See discussions, stats, and author profiles for this publication at: [Link]

net/publication/248768479

The Behavioral Consequences of Service Quality

Article  in  Journal of Marketing · April 1996

DOI: 10.2307/1251929

CITATIONS READS

6,358 37,212

3 authors, including:

Valarie A. Zeithaml Leonard L Berry

University of North Carolina at Chapel Hill Texas A&M University
114 PUBLICATIONS   86,063 CITATIONS    147 PUBLICATIONS   65,749 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Cost-Effective Service Excellence View project

Digitally Rich Experiences View project

All content following this page was uploaded by Valarie A. Zeithaml on 14 October 2014.

The user has requested enhancement of the downloaded file.

 ' ValarieA. ZeithamI, Leonard L Berry, & A. Parasuraman

The Behavioral Consequences of

Service Quality
If service quality relates to retention of customers at the aggregate level, as other research has indicated, then ev-
idence of its Impact on customers' behavioral responses should be detectable. The authors offer a conceptual
modei of the impact of service quality on particular behaviors that signai whether customers remain with or defect
from a company. Results from a multicompany empirical study examining relationships from the model concerning
customers' behavioral intentions show strong evidence of their being influenced by service quality. The findings
also reveal differences in the nature of the quality-intentions link across different dimensions of behavioral inten-
tions. The authors' discussion centers on ways the results and research approach of their study can be helpful to
researchers and managers. ...

D
elivering quality service is considered an essential Research on the relationship between service quality and
strategy for success and survival in today's competi- profits has begun to accumulate, and one thing is clear: The
tive environment (Dawkins and Reichheld 1990; link between service quality and profits is neither straight-
Parasuraman, ZeithamI, and Ben7 1985; Reichheld and forward nor simple (Greising 1994; Zahorik and Rust 1992).
[Link] 1990; ZeithamI, Parasuraman, and Berry 1990). Dur- The intermediate links between service quality and profits
ing the 1980s, the primary emphasis of both academic and have not been well understood. To delineate the complex re-
managerial effort focused on determining what service qual- lationship between these two variables, researchers and
ity meant to customers and developing strategies to meet cus- managers must investigate and understand many other rela-
tomer expectations (e.g., Parasuraman, ZeithamI, and Berry tionships, each of which is an integral part of the composite.
1985. 1988). Since then, many organizations—including One such relationship—between service quality and behav-
those whose primary offerings involve physical goods such ioral intention.s—is the primary focus of our present re-
as automobiles or computers^—have instituted measurement search. In the remainder of this introductory section, we pro-
and management approaches to improve their service. The
vide a general overview of the extant knowledge about the
service-quality agenda has now shifted and reconfigured to
link between service quality and profits. We then outline our
include other issues. The issue of highest priority today in-
specific objectives and how our study attempts to extend
volves understanding the impact of service quality on profit
current knowledge.
and other financial outcomes of the organization (Greising
1994; Rust, Zahorik, and Keiningham 1995). Seminal studies using the PIMS (Profit Impact of Mar-
ket Strategy) data set have uncovered significant associa-
Executives of many companies in the 1980s were will- tions among service quality, marketing variables, and prof-
ing to trust their intuitive sense that better service would itability. Findings from these studies show that companies
lead to improved financial success and thus committed re-
offering superior service achieve higher-than-nonnal market
sources to improving service prior to having documentation
share growth (Buzzell and Gale 1987), that the mechanisms
of the financial payoff. Some of these companies, such as
by which service quality infiuences profits include increased
Federal Express and Xerox, have been richly rewarded for
market share and premium prices (Phillips, Chang, and
their efforts (Germano 1992; Reams and Nadler 1992). But
executives in other companies have been reluctant to invest Buzzel! 1983), and that businesses in the top quintile of rel-
in service improvements without solid evidence of their fi- ative service quality on average realize an 8% higher price
nancial soundness. And in the current era of downsizing and than their competitors (Gale 1992). Evidence from compa-
streamlining, interest in tools to ascertain and monitor the nies large enough to have multiple outlets also suggest a
payoff from service investments is high. positive quality-profitability relationship: The Hospital Cor-
poration of America found a strong link between perceived
quality of patient care and profitability across its many hos-
Valarie A. Zeithami is Principal, Partners for Service Excellence, a consult- pitals (Koska 1990); and the Ford Motor Company has
ing firm specializing in strategy, measurement, and implementation of ser-
demonstrated that dealers with high service-quality scores
vice quality. Leonard L. Berry is JCPenney Chair of Retailing Studies and
Professor of Marketing, Texas A&M University. A. Parasuraman is Professor have higher-than-normal profit, retum on investment, and
and Holder of the James W. McLamore Chair in Marketing, University of profit per new vehicle sold (Ford Motor Company 1990).
Miami. The authors thank the editor and five anonymous JM reviewers tor
Although the previous findings document the financial
their constructive comments and suggestions on earlier drafts of this arti-
cle. They also thank the Marketing Science Institute and four of its corpo- and strategic impact of service quality across firms or out-
rate sponsors for supporting the research on which this article is based. lets, the evidence is often too general to answer the ques-
tions foremost in executives' minds: If I invest in service

Journal of Marketing
Vol. 60 (April 1996), 31-46 Service Quality/31
quality, will it pay off for my company? How will service 4. To suggest a research agenda whereby information about in-
quality pay off"? How much should we invest in service qual- dividual-level behavioral consequences of service quality
ity to receive the best retum? In addressing such questions, can t>e monitored and linked to sales and customer-retention
data to provide ongoing evidence of the financial impact of
researchers (Fomell and Wemerfelt 1987. 1988; Rust and
service quality.
Zahorik 1993; Zahorik and Rust 1992) distinguish between
offensive effects (capturing new customers) and defensive In addressing these objectives, we provide a concise
effects (retaining customers). Determining the offensive im- synthesis of the extant literature on the subject and extend
pact of service quality parallels the age-old search for the the literature in three significant ways. First, our study in-
advertising-sales connection. Service quality's effect.s— volves a comprehensive (multicompany/multi-industry) ex-
similar to advertising's effects—are cumulative, and there- amination of service quality's impact at the individual-con-
fore evidence of the link may develop slowly. And, similar sumer level rather than at the company/industry level, as is
to advertising, service quality is one of many variables—in- the case in most previous studies. Second, in addition to ex-
cluding pricing, advertising, efficiency, and image—that si- amining the general relafionship between service quality
multaneously influence profits. Furthermore, spending on and behavioral intentions, we explore changes in the
service per se does not guarantee results, because strategy strength of this relationship that are due to potential moder-
and execution must both be considered. ating effects of different levels of service relative to cus-
On the other hand, evaluating the defensive impact of tomers' expectation levels. Third, we incorporate a more ex-
service quality through customer retention promises to help tensive multiple-item behavioral-intentions measure than
companies gauge the financial impact of service quality. The has been used in previous research and examine service
relationship between retention and profits recently has been quality's impact on specific types of behavioral intentions.
estimated by a variety of researchers (e.g., Anderson and
Sullivan 1990; Fomell and Wemerfelt 1987, 1988; Reich-
held and Sasser 1990) and companies (e.g., IBM). If the re- Conceptual Framework and
lationship between service quality and retention can be sim- Hypotheses
ilarly documented, the financial implications for a given
company or even a given service initiative can be calibrat- Background
ed. Zahorik and Rust (1992) distinguish among five tasks Lowering customer defection rates can be profitable to com-
that must be completed to model the impact of service on panies. In fact, research has shown that it is a more prof-
profits: (I) identifying the key service attributes to include itable strategy than gaining market share or reducing costs.'
in the model, (2) selecting the most important attributes, (3) For exatnple, in an empirical study linking customer satis-
modeling the link between programs and attitudes, (4) mod- faction to profits, Fomell and Wemerfelt (1987, 1988) ex-
eling behavioral response to service programs, and (5) mod- amine the impact of complaint-handling programs on cus-
eling the impact of service programs on profits. tomer retention and conclude that marketing resources are
The research we describe involves the first four tasks better spent keeping existing customers than attracting new
that Zahorik and Rust (1992) propose and concentrates on ones. In support of this position. Reichheld and Sasser
the fourth, namely, modeling behavioral response to quality (1990, p. 105) assert that customer defections have a
service. All four of tbese tasks are firmly in the domain of stronger impact on a company's profits than "scale, market
share, unit costs, and many other factors usually associated
marketing and the first three have been studied extensively
with competitive advantage." For this reason, they extol the
in the last decade (for a review, see Zahorik and Rust 1992).
benefits of zero customer defections as an overall company
In contrast, the fourth attribute, the impact of service quali-
performance standard:
ty on behavioral response, has been the subject of only a few
marketing studies to date (Boulding et al. 1993; Cronin and Ultimately, defections should be a key performance mea-
Taylor 1992). sure for senior management and a fundamental component
The underlying premise of our article is that if service of incentive systems. Managers should know the compa-
quality relates to retention of customers at the aggregate ny's defection rate, what happens to profits when the rate
(i.e., firm) level, as other research has suggested, then evi- moves up or down, and why defections occur (p. 111).
dence of its impact on customers' behavioral responses Research and company efforts to quantify the financial impact
should be detectable. The consequences of service-quality of defection and retention have intensified in recent years.
perceptions on individual-level behavioral intentions can be
viewed as signals of retention or defection and are desirable Financial impact of defection. When customers are lost,
to monitor. With that in mind, our objectives are four-fold: new ones must be attracted to replace them, and replacement
comes at a high cost. Capturing new customers is expensive
1. To -summarize existing evidence about the behavioral con-
sequences of service quality at the individual customer 'This is not to say that companies should focus on customer re-
level. tention to the exclusion of strategies to attract new customers. For
2. To otTer a conceptual model of the impact of service quality instance, share-building strategies should be a high priority for
on particular behaviors that signal whether customers re- companies that are new entrants or operate in emerging markets.
main with or detect from the company. However, for companies wilh an established customer base (espe-
3. To report the results of an empirical study examining rela- cially in mature markets with entrenched competitors) the net re-
tionships between service quality and customers' behavioral turn on investments could be much higher for retention strategies
intentions. than for strategies lo attract new customers.

32 / Journal of Marketing, April 1996

 FIGURE 1
The Behavioral and Financial Consequences of Service Quality

±i
SERVICE BEHAVIORAL Ongoing Revenue
QUALITY INTENTIONS Increased Spending
Price Premium
Remain —^ Referred Customers
BEHAVIOR FINANCIAL
CONSEQUENCES
Defect —^ =i
Decreased Spending
Lost Customers
Costs to Attract New
Focus of present study
Customers
Empirical links demonstrated in macro
studies

for it involves advertising, promotion, and sales costs, as a conceptual model focusing on individual-level behavioral
well as start-up operating expenses. New customers are consequences of service quality.
often unprofitable for a period of time after acquisition: In
the insurance industry, for example, the insurer typically A Model of the Behavioral Consequences of
does not recover selling costs until the third or fourth year of Service Quality
the relationship. Capturing customers from other companies Figure I is a conceptual model that depicts the behavioral
is also an expensive proposition: Anderson and Sullivan consequences of service quality as intervening variables he-
{1990) find that a greater degree of service improvement is tween service quality and the financial gains or losses from
necessary to make a customer switch from a competitor than retention or defection. The left portion of the model is at the
to retain a current customer. level of the individual customer and proposes that setrice
Financial impact of retention. The longevity of a cus- quality and behavioral intentions are related and. thus, that
tomer's relationship favorahly inHuences profitability. Cus- service quality is a determinant of whether a customer ulti-
mately remains with or defects from a company.
tomers who remain with a firm for a period of years hecause
they are pleased with the service are more likely than short- Starting on the left, the model begins with a customer's
assessment of service quality and posits that when service
term customers to buy additional services and spread favor-
quality assessments are high, the customer's behavioral inten-
able word-of-mouth communication. The firm also may he
tions are favorable, whicb strengthetis his or her relationship
ahle to charge a higher price than other companies charge,
with the company. When service quality assessments are low.
because these custotners value maititaining the relationship.
the customer's behavioral intentions are unfavorable and the
The initial costs of attracting and establishing these cus-
relationship is more likely to be weakened. Behavioral inten-
tomers have already been absorbed and. due to experience- tions can be viewed as indicators that signal whether cus-
curve effects, they often can be served more efficiently (Re- tomers will remain with or defect from tbe company.
ichheld and [Link] 1990). Rose (1990) supports this view, Some of the links in Figure I (shown by dotted arrows)
contending that profit on credit card services purchased by a bave been demonstrated empirically in several aggregate-
ten-year customer is on average three times greater than for level studies using overall multicompany analysis (e.g.,
a five-year customer. Buzzeil and Gale 1987; Gale 1992; Reichheld and Sasser
Although the financial impacts of defection and reten- 1990). However, the mediating roles of behavioral inten-
tion have been studied at a macro level (i.e., company or in- tions and actual behavior on the relationship between ser-
dustry level), the micro-level (i.e., individual-level) process- vice quality and financial pertormance are not well under-
es through which these impacts occur have not been well stood, especially at the individual-customer level. We at-
understood. To attempt to fill this void, we develop and test tempt to add to our knowledge in this regard by undertaking

Service Quality / 33
an in-depth conceptual and empirical examination of the By integrating research findings and anecdotal evidence,
first link in the sequence of effects posited in Figure 1. As a list of specific indicators of favorable behavioral inten-
we [Link] in subsequent sections, multiple measures of ser- tions can be compiled. These include saying positive things
vice quality and behavioral intentions were operationalized about the company to others (Boulding et al. 1993), recom-
and used in surveys of customers from four different com- mending the company or service to others (Parasuraman.
panies. For ease of exposition in this section, the dependent Berry, and Zeithaml 1991a; Parasuraman. Zeithaml. and
construct is split broadly into favorable and unfavorable be- Berry 1988; Reichheld and Sasser 1990), paying a price pre-
havioral intentions. mium to the company, and remaining loyal to the company
(LaBarberaand Mazursky 1983; Newman and Werbel 1973;
Favorable behavioral intentions. Certain behaviors sig- Rust and Zahorik 1993). Loyalty may be manifested tn mul-
nal that customers are forging bonds with a company. When tiple ways; for example, by expressing a preference for a
customers praise the firm, express preference for the compa- company over others, by continuing to purcbase from it. or
ny over others, increase the volume of their purchases, or by increasing business with it in the future.
agreeably pay a price premium, they are indicating behav-
iorally that they are bonding with the company. Recent re- Unfavorable behavioral intentions. Customers perceiv-
search offers some evidence that customer satisfaction and/or ing service performance to be inferior are likely to exhibit
service-quality perceptions positively affect intentions to he- bebaviors signaling tbey are poised to leave tbe company or
have in these ways. However, most of the research opera- spend less with the company. Tbese behaviors include com-
tionalizes behavioral intentions in a unidimensional way plaining, which is viewed by many researchers as a combi-
nation of negative responses that stem from dissatisfaction
rather than delineate specific types of behavior. For example.
and predict or accompany defection (Riehins 1983;
Cronin and Taylor (1992), using a single-item purchase-in-
Scaglione 1988).
tention scale, find a positive correlation with service quality
and custotner satisfaction. Anderson and Sullivan (1990). in Complaining behavior itself is conceptualized as multi-
analyzing data from a study of customer satisfaction among faceted. According to Singh (1988), dissatisfaction leads to
consumer-complaining behavior (CCB) that is tnanifested in
Swedish consumers, find that stated repurchase intention is
voice responses (such as seeking redress from the seller),
strongly related to stated satisfaction across product cate-
private responses (negative word-of-mouth communica-
gories. A study conducted hy Woodside. Frey, and Daly
tion), or third-party responses (taking legal action). His
(1989) uncovers a significant association between overall pa-
three-dimensional typology of complaining bebavior.
tient satisfaction and intent to choose the hospital again.
founded on the object of the complaints (seller, friend, third
Several studies have examined the association between party), is statistically superior to previous models of CCB.
service quality and more specific behavioral intentions. In Maute and Forrester (1993) find strong supptm for a three-
previous studies (see Parasuratnan. Berry, and Zeilhaml way classification of dissatisfaction responses based on Hir-
I99la; Parasuraman. Zeithaml, and Berry 1988). we find a sbman's (1970) exit, voice, and loyalty responses (loyalty
positive and significant relationship between customers' being tbe decision to remain with the company despite dis-
perceptions of service quality and their willingness to rec- satisfaction). Solnick and Hemenway (1992) [Link] that
ommend the company. Boulding and colleagues (1993). in though voice and exit (in their view the two main hebavioral
one of two studies they conducted, find a positive correla- manifestations of dissatisfaction) can be substitutes for each
tion between service quality and a 2-item measure of repur- other, they often occur together. In the context of a health
[Link] intentions and willingness to recommend. In a second maintenance organization, they find tbat complainitig cus-
study involving university students, tbey find strong links tomers were four and one-half times more likely to leave the
between service quality and behavioral intentions that are of plan voluntarily than noncomplaining customers.
strategic itiiportance to the school, including saying positive Specific indicators of unfavorable behavioral intentions
things about tbe school, planning to contribute money to tbe suggested by the preceding discussion include different
class pledge on graduation, and planning to recommend the types of complaining (e.g., complaining to friends or exter-
school to employers as a place from which to recruit. nal agencies) and contemplation of switching to competi-
Individual companies are also monitoring the impact of tors. Another indicator of eventual defection is a decrease in
service quality on selected behavioral intentions. For exam- the amount of business a customer does with a company.
ple. Northwest Airlines found that the preference index (i.e., Differential impact of service-quality levels. Although
the preference for Northwest Airlines as the airline passengers superior service is likely to foster favorable behaviors and
like to fly) increased substantially in 1992, compared to 1991, reduce the likelihood of unfavorable bebaviors, an impor-
following a major company effort to improve service. As tant unresolved issue is the service-quality level tbat com-
measured in random surveys, preference rose in Minneapolis panies must target to have the desired impact on behaviors.
(from 70% to 75%). Detroit (from 49% to 59%). and Mem- How tiiuch service quality is enough to retain customers? Is
phis (from 48% to 63%) (Executive Report on Customer Sat- there a level of service beyond which there are diminishing
isfaction 1992). Toyota found that intent to repurchase a Toy- retums in terms of strengthening behavioral intentions?
ota automobile increased from a base of 37% to 45% with a Does the degree of association between service quality and
positive sales experience, from 37% to 79% with a positive bebavioral intentions change at different quality levels?
service experience, and from 37% to 91% with both positive Little published evidence directly addresses tbese ques-
sales and service experiences (McLaughlin 1993). tions. However, a study by Gale (1992), whicb quantitative-

34 / Journal of Marketing, April 1996

ly [Link] the relationship between level of service quality the slope of the service performance-loyalty relationship,
and willingness to purchase at AT&T, offers some indirect our prior recommendation implies an upward sloping
insight. Of AT&T's customers who rated the company's (rather than tlat) relationship within the zone of tolerance.
overall quality as excellent, over 90% expressed willingness Available evidence suggests that the sensitivity of behav-
to purchase from AT&T again. For customers rating the ser- ioral intentions to changes in service quality is likely to vary
vice as good. fair, or poor, the percentages decreased to from below to within to above the zone of tolerance, though
60%, 17%. and 0%. respectively. According to these data, there is no [Link] about the nature of this variation across
willingness to repurchase increased at a steeper rate (i.e., by the three regions of quality. A key empirical question is
43%) as the service-quality rating improved from fair to whether the relationship between behavioral intentions and
good than when it went from poor to fair (17%) or from service quality is flat or upward sloping witbin the zone of
good to excellent (30%). These results suggest that the im- tolerance and. if it is upward sloping, whether or not it is
pact of service quality on willingness to repurchase is most steeper than the relationship below and above the zone.
pronounced in some intermediate level of service quality. The discussion in the preceding sections implies that,
Coyne (1989, p. 73), however, makes the opposite predic- though service quality is positively associated with favor-
tion on the basis of research relating to the impact of customer able behavioral intentions and negatively related to unfavor-
satisfaction with service in a consumer-durable context; able behavioral intentions, customers' perceptions of the
There appear to be thresholds of service for affecting cus- service relative to their adequate and desired service levels
tomer behavior.... When satisfaction rose above a certain moderate these associations. More formally, we posit.
threshold, repurchase loyalty climbed rapidly. In contrast,
when satisfaction fell below a different threshold, cus- H|: The .service quality-tiehavioral intentions relationship (a)
tomer loyalty declined equally rapidly. However, between is positive (negative) tor favorable (unfavorable) behav-
these ihresholds, loyalty was relatively flat. I believe this ioral intentions and (b) has a different slope below and
iwin threshold framework applies to a wide variety of ser- above the zone of tolerance relative to within it.
vice situations.
Impact of problem [Link] and resolution. Another
A similar categorization of service levels follows one de- aspect of service provision that can influence behavioral in-
finition of service quality in the literature—the extent to tentions involves the problem experience of customers.
which a service meets or exceeds customer expectations When customers encounter service problems, these experi-
(Parasuraman, Zeithaml. and Berry 1985, 1988)—^and from ences are likely to affect bebavioral intentions adversely.
recent research explicating the expectations construct as two However, the impact of problem resolution on customers'
levels of expectations (Zeithaml, Berry, and Parasuraman intentions is less clear One view, based primarily on anec-
1993). The first level Is desired service, which Is the level of dotal evidence, is that superior problem resolution forges
service the customer hopes to receive, consisting of a blend of stronger bonds between customers and the company tban
what the customer believes can and should be delivered. Tbe would exist had no service problem occurred. For example,
second, lower level of expectations is adequate .service, J. W. Marriott, chief executive officer of the Marriott hotel
which is the level of service the customer will accept. Ade- chain, states: "Sometimes those [disgruntledl customers
quate service is the minimum service a company can provide whom you make that extra effort to gain back become the
and still hope to meet customers' basic needs. A zone of tol- most loyal customers that you have" (Lovelock 1994, p.
erance, bounded on the lower end by adequate service and on 214). The reasoning underlying this view seems to be tbat a
the upper end by desired service, captures the range of service service problem gives a company the opportunity to demon-
within which a company is meeting customer expectations. strate its commitment to customer service through excellent
Although the zone-of-tolerance framework seems struc- recovery efforts. On the other hand, empirical evidence sug-
turally similar to Coyne's (1989) twin-threshold framework, gests that service failures may weaken the customer-compa-
the managerial implications of the two frameworks are dif- ny bond even when the problem is resolved satisfactorily
ferent. Coyne, invoking the flat satisfaction-loyalty relation- (Boiton and Drew 1992). We report (see Zeithaml. Parasur-
ship he hypothesizes between the two thresholds, suggests aman. and Berry 1990) that customers who experienced no
that unless a company already has a strong reputation for recent service problem with a company have significantly
service, it may not benefit by improving service much be- better service-quality perceptions than customers who expe-
yond the lower threshold: "If a company is already above rienced a recent service problem that was satisfactorily re-
the minimum acceptable threshold, but nearer the lower end solved. A plausible explanation for this finding is that satis-
of the service satisfaction band, investments to incremental- factory problem-resolution service, though perhaps pleasing
ly change position may not be warranted" (p. 75). In con- to customers, does not cause them to forget the service fail-
trast, we have argued previously (see Parasuraman, Berry, ure. And the memory of the failed service negatively affects
and Zeithaml 1991b, p. 47) that firms operating within the customers' overall perception of the company's service. The
zone of tolerance, whiie possibly enjoying competitive ad- existing empirical evidence on this question leads to our
vantage, sbould continue to improve service, even to the second hypothesis:
point of exceeding the desired service level: "To develop a
true customer franchise—unwavering customer loyalty— Hi: Favorable (unfavorable) behavioral intentions are (a) high-
firms must exceed not only the adequate service level but est (lowest) for customers experiencing no service prob-
lem; (b) next highest (lowest) for customers experiencing
also the desired service level." Although we do not refer lo
service problems that are resolved, and (c) lowest (highest)

Service Quality/35
 for customers experiencing service problems that are not Surveys were mailed with a cover letter and postage-
resolved. paid retum envelope to all customers in the sample. The
In summary, the first hypothesis suggests a positive cover letter appeared on company letterhead and was signed
(negative) relationsbip between service quality and favor- by a senior company official. Respondents were requested
able (unfavorable) behavioral intentions, the strength of to retum completed questionnaires to a marketing research
which is different below and above the zone of tolerance rel- company hired to assist with data collection and coding. A
ative to that within it. Hi, along with H2, is depicted in Fig- reminder postcard was sent two weeks after mailing the
ure 2. which details the portion of the behavioral conse- questionnaires.
quences model on which we focus in the present study. Overall response rate was 25% (3069 questionnaires).
Company-specific response rates were 30% (1566 question-
naires) for the computer manufacturer; 22% (522 question-
Methodology naires) for the retail chain; 24% (568 questionnaires) for the
automobile insurer; and 17% (413 questionnaires) for the
Sample Design and Mail Survey life insurer. Demographic profiles of the respondent samples
Four companies that provide services to end or business cus- were reviewed by managers in the respective companies and
tomers were sponsors of the research study. Questionnaires considered to be representative of their customer bases.
were mailed to business customers of a computer manufac-
turer, as well as to end customers of a retail chain, automo- Survey Instrument
bile insurer, and life insurer. The sponsoring companies gen- Operationalization of service quality. Several measures
erated mailing lists from their cunent customer bases. The of service quality were included in the questionnaire: (1) an
retail chain, automobile insurer, and lite insurer each pro- overall, single-item rating scale with anchors at I (extreme-
vided random samples of 24(X) customers. The computer ly poor) and 9 (extremely good); (2) a multiple-item scale of
manufacturer provided a larger random sample of 5270 cus- perceived service from an expanded version of the
tomers, because it wanted to conduct its own detailed, seg- SERVQUAL scale we originally developed (see Parasura-
ment-by-segment analysis following the completion of tbe man, Zeithaml, and Berry 1988) and later refined (see Para-
main study. A total of 12,470 questionnaires were mailed. suraman, Berry, and Zeithaml 1991a); and (3) two categori-

FIGURE 2
Hypothesized Effects of Service Quality on Behavioral Intentions

Performance
Relative to Adequate
and Desired Service

Service Quality Behavioral Intentions

J

Perceived Service Favorable

Performance High
* Say positive things
Medium • Recommend company
Problem Experience - Remain loyal to company
- Spend more with company
and Resolution Low - Pay price premium

Experienced
Problem? No High Unfavorable
1 * Say negative things
Yes Yes Medium - Switch to another company
- Complain to external agencies
Low - Do less business with company
Problem H2.
No
Resolved?
~^

36 / Journal of Marketing, April 1996

cal questions to measure whether respondents had experi- tentions when service problems occur. The 13 items were
enced a recent service problem with tbe company and, if so, grouped into four a priori categories: word-of-moutb com-
whether the problem was resolved to their satisfaction. munications, purchase intentions, price sensitivity, and com-
Tbe second measure (i.e., the revised SERVQUAL bat- plaining bebavior. (These groupings were not made known
tery) represented tbe service dimensions of reliability (five to respondents.) The last two categories contained items not
items), responsiveness (tbree items), assurance (four items), included in prior service-quality research. Eacb item was ac-
empatby (four items), and tangibles (five items). Consistent companied by a 7-point likelihood scale (I = not at all like-
witb the expanded conceptualization of customers' service ly, and 7 = extremely likely).
expectations (ZeithamI, Berry, and Parasuraman 1993). re-
spondents were asked to indicate tbeir adequate- and de-
sired-service levels in addition to their perceptions of each Analyses, Results, and Discussion
SERVQUAL item. Thus, separate ratings of adequate, de-
sired, and perceived service were obtained on tbree 9-point Dimensions of Behavioral intentions
scales (1 = low, 9 = high) arranged as three adjacent Factor analysis of the behavioral-intentions battery was con-
columns next to the SERVQUAL battery on the ducted to examine the dimensionality of the items. Because
[Link].- the battery was designed to represent four categories of be-
Tbe questionnaire containing tbe SERVQUAL battery havioral intentions, a four-factor solution was obtained sep-
with the three columns of ratings used in this study was one arately for eacb company and subjected to oblique rotation
of three different questionnaire formats evaluated in a larg- to allow for potential cotrelation among the categories. The
er methodological study (Parasuraman, ZeithamI, and Berry item clusters implied by the factor loadings differed from
1994a). As such, the adequate-, desired-, and perceived-ser- the a priori clusters and varied somewbat across the four
vice scores used in tbe present study were based on a partial companies, Tbe general pattems of loadings suggested tbat
sample from each company (the otber two questionnaire for- a five-factor solution may belp reconcile these differences.
mats did not produce separate scores for tbese variables). A A five-factor solution produced an unambiguous factor pat-
total of 1009 questionnaires contained scores for tbe ade- tem that was consistent across ail companies. This consis-
quate-, desired-, and perceived-service variables: 498 from tent pattem suggested a reconfiguration ofthe 13 items into
the computer manufacturer, 188 from the retail chain, 191 five dimensions: loyalty to company (loyalty), propensity to
from the automobile insurer, and 132 from the life insurer. switch (switch), willingness to pay more (pay more), exter-
All three questionnaire formats contained measures for the nal response to problem (external response), and intemal re-
remaining study variables (overall service quality, behav- sponse to problem (internal response). In Table I. we present
ioral intentions, and incidence of service-problem experi- tbe reconfigured bebavioral-intentions battery, and in Table
ence and satisfactory problem resolution). Therefore, scores 2, we present the factor-load ing matrices supporting it,
for these variables were based on the full sample. along with reliability coefficients for its multiple-item
components.
Operationalization of behavioral intentions. Previous re-
search bas not captured the full range of potential behaviors Of the five factors, loyalty (with five items) and pay
more (witb two items) exbibit consistent pattems of load-
likely to be triggered by service quality, Cronin and Taylor
ings across the four companies. Switch (with two items) and
(1992) focus solely on purcbase intentions and measure tbe
extemal response (witb tbree items) also display a moderate
construct witb a single-item scale. In tbe tlrst of two studies
to high degree of uniformity in factor loadings. The final di-
by Boulding and colleagues (1993), repurcbase intentions
mension, intemal response, contains just one item that loads
and willingness to recommend were the only two behavioral on tbe fifth factor.
intentions measured. In tbe second study, involving service
quality of an educational institution, tbey used a 6-item scale Tbe 5-item loyalty scale has excellent intemal consis-
comprised largely of education-specitlc items, such as intent tency, which is evidenced by alpbas ranging from .93 to .94
to contribute money to the class pledge and intent to recom- across tbe four companies. The 3-item extemal-response
mend tbe school to employers as a place to recruit. scale bas alphas of at least .6, with tbe values in two of tbe
A 13-item battery was developed to gauge a wider range four companies exceeding the threshold of ,7 thai Nunnally
of behavioral intentions tban have been suggested in the lit- (1978) suggested. The 2-item scales measuring switch and
erature or by anecdotal evidence from companies.^ This bat- pay more have somewhat weaker alphas, with several values
tery included items to capture several facets of behavioral falling below .6, perhaps because of too tew items in tbe re-
intentions not incorporated in previous service-quality stud- configured factors. In general, the alpha score for loyalty is
ies: likelihood of paying a price premium and remaining high, particularly for an early study, Tbe alpba scores tor tbe
loyal to a company even when its prices go up, intent to do other three factors with multiple items range from adequate
more business with the firm in the future, and complaint in- to weak, indicating the need to add items to tbe scale in fur-
ther research.
Although the factor structure of the behavioral-inten-
e adequate- and desired-level service ratings were used in tions battery differs somewbat from the a priori specifica-
defining the lower and upper boundaries of the ^one of tolerance to
tion, the loadings suppoti the dichotomy in behavioral in-
verify the tiiflerential slopes predicted by H||, tor the quality-in-
tentions relationship. tentions of favorable and unfavorable categories. The largest
^A copy of the instrument containing the behavioral-intentions factor, loyalty, contains five favorable behavioral-intentions
and service-quality question.s can be obtained from the third author. items: saying positive things about the company, recom-

Service Quality / 37
 TABLE 1
Behavioral-Intentions Battery^
Behavioral-
Intentions Item
Dimension Label Item Wording

Loyalty 11 Say positive things about XYZ to other people.

12 Recommend XYZ to someone who seeks your advice.
13 Encourage friends and relatives to do business with XYZ.
14 Consider XYZ your first choice to buy services.
15 Do more business with XYZ in the next few years.
Switch 16 Do less business with XYZ in the next few years (-).
17 Take some of your business to a competitor that offers better prices (-).
Pay More 18 Continue to do business with XYZ if its prices increase somewhat.
19 Pay a higher price than competitors charge for the benefits you currently
receive from XYZ.
no Switch to a competitor if you experience a problem with XYZ's service.
External Response Complain to other customers if you experience a problem with XYZ's service.
111
112 Complain to external agencies, such as the Better Business Bureau, if you
experience a problem with XYZ's service.
113 Complain to XYZ's employees if you experience a problem with XYZ's service.
Internal Response
^The items were grouped as follows in the a priori categorization of the battery: Word-of-Mouth Communications—II, 12, 13; Purchase Inten-
tions — 14,15,16; Price Sensitivity — 17,18, 19; Complaining Behavior — 110, 111, 112,113, Each item was accompanied by a 7-point likelihood
scale (1 = not at ait likely and 7 = extremely likely), items identified with a "-" were reverse scored.

mending the company to someone who seeks advice, en- zone of tolerance relative to within it. This hypothesis was
couraging friends and relatives to do business with the com- tested by using multiple regression analysis to examine si-
pany, considering the company the first choice from which multaneously (I) whether the slope of the relationship with-
to buy services, and doing more business with the company in the zone of tolerance was significantly different from zero
in the next few years. Pay more contains two favorable and (2) whether this slope differed significantly from the
items: continuing to do business with the company even if slopes below and above the zone of tolerance. In accordance
its prices [Link] somewhat and paying a higher price than with procedures discussed by Cohen and Cohen (1983,
competitors charge for the benefits currently received from Chapter 8) for conducting this type of analysis, the follow-
the company. ing regression equation was estimated:
The second and fourth factors comprise all unfavorable (I) Y= B,X
behavioral-intentions items. Switch contains two of these:
doing less business with the company in the next few years where
and taking some business to a competitor that offers better Y = behavioral-intentions score;
prices. Extemal response includes items that relate to expe- X = service-quality score;
riencing a service problem: switching to a competitor, com- dl = dummy variable with a value of 1 if the perceived service is
plaining to other customers, and complaining to extemal below the zone of tolerance. 0 otherwise;
agencies such as the Better Business Bureau. d2 = dummy variable with a value of 1 if the perceived service is
The interpretation of intemal response, the fifth factor above the zone of tolerance, 0 otherwise;
with one item (complaining to the company's employees if Bs = unstandardized regression coefficients; and
a service problem Is experienced), is unclear. Customers
€ = error term.
more favorably disposed toward a company may be more
likely to complain intemally to give the company a "second The coefficients in Equation 1 that are relevant for examin-
chance." Conversely, disgruntled customers with an unfa- ing the first hypothesis are B,, B,, and B,. Specifically, B,
vorable image of the company may be more likely to com- represents the slope of the quality-Intentions relationship
plain intemally to vent their frustrations. The equivocal in- within the zone of tolerance, whereas Bi and B, represent
terpretation of this factor and its being represented by just changes in B, below and above the zone of tolerance, re-
one item undennine its meaningfulness on conceptual and spectively. Thus, B, + BT represents the slope below the
psychometric grounds. As such, we deleted this single-item zone, and B, + B3 represents the slope above the zone.
measure from all subsequent analyses. Service quality (the key independent variable X) was
operationalized in two ways: as the rating on the 9-point
Relationship Between Service Quality and overall-quality (OQ) scale and as a weighted-average per-
Behavioral intentions ceived perfonnance (WP) score across the SERVQUAL di-
The first hypothesis predicted a positive (negative) quality- mensions. Of late there has been debate in the literature
intentions relationship for favorable (unfavorable) behav- about the most appropriate way to operationalize service
ioral intentions, with different slopes below and above the quality (cf Brown, Churchill, and Peter 1993; Cronin and

38 / Journal of Marketing, April 1996

 13 C

a>
Q.
E 2
a
0)
•a
9
c
E
o
o
EZ
a
(0 CO
> CVI
c
o 2.
'« in c
u. O) o
c
0) c
t> (5
E 03
u.
o d> o
tf) •-1
o
Life Insi

c
ehaviiora I-In1tent ion

act
o
•c
CM
U. o c
<u
o

itii
(D ffl
[.93]

c n
("
01 U-

3
o (A u.
«^ C

(0 £1 (0
JZ a U.

a. o C) o
E
o
3 s:
m £
< -Si
EZ
effi

[•94]

o vO O
o
>.
:^ c
'm
!o .c "^
.2 O
n
OC
0)
n s-s
i
a>
•D CC et
C
(0
Ii.
s
tf) ? - o
0) u. bi f i CO
o
•o i i
Mati

[.94]

^- o
« (I) Q.
D) £
C 3
O G 2 c
'S
n 3 E E8 r^ S
o C
(D o
-I E
C
autO:
lrouQ

o :5
3
e
CO
- —
I 1 Q. SJ 'c^
E
o XT
o LZ (D ra "ai
aii

i2 Q .Q
n j Q) i5
o (A
CO
S O

% "a"
B-iltern

illy[

XJ O
E (/I
o .£ Q.m

Service Quality / 39
 Taylor 1992; Parasuraman. Berry, and ZeitbamI 1993; Para- switch and extemal response—and, with few exceptions, are
suraman, ZeitbamI, and Berry 1994b; Teas 1993). Tbe cen- statistically significant at/) < .01. The pattem of B| values
tral issue in this debate is whether service quality should be across the four companies suggests tbat the effects are gen-
measured as the difference between customers" perceptions erally stronger for loyalty and switch tban for pay more and
and expectations ratings or simply as the perceptions rat- extemal response. Tbe results for the combined sample pro-
ings. Altbougb tbis issue continues to be debated, there is vide additional insight into the relative infiuences of service
some agreement that a study's purpose may influence the quality on the four behavioral-intentions dimensions: the
cboice of which measure to use: The perceptions-only oper- strongest effects of both WP and OQ are on loyalty (.70 and
ationalization is appropriate if the primary purpose of mea- .55), followed by switcb (-.67 and -.47), pay more (.43 and
suring service quality is to attempt to explain the variance in .37), and extemal response (-.28 and -.21) in that order.
some dependent construct; the perceptions-minus-expecta-
The regression coefficients in Table 3 in the columns for
tions difference-score measure is appropriate if the primary
B2 and B, values pertain to the differential effects predicted
[Link] is to diagnose accurately service shortfalls (Para-
by H| (the statistically significant BT and B^ coefficients are
suraman, Zeiiliaml. and Berry 1994a). The purpose of our
boldfaced). For eacb bebavioral-intentions dimension wiib-
present study is tbe former. Moreover, as we discuss subse-
in a given company, the B2 coefficients for WP and OQ have
quently, the two expectations measures (i.e., the adequate-
the same sign except in a few instances. Similarly, the signs
and desired-service levels) were independently incorporated
of the B3 coefficients are identical for WP and OQ. The sta-
into the analysis to operationalize the two dummy variables
bility in the signs of tbe slope-change coefficients across
di and di. Therefore, the ratings from tbe SERVQUAL por-
two different service-quality measures is encouraging in
tion ofthe survey were used to operationalize service quali-
terms of drawing inferences about the direction of changes
ty as weighted-average performance scores, ratber than dif-
in the quality-intentions link below and above the zone of
ference scores.
tolerance. However, support for the strength of these
To determine WP, a perceived performance rating was changes is mixed, as is evidenced by tbe pattern of statisti-
first computed lor each SERVQUAL dimension by averag- cal significance of these coefficients. Tberefore, based on
ing the ratings on the items forming the dimension. (The co- the presence of significant coefficients for at least one of the
efficient alpha values for reliability [five items], responsive- two service-quality measures (WP and OQ), only the fol-
ness [three items), assurance [four items), empathy [four lowing inferences seem warranted.
items] and tangibles [five itcnis[ ranged from .80 to .96 In the compu ter-manufacturer sample, tbe quality-inten-
across tbe four samples.) To obtain the WP score, the aver- tions relationship for loyalty and switch is flatter above tbe
age performance ratings for the dimensions were then zone of tolerance (implying diminished sensitivity to quali-
weighted by tbe relative importance of the dimensions. To ty improvements beyond the desired-service level), but is
measure tbe relative importance of tbe five dimensions, re- unchanged below the zone of tolerance. The relationship for
spondents were asked to allocate 100 points among tbe di- pay more is flatter both below and above the zone. In the re-
mensions according to how important each dimension was tail-chain sample, the relationship for loyalty, switch, and
to them in evaluating a company's service. The relative extemai response is flatter below tbe zone of tolerance but
points allocated to the dimensions were used as weights in remains unchanged above the zone of tolerance (implying
computing the WP score. undiminisbed retums for quality beyond the desired service
The dummy variables d| and dj were operationalized by level). In the automobile-insurer sample, the relationship for
comparing each respondent's WP score with his or her loyalty is steeper below the zone of tolerance but remains
weighted-average adequate- and desired-service scores unchanged above tbe zone. However, the relationship for
(computed using a procedure similar to tbat used in deter- switch is fiatter below tbe zone and considerably steeper
mining WP). The d| value was 1 if WP was less tban the above tbe zone, whicb implies that there are increasing pay-
weigbted-average adequate-service score, 0 otherwise. The offs as service improves from below to within to above the
di value was I if WP was greater than the weighted-average zone. Tbe relationship for extemal response in the automo-
desired-service score, 0 otherwise. bile-insurer sample is similar to that in the retail-chain sam-
The regression analysis was performed separately for ple (i.e., fiatter below the zone but unchanged above it). All
the four companies, a.s well as for the combined sample. In of the slope-change coefficients in the life-insurer sample
eacb instance, two equations were estimated for eacb be- are nonsignificant. This lack of significance may be due to
havioral-intentions dimension: one using WP scores and the insufficient data points below and above the zone—only 15
second using OQ scores as values for tbe independent vari- respondents in the life-insurer sample bad WP scores below,
able X. The average score across items comprising the be- and only 8 had WP scores above the zone. A similar defi-
havioral-intentions dimension represented tbe dependent ciency may account for the lack of significance of any of tbe
variable Y. In Table 3, we summarize the regression-analy- B3 coefficients in tbe retail-cbain sample; only 16 respon-
sis results pertaining to the first hypothesis. dent.s had WP scores above the zone.
Tbe regression coefficients in the first two columns of In tbe combined sample, tbe quality-intentions relation-
Table 3 (B| values) offer strong support for tbe hypothesized ship for loyalty and switch is fiatter below but remains un-
qualiiy-intcntions links within the zone of tolerance. The co- changed above the zone. Thus, exceeding tbe adequate-ser-
efficients for WP and OQ are all in tbe hypothesized direc- vice threshold can sharply increase tbe payoffs (in terms of
tions—positive for loyalty and pay more and negative for fostering customer loyalty and curtailing propensity to

40 / Journal of Marketing, April 19%

 in
o
V

v>
<o in en OJ o eg o CO O CO
^ -- o o • ^ CM • • -
CD to ' -

5.5
2«
< 3 V

CO
en o en o O CO
o GO CO CO eo in »- to
o 'T eg 1-
o CO CM 1 - CM
i n CM
OJ to eg o

IO
t-; *-; q
q
V V V V
(O CO tn (0 "co" tn Wl
333 c c c c c c c c
3 c
CO t o OJ CO CO CJ) CO in T- Ol to eg eg o in in CM
OJ CO U l o e y e g CM O <D »— CM CD OJ
o CO

r rr
3
.—. V -—-
to CA tA tn (0 (A tA tn _ tn
c c c c Q. c
n CO ,_ O} in III to ^ CJ CJ i n in o in in
CJ CJ eg CO 1 - t-; T- 1/1 CD CM to ^ T - CJ
I'
(/I r r r I* r I* r • r
3
V> _C

CC 0)
O)
c IO to
.05

.05
as
CO JC o o
O V V V
.05 (ns)

.13 (ns)

.03 (ns)

-.07 (ns)
(su)

(ns)
tu o en to ' ay ta
o.
tn tA
_l c o -SSS a e: 3
en o
.14
.25

^ OJ CVJ at

.06
.52

CQ OJ CO i n
< o o w q T- ^ O Ol CM q T-

r r

to
GC oV •—• V — V
SSB S
O) CO O GO P ) *O CJ CO in in Ui o Tt in Ol IO O
O CJ OJ O <O ^ CM q tt T- in 1 - CO ^
^ 1- O

in
q 0 O
V

C 0)
3 <D O
GO •^ eg t O CO CO in O) t o CM evt r - CO in 1
in in 1- i n CO - - CM •^ to to i n CO CO in CO

r
I!
tf) O
V
c
o in to
in in
CO O) CO to N. en
r^ 00 to in
CM i n O CM o h- CO
N r^ to "* r^ '3" ^ CO l*~- CD ^
r

CD <D 0)
tn o Cfl
0) tn to tr
} adiu:Rted
Res pon
Res pon

)ani
ore

(0
ore
ore

c
Insu rer
}ile

ro ro ro "S
^ o
/
oyalti

xtern

Resi
witch

xtern

witch

JLUOC

xtern
lsure

o ra
a. E Q. m
I E
o •
O
5 ro ^
C/) CL l U
CC
o
3
>.
_oi (/>
to
CL UJ
'3
_ i CO Q.
>s
ro
UJ
<
5 0) Q. LU .c
•0

Service Quality / 41
 TABLE 4
Mean Scores for Service Quality and Behavioral Intentions
Service Qualitya Behavioral Intentions'*

External
Company WP OQ Loyalty Switch Pay More Response

Computer Manufacturer 7.3 6.8 5.3 3.8 3.9 3.6

Retail Chain 6.5 6.2 4.9 3.9 3.2 4.3
Automobile Insurer 7.8 7.3 5.6 3.1 3.5 4.2
Life Insurer 7.6 7.0 5.0 3.1 3.4 3.8
aMean scores on a 9-point scale.
I'Mean scores on a 7-point scale.

switch). However, the combined-sample results for the pay more prone to switch and complain externally—than the au-
more dimension reveal considerable fiattening of the quali- tomobile insurer's customers.
ty-intentions relationship above the zone of tolerance. In
fact, the slope for the OQ-pay more relationship changes Impact of Service-Problem Experience and
from .37 below the desired-service level to just .03 (.37 - Resolution on Behavioral Intentions
.34) above. Thus, companies wishing to improve service be- Hi predicts that customers experiencing no service problems
yond Ihe desired-service level should do so cautiously and have the best behavioral-intentions scores (highest for fa-
cost-effectively, because recouping the added expense by vorable intentions and lowest for unfavorable intentions—
charging price premiums may not be a viable option. The Hia), customers experiencing problems that were resolved
quality-intentions relationship for external response— would have intermediate scores (H^^)- antl customers with
which, as indicated by its B| coefficients, is flatter within unresolved service problems would have the worst scores
the zone than for the other three dimensions—remains un- (H,^.). To test this hypothesis, the combined sample was clas-
changed below and above the zone as well. Thus, relative to sified into three groups of respondents; [Link] experiencing
the other dimensions, extemal response appears much less no recent service problems; those experiencing problems
affected by changes in quality over a wide range. that were resolved; and those experiencing problems that
Tbe pattem of adjusted R-squared values in the last two were not resolved. Analysis of variance was conducted to
determine whether scores on each behavioral-intentions di-
columns of Table 3 offer two noteworthy insights based on
mension differed across the groups. The F-values for all four
the overall ability of service-quality-related variables (d|,
ANOVAs were significant at p < .001. Eight prcspecifed
d2, and X) to explain the variation in scores on each behav-
contrasts (first-group mean versus second group mean and
ioral-intentions dimension. First, the relationship of quality
.second-group mean versus third group mean for each of the
(both WP and OQ) with loyalty and switch is consistently
four behavioral-intentions dimensions) were also evaluated.
stronger in the two pure-service companies (automobile and In Table 5, we present the group means and the significance
life insurers) than in the two product companies (computer levels for the planned contrasts.
manufacturer and retail chain); however, the reverse is true
for the quality-pay more relationship: The relationship is The alpha level for testing the significance of individual
consistently stronger in the two product companies than in contrasts was reduced by applying the Bonferroni correction
the two pure-service companies (for additional analyses, see to ensure that the overall probability of Type I error across
all eight contrasts did not exceed .05 (for details, see foot-
the Appendix). Second, the quality-pay more relationship is
note b in Table 5). The evidence in Table 5 fully supports the
consistently weaker than the quality-loyalty relationship in
second hypothesis for the loyalty, switch, and extemal re-
all four companies and the combined sample. We examine
sponse dimensions, and partially supports it for the pay
the implications of these insights subsequently.
more dimension. The findings clearly show that customers
In Table 4, we summarize the mean scores for service experiencing no service problems have the strongest levels
quality and behavioral intentions by company. An across- of loyalty intentions and the weakest switch and extemal re-
company comparison of the mean-score pattems provides sponse intentions. However, their pay more intentions are
additional support for inferring that service quality is asso- not significantly higher than those of customers experienc-
ciated positively with favorable behavioral intentions and ing service problems that were [Link] satisfactorily.
negatively with unfavorable behavioral intentions. With few Among customers experiencing recent service problems,
exceptions, the better a company's service-quality scores, those receiving satisfactory resolution have significantly
the higher are its loyalty and pay more means and the lower higher loyalty and pay more intentions, and significantly
are its switch and extemal response means. To [Link], the lower switch and extemal response intentions, than those
retail chain's WP and OQ scores are considerably lower than with unresolved problems. Thus, effective service recovery
the corresponding scores for the automobile insurer. Match- significantly improves all facets of behavioral intentions.
ing behavioral-intentions data show that the retail chain's However, with the possible exception of tbe pay more di-
customers are less loyal and less willing to pay more—and mension, the improvements do not restore intentions to the

42 / Journal of Marketing, April 1996

levels expressed by customers not experiencing service viously mentioned, customers favorably disposed toward a
problems. These results are consistent with those from a company may complain to give it a .second chance, while
study in which Bolton and Drew (1992) examine the impact customers unfavorably disposed may also complain merely
of problem experience and resolution on telephone cus- to vent their frustrations. Therefore, in expanding this com-
tomers' evaluation of billing service: Customers rated the ponent, it would be useful (from a diagnostic standpoint) to
service substantially lower if (hey had experienced a billing add items that capture why customers are likely or unlikely
problem, and the effect of satisfactorily resolving the prob- to complain. For example, respondents could be asked to
lem did not completely offset its negative impact. rate the likelihood of the following {on the same 7-point
scale used in the behavioral-intentions battery):
Discussion and Implications • Complain to XYZ"s employees aboul a service problem be-
cause [ am confident they will resolve the problem.
We developed a conceptual model of the behavioral and fi-
nancial consequences of service quality (Figure I. A portion • Complain to XYZ's employees about a .service problem to
help relieve my frustration (reverse scored).
of the model—the quality-intentions link—^was empirically
examined at the individual-customer level in a multicompa- Second, additional research is needed to examine further
ny context. Two distinctive features of the study's empirical our tentative insights pertaining to intercompany differences
component were the development of a more extensive be- in the quality-intentions relationship and the changes in its
havioral-intentions battery than has been used in previous slope below and above the zone of tolerance relative to
research and the investigation of changes in the quality-in- within it. Although the total sample size was large for each
tentions link at different service levels relative to customers' company, the subsamples of respondents below and above
expectations. The study's findings have important implica- the zone were relatively small, and this possibly contributed
tions for researchers and managers. to the lack of significance of some of the slope-change co-
efficients. Obtaining larger samples of respondents below
Directions for Further Research ;^ ': , and above the zone in further studies would facilitate a more
The distinctive aspects of the empirical study contribute robust examination of changes in the quality-intentions rela-
several new insights whose implications we subsequently tionship. One option for doing so is to select samples from
explore. However, our findings also reveal certain weak- companies that are well known for their excellent (or poor)
nesses with methodological implications. First, the behav- service. Another option is to devise a suitable quota-sam-
ioral-intentions battery developed here, though more com- pling procedure to ensure large enough subsamples below,
prehensive than intentions scales used in previous studies, within, and above the zone. Mullicompany research using
needs further development. In particular, more items are such sampling procedures is needed for more definitive con-
needed to strengthen the reliability of three of its compo- clusions about the intriguing differences uncovered in this
nents, namely, switch, pay more, and extemal response. study, which concem changes in the quality-intentions link
With additional items, the scales should be reevaluated for within and across dimensions and companies.
their psychometric properties. Consideration should also be In addition to addressing the previous issues, further re-
given to augmenting and including in the battery intemal re- search should also focus on aspects of the conceptual model
sponse—the component that was eliminated because it had not examined here. For example, the association between
only one item subject to equivocal interpretation. As we pre- behavioral intentions atid remaining with or defecting from

TABLE 5
Mean Behavioral-Intentions Scores for Respondents Classified According to Service Problem Experience
Significance Levels for
Mean Scores for Customers Experiencing^ Planned Ct)ntrastsb
Service Problems Service Problems
Behavioral- No Service That Were That Were Group 1 Group 2
Intentions Problems Resolved Not Resolved Mean versus Mean versus
Dimension (Group 1; n = 2153) (Group 2; n = 455) (Group 3; n = 346) Group 2 Mean Group 3 Mean

Loyalty 5.47 5.01 4.11 -000 .000

Switch 3.35 4 ^ 4.49 .000 " .000
Pay More 3.76 3;63 3.11 .036 .000
External
Response 3.70 3.95 4.43 .000 .000
«The behavioral-intentions scores are on a 7-point likelihood scale.
I'The reported significance levels are for one-tailed tests, because Hj implies directional comparisons of group means. Because multiple con-
trasts were evaluated to test this hypothesis, the Bonferroni correction was applied to the customary alpha level of .05 to control the Type I
error rate. Specifically, the alpha level was lowered by a factor of eight (the total number of planned contrasts) to yield a critical alpha level of
.006 for testing the significance of each contrast (Myers 1979, pp, 298-300). At this reduced alpha level, seven of the eight planned contrasts
are significant; the sole exception is the Group 1 versus Group 2 contrast for the pay more dimension.

Service Quality / 43
(he company merits study. Rust and Zahorik (1993) suggest Managerial Implications
ways to investigate this link, including panel data, longitu-
The overall findings offer strong empirical support for the
dinal analysis with customers, and cross-sectional surveys
intuitive notion that improving service quality can increase
asking customers about previous and current providers. Ad-
favorable behavioral intentions and decrease unfavorable in-
ditional cross-sectional research might ask customers to in-
tentions. The findings demonstrate the importance of strate-
dicate not only their behavioral intentions but also their ac-
gies that can steer behavioral intentions in the right direc-
tual behaviors. For example, customers could be asked
tions, including striving to meet customers' desired-service
whether they have said positive things about the company
levels (rather than merely performing at their adequate-ser-
(actual behaviors) instead of how likely they would be to
vice levels), emphasizing the prevention of service prob-
say positive things (behavioral intentions). Such research
lems, and effectively resolving problems that do occur.
also needs to be supplemented with longitudinal research to
However, multiple findings suggest that companies wanting
verify the causal direction of the quality-intentions link.
to improve service, especially beyond the desired-service
Data from studies tracking service quality and behavioral in-
level, should do so in a cost-effective manner: the quali-
tentions over time can be analyzed to determine the impact
ty-pay more relationship in the combined sample, while up-
of service quality in a given period on behavioral intentions
wardly sloping below the desired service level, becomes vir-
in subsequent periods.
tually flat above that level (Table 3); the adjusted R-squared
If the longitudinal data set also contains information for values for the quality-pay more regressions are weaker than
individual customers on variables such as purchase frequen- for the quality-loyalty regressions across all companies and
cy and volume and new-customer referrals, the impact of for both measures of service quality (Table 3); and, in each
service quality on actual behavior can be traced. Companies of the four companies, the mean score for pay more inten-
that have information systems linking customer data and tions is considerably lower than for loyalty intentions (Table
purchase data could also examine increases or decreases in 4).
spending that result from different levels of service quality.
This type of research would provide direct evidence of the In addition to these general implications, the conceptual
financial impact of service quality at the individual level. model and the empirical findings have specific implications
for firms' research and resource-allocation decisions per-
An intriguing finding worthy of further research is the taining to improving service quality. A salient issue on the
pattem of across-company differences implied by the differ- service quality research agenda of many companies is un-
ences in the adjusted R-squared values for the various qual- derstanding the impact of service quality on profits. One of
ity-intentions regression equations (Table 3). As previously the reasons it has been difficult for individual firms to cali-
highlighted, the quality-intentions link for the loyalty and brate this impact is that the relationship is complex and con-
switch dimensions is consistently stronger for the two pure- sists of multiple intervening relationships. As was suggested
service companies than for the two product companies, in our model of individual customer response, a chain of re-
whereas the reverse is true for pay more. Is it possible that lationships is integral to the overall impact.
the role of service within a firm's total offering (i.e.. core
Companies first must examine the impact of their ser-
versus supplemental component) is a plausible explanation
vice-quality provision on customers' responses, including
for this pattem of differences? Because service is all that a
intentions signaling behaviors that are potentially favorable
pure-service provider, such as a life insurance company, de-
or unfavorable to the company. For most companies, a bat-
livers in exchange for customers" money, customers' com-
tery of behavioral-intentions questions could be incorporat-
mitment to the company might be extremely responsive to
ed easily into the measurement systems currently used to
service-quality improvements; however, these customers'
capture service-quality assessments. Doing so provides a
willingness to pay more may not be as responsive, because
continuous source of infomiation relating to such questions
they may feel they have, in effect, already paid for high-
as.
quality service. Alternatively, the pay more findings may
simply reflect customers' general reluctance to pay for in- •What levels of service quality must we deliver to retain
surance services and may not apply to pure services overall. customers?
In contrast, because service is not the core of what a •What service initiatives should we undertake to encourage
product company sells to customers, their commitment to customers to recommend the company, spend more with the
company, or pay a price premium?
the company may be less sensitive to changes in service
quality (especially if product quality is mediocre); however, •What auributes should we focus on to reduce the likcHhood
customers may be somewhat more willing to pay more for of customers .spreading negative word-of-mouth communica-
tions when service problems occur?
better service, because they may consider service to be an
extra feature. To what extent and under what circumstances • To retain customers, should we spend our money on proac-
tive service improvements or on handling complaints?
are these speculations likely to be true? Furthemiore. what
is the nature and extent of the impact of factors other than In essence, behavioral intentions become dependent vari-
the service component (e.g.. price, product characteristics) ables with potentially higher validity (because they are more
on customers' behavioral intentions? Additional conceptual closely related to actual behaviors) and richer diagnostic
and empirical research addressing these issues can improve value than the "overall service quality" or "customer satis-
our understanding of the behavioral consequences of service faction" variables currently being used in most measure-
quality. ment programs.

44 / Journal of Marketing, April 1996

 Companies also can use surveys eliciting behavioral in- Appendix
tentions as an early warning system to identify customers in
danger of defection and lo take timely corrective action. It To formally test the significance of these and other inter-
has long been speculated that many service customers ex- company differences implied in our discussion, an extended
hibit "spurious loyalty"^—they remain witb companies they version of regression Equation I was estimated for the com-
are dissatisfied with because tbey see no alternatives. Local bined sample:
telephone customers, with no choice in service providers,
are often cited as customers exhibiting spurious loyalty. Be-
eause of the deregulation of that industry—and the immi- B4X +
nent competition among cable companies, long-distance B7X +
companies, and other regional Bell companies—firms sup- B|,,X B|,d,X
plying local telephone service mighl be able to strengthen
customer retention by identifying customers most in danger where
of defection.
gi = dummy variable with a value of I when
Another measurement implication for companies to un- the company is the retail chain. 0
derstand better and leverage the relationships in the concep- otherwise;
tual model (Figure I) is to query customers about wbetber g2 = dummy variable with a value of I when
they have manifested, rather than merely would manifest, the company is the aulomobile insurer, 0
favorable or unfavorable behaviors. Companies sbould also otherwise;
consider connecting their marketing database to their ac- g, = dummy variable with a value of I when
counting information systems to explore the links between the company is ihe life insurer, 0 other-
wise; and
customers' assessments of service quality and tbeir purchase
behavior L. L. Bean has accounting systems that track vita! Y. X, di, di. B's. and € = same as those defined for Equation I.
purchasing pattems of individual customers: what they pur- Wben g|, g2. and g3 are 0. Equation Al pertains to the
chase, when they buy. and how much they spend over time. computer manufacturer, witb B, representing tbe slope of the
Companies witb such systems could benefit from integrat- qitality-intentions relationship within the tolerance zone and
ing their customer-research database witb actual purcbase B2 and B3 representing changes in thai slope below and
data lo enhance their knowledge of how and to wbat extent above the zone, respectively. Using the computer manufac-
behavioral intentions act as precursors of purchasing turer as the reference company. B4 through B^,. B7 through
behavior. By, and Bm through B|2 are the corresponding slope-change
Yet another facet of ihe conceptual model that merits coefficients for the retail chain, aulomobile insurer, and life
thorough analysis at the individual company level involves insurer, respectively. As in the case of Equation I. Equation
actual increases or decreases in revenue from retention or AI was estimated for each of the four behavioral-intentions
defection of customers. Reichheld and Sasser (1990) call for dimensions. The results show consistent statistical signifi-
accounting systems that capture the expected cash flows cance (atp < .05) only for B,. With a few exceptions (that re-
from and life-time value of a loyal customer. This call could vealed no meaningful pattem). the slope-change coefficients
be expanded to include an examination of the revenue im- B2 through B|2 are not significant. An inadequate number of
plications of individual bebavioraUintentions components sample units with nonzero values for the variables corre-
as well. For example, what is the potential revenue differen- sponding to these eleven coefficients (and the consequent
tial between customers indicating a low versus a high low variance for these variables) is a bighly plausible expla-
propensity to switch? Is there a difference in the revenue nation for this lack of significance: Tbe percentage of re-
streams from customers expressing a strong versus a weak spondents (relative to the total sample) witb nonzero values
inclination to pay a price premium for a company's service? was less tban 5% for five of tbe variables, less tban 10% for
Examining such questions will facilitate a more detailed cal- another five variables, and less than 20% for the remaining
ibration and a clearer understanding of tbe financial conse- one variable. A set of tables summarizing these regression re-
quences of service quality. sults can be obtained from tbe third author.

REFERENCES
Anderson. Eugene W. and Mary W. Sullivan (1990). "Customer Brown. Tom J., Gilbert A. Churchill. Jr.. and J. Paul Peter (1993),
Satisfaction and Retention Across Firms," presentation at the "Improving the Measurement of Service Quality." Journal of
TIMS College of Marketing Special Interest Conference on Retailing. 69 (Spring). 127-39.
Services Marketing. Nashville, TN. (September). Bu/zell, Robert D. und Bradley T Gale (1987), The PIMS Princi-
Bolton. Ruth N. and James H. Drew (1992)/'Mitigating the Effect ples. New York: The Free Press.
of Service Encounters," Marketing Letters. 3(1), 57-70, Cohen. Jacob and Patricia Cohen (\9^i). Applied \fultivariate Re-
[Link]^Correlation [Link] for the Behavioral Sciences. 2d
Boulding, William, Ajay Kalra, Richard Staelin, and Valarie A.
ed. Hillsdale, NJ: Lawrence Eribaum Associates.
Zeitham! (1993), "A Dynamic Process Model of Service Qual- Coyne. Kevin P (1989). "Beyond Service Fads: Meaningful Strate-
ity: From Expectations to Behavioral Intentions," Journal of gies for the Real World," Sloan Management Review (Sum-
Marketing Research, 30 (February). 7-27. mer). 69-76.

Service Quality / 45
Cronin. J. Joseph. Jr. and Steven A. Taylor (1992). "[Link] Ser- . Valarie A. ZeithamI. and Leonard L. Berry (1985). "A
vice Quality: A Reexamination and Extension." Journal of Conceptual Model of Service Quality and Its Implications for
Marketing, 56 (July), 55-68. Future Research," Journal of Marketing. 49 (Fall). 41-50.
Dawkins, P and F. Reichheld (1990), "Customer Retention as a . , and (1988). "SERVQUAL: A Multiple-
Competitive Weapon." Directors and Boards. 14 (Summer), Item Scale for Measuring Consumer Perceptions of Service
42^7. Quality," Journal of Retailing, 64 (1). 12-40.
Executive Report on Customer Satisfaction (1992). "Northwest —~. — ; and (1994a), "Moving Forward in Ser-
Finds Direct Correlation between Service and Customer Prefer- vice Quality [Link]: [Link] Different Levels of Cus-
ence," (November 15). New York: Alexander Research and tomer Expectations. Comparing Altemative Scales, and Exam-
Communications. I. ining the Perfonnance Behavioral Intentions Link," Marketing
Ford Motor Company (199()), Memorandum to dealers. (October 3). Science Institute Working Paper, Report No. 94-114, (Septem-
Fornell, Claes and Birger Wemerfelt (1987). "Defensive Marketing ber).
Strategy by Customer Complaint Management: A Theoretical , , and (1994b), "Reassessment of Expecta-
Analysis," Journal of Marketing Research, 24 (November), tions as a Comparison Standard in Measuring Service Quality:
337-^6. Implications for Funher Research." Journal of Marketing, 58
and (1988), "A Model for Customer Complaint (January), 111-24.
Management," Marketing Science. 1 (Summer). 271-86. Phillips, Lynn, D., Dae R. Chang, and Robert Buzzell (1983),
Gale. Bradley (1992). "Monitoring Customer Satisfaction and "Product Quality. Cost Position and Business Performance: A
Market Perceived Quality." American Markeiing Association Test of Some Key Hypotheses." Journal of Marketing, 47
Worth Repeating Series, Number 922CSO I. Chicago: Ameri- (Spring). 26-43.
can Marketing AssiKiation. Reichheld, Frederick and W. Earl [Link], Jr. (1990), "Zero Defec-
Genmano, Charles A. (1992). The Bottom Line of Quality. Mem- tions: Quality Comes to Services," Harvard Business Review,
phis. TN: Federal Express Corporation. 68 (September/October). 105-11.
Greising, David (1994), "Quality: How to Make ll Pay." Business Richins. Marsha (1983), "Negative Word-of-Mouth by Dissatisfied
Week (August 8), 54 59. Consumers: A Pilot Study," Joumal of Marketing, 47 (Winter),
Hirschman, A. O. (1970), Exit, Voice and Loyalty Responses to De- 68-78.
clines in Firms. Organizations and States. Cambridge. MA: Rose. S. (1990). "The Coming Revolution in Credit Cards." Jour-
Harvard University Press. nal of Retail Banking, 12 (Summer). 17-19,
Keams, David T. and David A. Nadler (1992). Prophets in the Rust. Roland T. and Anthony J, Zahorik (1993). "[Link] Satis-
Dark: How Xerox Reinvented Itself and Beat Back Japanese. faction, Customer Retention, and Market Share," Joumal of Re-
New York: HarperCollins I*ublishers. tailing, 69 (2), 193-215.
Koska. Mary T. (1990) "High-Quality Care and Hospital Profits: Is , , and Timothy L. Keiningham (1995), "Retum on
There a link?" Hospitals. 64 (March 5). 62-63. Quality (ROQ): Making Service Quality Financially Account-
LaBarbera, Priscilla and David Mazursky (1983). "A Longitudinal able," Journal of Marketing. 59 (April), 58-70.
Assessmetit of Consumer Satisfaction/Dissatisfaction: The Dy- Scaglione, F. (1988), "Two-Way Communication: Tapping into
namic Aspect of the Cognitive Process," Journal of Marketing Gripes and Profits," Management Review, 77 (September),
Reseanh, 20, (November). 393-404. 51-53.
Lovelock, Christopher (1994), Product Plus: How Product + Ser- Singh. Jagdip (1988), "Consumer Complaint Intentions and Be-
vice = Competitive Advantage. New York: McGraw-Hill. havior: Definitional and Taxonomieal Issues." Journal of Mar-
Maute. Manfred and William R, Forrester. Jr. (t993), "The Struc- keting, 52 (January), 93-107.
ture and Determinants of Consumer Complaint Intentions and Solnick. Sara J. and David Hemenway (1992). "Complaints and
Behavior." Journal of Economic Psychology. 14 (3). 219-47. Dl,senrollment at a Health Maintenance Organization," Joumal
McLaughtin. John P. (1993). "Ensuring Customer Satisfaction is a of Con.'iumer Affairs. 26 (1). 90-103.
Strategic Issue, Not Just an OfKrational One." presentation at Teas. R. Kenneth (1993). "Expectations, Perfomiance Evaluation,
the AlC Customer Satisfaction Measurement Conference. and Consumer's Perceptions of Quality." Journal of Marketing,
Chicago. (December 6-7). 57 (October), 18-34.
Myers. Jerome L. (t979). Fundamentals of Experimental Design, Woodside, A., L. Frey, and R. Daly (1989). "Linking Service Qual-
3d ed. Boslon: Allyn and Bacon. ity, Customer Satisfaction and Behavioral Intention." Journal
Newman. Joseph W. and Richard A, Werbel (I 973), "Multivariate of Health Care Marketing. 9 (December). 5-17.
Analysis of Brand Loyalty for Major Household Appliances," Zahorik, Anthony J. and Roland T. Rust (1992), "Modeling the Im-
Journal of Marketing Research. 42 (November). 404—49, pact of Service Quality on Profitability: A Review." in Ad-
Nunnally. Jum C. (1978). Psychometric Theory. 2d ed. New York: vances in Services Marketing and Management, Vol. 1, Teresa
McGraw-Hiii Bonk Company. A. Swartz, David E. Bowen, and Stephen W. Brown, eds.
Parasuraman. A.. Leonard L. Berry, and Valarie A. ZeithamI Greenwich, CT: JAI Press, 247-76.
(1991a), "Retmemetit and Reassessment of the SERVQUAL Zeithami. Valarie A., A. Parasuraman. and Leonard L. Berry
Scale." Journal of Retailing, 67 (4), 420-50, (1990). Delivering Quality Service: Balancing Customer Per-
, , and (1991b), "Understanding Customer ceptions and Expectations. New York: The Free Press.
Expectations of Service," Sloan Management Review, 32 , Leonard L. Berry, and A. Parasuraman (1993), "The Na-
(Spring), 3 9 ^ 8 . ture and Determinants of Customer Expectations of Service,"
. , and (1993), "More on Improving Service Journal of the Academy of Marketing Science, 21 ( I ) , 1-12
Quality Measurement," Journal of Retailing. 69 (Spring),
140^7.

46 / Journal of Marketing, April 1996

View publication stats