 1173–82.

ass, Frank M. and Robert P. Leone (1986), “Estimating Micro Relationshipsfrom Macro Data: A Comparative Study of Two Approximations of theBrand Loyal Model Under Temporal Aggregation,” Journal of MarketingResearch, 23 (August), 291–7.

ass, Frank M. and Jerry Wind (1995), “Introduction to the Special Issue:Empirical Generalizations in Marketing,” Marketing Science, 14 (3) (Part 2of 2)

aumann, Chris, Suzan Burton and Greg Elliott (2005), “Determinants of Cus-tomer Loyalty and Share of Wallet in Retail Banking,” Journal of FinancialServices Marketing, 9 (3), 231–48.

eatty, Sharon E., Lynn R. Kahle and Pamela Homer (1988), “TheInvolvement–Commitment Model: Theory and Implications,” Journal ofBusiness Research, 16 (2), 149–67.

emmaor, Albert C. (1995), “Predicting Behaviour from Intention-to-BuyMeasures: The Parametric Case,” Journal of Marketing Research, 32 (May(9)), 176–91.

ennett, Rebekah and Sharyn Rundle-Thiele (2005), “The Brand Loyalty Life
Cycle: Implications for Marketers,” Journal of Brand Management, 12 (4),250–63.
erman, Barry (2005), “How to Delight Your Customers,” California Manage-ment Review, 48 (1), 129–51.

F

G

ling 89 (3, 2013) 246–262

iong, Herald (1993), “Satisfaction and Loyalty to Suppliers Within the GroceryTrade,” European Journal of Marketing, 27 (7), 21–38.

loch, Peter H. and Marsha L. Richins (1983), “A Theoretical Model forthe Study of Product Importance Perceptions,” Journal of Marketing, 47(Summer), 69–81.

loemer, Josee and Ko de Ruyter (1998), “On the Relationship Between StoreImage, Store Satisfaction and Store Loyalty,” European Journal of Market-ing, 32 (5/6), 499–513.

olton, Ruth and James H. Drew (1991), “A Longitudinal Analysis of the Impactof Service Changes on Customer Attitudes,” Journal of Marketing, 55 (1),1–9.

olton, Ruth, Katherine N. Lemon and Peter C. Verhoef (2004), “The Theo-retical Underpinnings of Customer Asset Management: A Framework andPropositions for Future Research,” Academy of Marketing Science, 32 (3),271–92.

olton, Ruth N. (1998), “A Dynamic Model of the Duration of the Customer’sRelationship with a Continuous Service Provider,” Marketing Science, 17(1), 45–65.

olton, Ruth N., P.K. Kannan and Matthew D. Bramlett (2000), “Implicationsof Loyalty Program Membership and Service Experiences for CustomerRetention and Value,” Journal of the Academy of Marketing Science, 28 (1),95–108.

olton, Ruth N. and Katherine Lemon (1999), “A Dynamic Model of Cus-tomers’ Usage of Services: Usage as an Antecedent and Consequence ofSatisfaction,” Journal of Marketing Research, 36 (May), 171–86.

owen, John T. and Shiang-Lih Chen (2001), “The Relationship BetweenCustomer Loyalty and Customer Satisfaction,” International Journal ofContemporary Hospitality Management, 13 (5), 213–7.

owman, Douglas and Das Narayandas (2001), “Managing Customer InitiatedContacts with Manufacturers: The Impact of Share of Category Require-ments and Word-of-Mouth Behavior,” Journal of Marketing Research, 38(3), 281–97.

and (2004), “Linking CustomerManagement Effort to Customer Profitability in Business Markets,” Journalof Marketing Research, 41 (4), 433–47.

rown, Tom J., Thomas Barry, Peter Dacin and Richard F. Gunst (2005),“Spreading the Word: Investigating Antecedents of Consumers’ PositiveWord-of-Mouth Intentions and Behaviors in a Retailing Context,” Journalof the Academy of Marketing Science, 33 (Spring), 123–38.

apraro, Anthony J., Susan Broniarczyk and Rajendra K. Srivastava (2003),“Factors Influencing the Likelihood of Customer Defection: The Role ofConsumer Knowledge,” Journal of the Academy of Marketing Science, 31(2), 164–75.

handon, Pierre, Vicki G. Morwitz and Werner J. Reinartz (2005), “Do Inten-tions Really Predict Behavior? Self-generated Validity Affects in SurveyResearch,” Journal of Marketing, 69 (April), 1–14.

handrashekaran, Murali, Kristin Rotte, Stephen Tax and Rajdeep Grewal(2007), “Satisfaction Strength and Customer Loyalty,” Journal of MarketingResearch, 49 (February), 153–63.

ooil, Bruce, Timothy L. Keiningham, Lerzan Aksoy and Michael Hsu (2007),“A Longitudinal Analysis of Customer Satisfaction and Share of Wallet:

Investigating the Moderating Effect of Customer Characteristics,” Journalof Marketing, 71 (January), 67–83.

ronin, Joseph, Brady Michael and Tomas Hult (2000), “Assessing theEffects of Quality, Value and Customer Satisfaction on Consumer Behav-ioral Intentions in Service Environments,” Journal of Retailing, 76 (2),193–218.

e Matos, Celso Augusto and Carlos Alberto Vargas Rossi (2008), “Word-of-Mouth Communications in Marketing: A Meta-analytic Review of theAntecedents and Moderators,” Journal of the Academy of Marketing Science,36 (4), 578–96.

ick, Alan S. and Kunal Basu (1994), “Customer Loyalty: Toward an IntegratedConceptual Framework,” Journal of the Academy of Marketing Science, 22(2), 99–113.

ornell, Claes (1992), “A National Customer Satisfaction Barometer: TheSwedish Experience,” Journal of Marketing, 56 (January), 6–21.

anesh, Jaishankar, Mark Arnold and Kristy E. Reynolds (2000), “Under-standing the Customer Base of Service Providers: An Examination of the

Retai

G

G

G

G

G

G

H

H

H

H

H

H

I

J

J

J

J

K

K

K

K

K

K

K

L

L

L

L

L

L

L

M

M

M

M

M

M

M

N

O

O

O

P

R

R


Differences Between Switchers and Stayers,” Journal of Marketing, 64 (3),65–87.

arbarino, Ellen and Mark S. Johnson (1999), “The Different Roles of Sat-isfaction, Trust and Commitment in Customer Relationships,” Journal ofMarketing, 60 (2), 70–87.

upta, Sunil, Dominique Hanssens, Bruce Hardie, Wiliam Kahn, V. Kumar,Nathaniel Lin, Nalini Ravishanker and S. Sriram (2006), “Modeling Cus-tomer Lifetime Value,” Journal of Service Research, 9 (2), 139–55.

upta, Sunil, Donald R. Lehmann and Jennifer Ames Stuart (2004), “ValuingCustomers,” Journal of Marketing Research, 41 (1), 7–18.

upta, Sunil and Valerie Zeithaml (2006), “Customer Metrics and Their Impacton Financial Performance,” Marketing Science, 25 (6), 718–39.

ustafsson, Anders and Michael D. Johnson (2004), “Determining AttributeImportance in a Service Satisfaction Model,” Journal of Service Research,7 (2), 124–41.

ustafsson, Anders, Michael D. Johnson and Inger Ross (2005), “The Effects ofCustomer Satisfaction, Relationship Commitment Dimensions, and Triggerson Customer Retention,” Journal of Marketing, 69 (October), 210–8.

allowel, Roger (1996), “The Relationships of Customer Satisfaction, Cus-tomer Loyalty, and Profitability: An Empirical Study,” International Journalof Service Industry Management, 7 (4), 27–42.

eskett, James L., Thomas O. Jones, Gary W. Loveman, W. Earl Sasser Jr. andLeonard A. Schlesinger (1994), “Putting the Service Profit Chain to Work,”Harvard Business Review, (March–April), 164–74.

eskett, James L., W. Earl Sasser Jr. and Leonard A. Schlesinger (1997), TheService Profit Chain: How Leading Companies Link Profit and Growth toLoyalty, Satisfaction and Value, New York, NY: Free Press.

omburg, Christian and Andreas Furst (2005), “How Organizational ComplaintHandling Drives Customer Loyalty: An Analysis of the Mechanistic and theOrganic Approach,” Journal of Marketing, 69 (July), 95–114.

omburg, Christian and Annette Giering (2001), “Personal Characteristics asModerators of the Relationship Between Customer Satisfaction and Loyalty– An Empirical Analysis,” Psychology and Marketing, 18 (1), 43–66.

omburg, Christian, Nicole Koschate and Wayne Hoyer (2005), “Do Satis-fied Customers Really Pay More? A Study of the Relationship BetweenCustomer Satisfaction and Willingness to Pay,” Journal of Marketing, 69(April), 84–96.

ttner, Christopher D. and David F. Larcker (1998), “Are Nonfinancial MeasuresLeading Indicators of Financial Performance? An Analysis of CustomerSatisfaction,” Journal of Accounting Research, 36 (Suppl.), 1–35.

amieson, Linda F. and Frank M. Bass (1989), “Adjusting Stated Inten-tion Measures to Predict Trial Purchase of New Products: A Comparisonof Models and Methods,” Journal of Marketing Research, 26 (August),336–45.

.D. Power and Associates (2007), Satisfaction, How Every Great CompanyListens to the Voice of the Customer, USA: Portfolio, Penguin Group.

ohnson, Michael D., Andreas Herrmann and Frank Huber (2006), “The Evo-lution of Loyalty Intentions,” Journal of Marketing, 70 (April), 122–32.

ones, Thomas O. and W. Earl Sasser Jr. (1995), “Why Satisfied CustomersDefect,” Harvard Business Review, 73 (6), 88–91.

amakura, Wagner A., Vikas Mittal, Fernando de Rose and Jose Afonso Mazzon(2002), “Assessing the Service-Profit Chain,” Marketing Science, 21 (3),294–317.

eh, Hean Tat and Yih Hwai Lee (2006), “Do Rewards Programs Build Loyaltyfor Services? The Moderating Effect of Satisfaction on Type and Timing ofRewards,” Journal of Retailing, 82 (2), 127–36.

einingham, Timothy, Bruce Cooil, Lerzan Aksoy and Weiner Tor Andreassen(2007), “The Value of Different Customer Satisfaction and Loyalty Metricsin Predicting Customer Retention, Recommendation and Share of Wallet,”Managing Service Quality, 17 (4), 361–84.

einingham, Timothy L., Tiffany Perkins-Munn and Heather Evans (2003),“The Impact of Customer Satisfaction on Share-of-Wallet in a Business-to-Business Environment,” Journal of Service Research, 6 (1), 37–50.

umar, V. (2008), Managing Customers for Profit, Upper Saddle River, NJ:
Wharton School Publishing.
umar, V. and J.A. Petersen (2005), “Using a Customer-Level Marketing Strat-egy to Enhance Firm Performance,” Journal of the Academy of MarketingScience, 33 (4), 504–19.

R

ling 89 (3, 2013) 246–262 261

umar, V., Ilaria Dalla Pozza, J. Andrew Petersen and Denish Shah (2009),“Reversing the Logic: The New Path to Profitability,” Journal of InteractiveMarketing, 23 (2), 147–56.

aBarbera, Priscilla A. and David Mazursky (1983), “A Longitudinal Assess-ment of Consumer Satisfaction/Dissatisfaction: The Dynamic Aspect ofthe Cognitive Process,” Journal of Marketing Research, 20 (November),393–404.

am, Shun Yin, Venkatesh Shankar, M. Krishna Erramilli and Bvsan Murthy(2004), “Customer Value, Satisfaction, Loyalty and Switching Costs: AnIllustration from a Business to Business Service Context,” Journal of theAcademy of Marketing Science, 32 (3), 293–311.

iang, Chiung-Ju and Wen-Hung Wang (2004), “Attributes, Benefits, CustomerSatisfaction and Behavioral Loyalty, an Integrative Research of FinancialServices Industry in Taiwan,” Journal of Service Research, 4 (1), 57–91.

ongfellow, Timothy A and Kevin G. Celuch (1992), “A Comparison of HighVersus Low Involvement Profiles,” Journal of Business and Psychology, 6(Summer), 483–93.

oveman, Gary W. (1998), “Employee Satisfaction, Customer Loyalty, andFinancial Performance: An Empirical Examination of the Service ProfitChain in Retail Banking,” Journal of Service Research, 1 (August (1)),18–31.

uo, Xueming and Christian Homburg (2007), “Neglected Outcomes of Cus-tomer Satisfaction,” Journal of Marketing, 71 (April), 133–49.

uo, Xueming, Christian Homburg and Jan Wieseke (2010), “Customer Sat-isfaction, Analyst Stock Recommendations, and Firm Value,” Journal ofMarketing Research, 47 (6), 1041–58.

ägi, Anne W. (2003), “Share of Wallet in Retailing: The Effects of Cus-tomer Satisfaction, Loyalty Cards and Shopper Characteristics,” Journal ofRetailing, 79 (2), 97–106.

azurski, David and Aviva Geva (1989), “Temporal Decay inSatisfaction–Purchase Intention Relationship,” Psychology and Marketing,6 (3), 211–27.

ittal, Banwari and Myung-Soo Lee (1989), “A Causal Model of ConsumerInvolvement,” Journal of Economic Psychology, 10 (3), 363–89.

ittal, Vikas and Carly Frennea (2010), Customer Satisfaction: A StrategicReview and Guidelines for Managers, Cambridge, MA: MSI Publication.

ittal, Vikas and Wagner A. Kamakura (2001), “Satisfaction, RepurchaseIntent, and Repurchase Behavior: Investigating the Moderating Effect ofCustomer Characteristics,” Journal of Marketing Research, 38 (1), 131–42.

ittal, Vikas, Pankaj Kumar and Michael Tsiros (1999), “Attribute-LevelPerformance Satisfaction, and Behavioral Intentions Over Time: A Con-sumption System Approach,” Journal of Marketing, 63 (2), 88–101.

ittal, Vikas, William T. Ross Jr. and Patrick M. Baldasare (1998), “The Asym-metric Impact of Negative and Positive Attribute Level Performance onOverall Satisfaction and Repurchase Intentions,” Journal of Marketing, 62(January), 33–47.

gobo, Paul V. (1999), “Decreasing Returns in Customer Loyalty: Does ItReally Matter to Delight the Customer,” Advances in Consumer Research,26, 469–76.

liva, Terence A., Richard L. Oliver and Ian C. MacMillan (1992), “A Catas-trophe Model for Developing Service Satisfaction Strategies,” Journal ofMarketing, 56 (3), 83–95.

liver, Richard, Roland T. Rust and Sajeev Varki (1997), “Customer Delight:Foundations, Findings and Managerial Insight,” Journal of Retailing, 7 (3),311–36.

lsen, Svein Ottar (2002), “Comparative Evaluation and the RelationshipBetween Quality, Satisfaction, and Repurchase Loyalty,” Journal of theAcademy of Marketing Science, 30 (3), 240–9.

erkins-Munn, Tiffany, Lerzan Aksoy, Timothy L. Keiningham and DemitryEstrin (2005), “Actual Purchase as a Proxy for Share of Wallet,” Journal ofService Research, 7 (3), 245–56.

eichheld, Frederick F. (1996), “Learning from Customer Defections,” HarvardBusiness Review, 74 (March), 56–69.

eichheld, Frederick (2003), “The One Number You Need to Grow,” Harvard
Business Review, 81 (December), 46–54.
einartz, Werner and V. Kumar (2000), “On the Profitability of Long LifeCustomers in a Noncontractual Setting: An Empirical Investigation andImplications for Marketing,” Journal of Marketing, 64 (4), 17–35.

2 Retai

R

R

R

R

R

R

R

S

S

S

S

S

S

S

U

U

V

V

V

W

W

Y

Yim, Chi Kin and P.K. Kannan (1999), “Customer Behavioral Loyalty: A Seg-


and (2002), “The Mismanage-ment of Customer Loyalty,” Harvard Business Review, 80 (7), 86–94.

hian, Silvestro and Stuart Cross (2000), “Applying the Service Profit Chain ina Retail Environment: Challenging the “Satisfaction Mirror”,” InternationalJournal of Service Industry Management, 11 (3), 244–68.

ichins, Marsha L. and Peter H. Bloch (1986), “After the New Wears Off: TheTemporal Context of Product Involvement,” Journal of Consumer Research,13 (2), 280–5.

ucci, Anthony J., Steven P. Kirn and Richard T. Quinn (1998), “TheEmployee–Customer-Profit Chain at Sears,” Harvard Business Review,(January–February), 83–97.

ust, Roland T., Tim Ambler, Gregory S. Carpenter, V. Kumar and Rajen-dra K. Srivastava (2004a), “Measuring Marketing Productivity: CurrentKnowledge and Future Directions,” Journal of Marketing, 68 (4), 76–89.

ust, Roland T., Katherine N. Lemon and Valarie A. Zeithaml (2004b), “Returnon Marketing: Using Customer Equity to Focus Marketing Strategy,” Journalof Marketing, 66 (1), 109–27.

ust, Roland T. and Richard L. Oliver (2000), “Should We Delight the Cus-tomer?,” Journal of the Academy of Marketing Science, 28 (1), 86–94.

ust, Roland T., Anthony J. Zahorik and Timothy L. Keiningham (1995),“Return on Quality (ROQ): Making Service Quality Financially Account-able,” Journal of Marketing, 59 (2), 58–70.

chmitt, Bernd (1999), Experiential Marketing: How to Get Customers to Sense,Feel, Think, Act, and Relate, New York, USA: The Free Press.

chneider, Benjamin and David E. Bowen (1999), “Understanding CustomerDelight and Outrage,” Sloan Management Review, 41 (1), 35–45.

eiders, Kathleen, Glenn B. Voss, Dhruv Grewal and Andrea L. Godfrey (2005),“Do Satisfied Customers Buy More? Examining Moderating Influences in

a Retailing Context,” Journal of Marketing, 69 (4), 26–43.hankar, Venkatesh, Amy K. Smith and Arvind Rangaswamy (2003), “Cus-

tomer Satisfaction and Loyalty in Online and Offline Environments,”International Journal of Research in Marketing, 20, 153–75.

öderlund, Magnus (2002), “Customer Familiarity and Its Effects on Satis-faction and Behavioral Intentions,” Psychology and Marketing, 19 (10),
861–79.
uh, Jung-Chae and Youjae Yi (2006), “When Brand Attitudes Affect theCustomer Satisfaction–Loyalty Relation: The Moderating Role of ProductInvolvement,” Journal of Consumer Psychology, 16 (2), 145–55.

Z

ling 89 (3, 2013) 246–262

zymanski, David M. and David H. Henard (2001), “Customer Satisfaction:A Meta-analysis of the Empirical Evidence,” Journal of the Academy ofMarketing Science, 29, 1–16.

ncles, Mark, Andrew Ehrenberg and Kathy Hammond (1995), “Patterns ofBuyer Behavior: Regularities, Models, and Extensions (in Stochastic ChoicePhenomena),” Marketing Science, 14 (3), G71–8.

ncles, Mark D., Grahame R. Dowling and Kathy Hammond (2003), “CustomerLoyalty and Customer Loyalty Programs,” Journal of Consumer Marketing,20 (4), 294–316.

erhoef, Peter C. (2003), “Understanding the Effect of Customer RelationshipManagement Efforts on Customer Retention and Customer Share Develop-ment,” Journal of Marketing, 67 (October), 30–45.

erhoef, Peter C., Philip Hans Franses and Janny C. Hoekstra (2001), “TheImpact of Satisfaction and Payment Equity on Cross Buying: A DynamicModel for a Multi Service Provider,” Journal of Retailing, 77 (3), 359–78.

, and(2002), “The Effect of Relational Constructs on Customer Referrals andNumber of Services Purchased from a Multiservice Provider: Does Age ofRelationship Matter?,” Journal of the Academy of Marketing Science, 30 (3),202–16.

esel, Patrick and Vesna Zabkar (2009), “Managing Customer Loyalty ThroughMediating Role of Satisfaction in the DIY Retail Loyalty Program,” Journalof Retailing and Consumer Services, 16 (5), 396–406.

angenheim, Florian (2005), “Postswitching Negative Word of Mouth,” Jour-nal of Service Research, 8 (August), 67–78.

angenheim, Florian and Tomás Bayón (2003), “Satisfaction, Loyalty andWord of Mouth Within the Customer Base of a Utility Provider: DifferencesBetween Stayers, Switchers and Referral Switchers,” Journal of ConsumerBehavior, 3 (1), 211–20.

i, Youjae and Suna La (2004), “What Influences the Relationship BetweenCustomer Satisfaction and Repurchase Intention? Investigating the Effects ofAdjusted Expectations and Customer Loyalty,” Psychology and Marketing,21 (5), 351–73.

mentation Model and Analysis,” Journal of Business Research, 44, 75–92.aichkowsky, Judith L. (1985), “Measuring the Involvement Construct,” Jour-

nal of Consumer Research, 12 (December), 341–52.

Reversing the Logic: The Path to Profitability through Relationship Marketing

Documents