Industrial Marketing Management 33 (2004) 675 – 687

Customer value, overall satisfaction, end-user loyalty, and market

performance in detail intensive industries
Joseph M. Spiteria, Paul A. Dionb,*
a
6255 RFD, Long Grove, IL 60047, USA
b
Sigmund Weis School of Business, Susquehanna University, 514 University Avenue, Selinsgrove, PA 17870, USA

Received 21 October 2003; received in revised form 15 February 2004; accepted 28 March 2004
Available online 14 July 2004

Abstract

The objective was to test a customer value variable, as operationalized by a modification of Ulaga and Eggert’s scale, as a direct
explanatory concept in predicting satisfaction, loyalty, and market performance in a hybrid veterinary pharmaceuticals market. The market
was considered to be a hybrid because physicians purchase on behalf of their clients. The scale proved to be reliable and valid in a detail
intensive market using 220 veterinary physicians as respondents. To achieve an acceptable structural equation modeling fit, the customer
value variable had to be dropped. The final model indicated a direct influence by product, strategic, and personal benefits as well as perceived
sacrifices on the dependent variables. Management implications of the study are described.
D 2004 Published by Elsevier Inc.

Keywords: Customer value; Satisfaction; Ulaga – Eggert scale; Veterinary pharmaceuticals

1. Introduction and trust.’’ Methodologically, there is also a need to inves-

tigate how robust this new scale is outside its original setting
Marketing academics have identified customer value as of French industrial purchasing managers. The purpose of
one of the top research agendas. The Marketing Science this study was to empirically investigate the model shown
Institute has consistently included customer value in its list below in Fig. 1. This model was synthesized from several
of research priorities. The Institute for the Study of Business available research efforts.
Markets at the Pennsylvania State University and the Center Additionally, as the new consumer value scale of Ulaga
for Business and Industrial Marketing at Georgia State and Eggert (2002) has neither been revalidated nor tested for
University have also integrated research on customer value its predictive effectiveness in a new marketing setting, this
and view it as a key research program. Customer value is on newly developed scale was validated and then empirically
the mission statements of many large corporations (Exxon) tested in a detail intensive industry (DII) setting. In sum-
and has become a sought after source of competitive advan- mary, the research objectives were to:
tage. Now that the first scale for the operationalization of this
concept has been published (Ulaga & Eggert, 2002), research 1. empirically investigate the effect of perceived product
into customer value can become a more empirical area of benefits, perceived strategic benefits, perceived personal
inquiry. The needed theoretical areas of research for custom- benefits, perceived sacrifices, and perceived relationship
er value include ‘‘. . .major antecedents, mediating variables, benefits on the construct of customer value;
and consequences’’ (Ulaga & Eggert, 2002, p.15) and ‘‘to 2. empirically link the measurement of customer value to
investigate relationship customer-value concepts closely outcome measures, such as overall satisfaction, end-user
related to the construct such as commitment, satisfaction, loyalty, and market performance;
3. empirically investigate the role of overall satisfaction in
the model;
* Corresponding author. Tel.: +1-570-372-4034. 4. empirically investigate the effects of relationship benefits
E-mail address: [email protected] (P.A. Dion). and sacrifices on market performance;

0019-8501/$ – see front matter D 2004 Published by Elsevier Inc.

doi:10.1016/j.indmarman.2004.03.005
676 J.M. Spiteri, P.A. Dion / Industrial Marketing Management 33 (2004) 675–687

Fig. 1. Customer values’ main model.

5. finally, empirically test and validate Ulaga and Eggert’s uations that involve intense relationship-building activities
(2002) customer value scale in a DII market setting. by manufacturers’ representatives with the medical profes-
sional. This and other interesting aspects of DII have long
The research context employed here is veterinary pharma- attracted marketing academics (Krishnamurthi & Zoltners,
ceuticals. The industry that comprises ethical pharmaceut- 1994).
icals for both humans and animals is often called ‘‘detail The theoretical underpinnings for this research into cus-
intensive industries’’ because of its reliance on detailers tomer value in DII begin with the work of Carter (1997), who
(another name for ethical pharmaceuticals salespersons). DII researched the importance of relationship marketing activi-
offers one of the most salient examples of all the elements of ties in the human pharmaceutical market. Carter asserted that
business-to-business exchanges with product marketing sit- given the importance of detailing (presenting information on
 J.M. Spiteri, P.A. Dion / Industrial Marketing Management 33 (2004) 675–687 677

drugs) ‘‘if there is no personal contact, then the chances that followed by the processes that lead to the construct of
the customer will purchase the product is small.’’ It is customer value. These processes include the assessment of
through the personal selling activity that a long-term rela- product quality and price and how they lead to the concept
tionship between the manufacturer and their representative is of customer value followed by a look at the construction of
maintained with the customer—the medical professional. a recently released scale. After the antecedents and process
Therefore, the spectrum of personal selling activities is are presented, the literature is reviewed for possible out-
viewed as central to the relationship marketing efforts of comes to customer value. They are overall buyer satisfac-
the firm and one of the critical sets of antecedents to tion, end-user loyalty, and market performance. Finally,
customer value. Recently, drug manufacturers have invested there is an examination of marketing issues in DII.
in direct-to-consumer advertising (DTC), which is changing
the dynamics and the relative importance of detailing and 2.1. Antecedents of customer value
other relationship-building activities to DTC. DII have also
become examples of business to consumer marketing. In- There are many legitimate antecedents to the compre-
deed, DII is now a good example of a hybrid of business-to- hensive and complex concept of customer value. There are
business and business-to-consumer. 16 such influences documented in the literature. A detailed
The research was conducted in three phases. Using a discussion of each would be too lengthy to include in this
Delphi methodology on 22 respondents, Phase I tested the article but Table 1 gives a summary of the influences and
language or wording adaptation of the perceived strategic references findings.
benefits indicators variable in the veterinary market. Phase II
used a sample of 100 multi-informants to test the importance 2.2. Processes of customer value
of the items of benefits and sacrifices found in the scale of
customer value. Based on the findings of the two earlier Quality models comprise a large research stream in
phases, Phase III consisted of fielding the final version of the marketing thought. However, recently, quality models have
full questionnaire to a random sample of 623 respondents. been criticized for not explicitly including the effect of
perceived price or cost on the customers’ judgment of quality
(Iacobucci, Grayson, & Ostrom, 1994) and for not consi-
2. Literature and theory background dering sustainable competitive advantage (Butz & Goodstein,
1996). Since the continued search for competitive advantage
The literature review begins with a conceptual presenta- has been directed toward delivering superior customer value
tion of the antecedents to the customer value construct, (Band, 1991; Day, 1990; Gale, 1994; Naumann, 1995),

Table 1
Antecedents of customer value
Antecedents Findings Author(s)
Market orientation Directs firm to focus on gathering and Slater & Narver, 2000; Woodruff & Gardial, 1996
dissemination of customer value information
Learning Positively linked to implementation of Day, 1994a, 1994b; Kohli & Jaworski, 1990;
customer value information Slater & Narver, 1995
Competence/expertise Linked to customer value Doney & Cannon, 1997; Smith & Barclay, 1997
Communication Linked to customer value Anderson & Weitz, 1989; Gronroos, 1990
Alignment of sales compensation to value Linked to customer value Anderson et al., 1994
Equity/fairness in exchange Linked to success of buyer/seller Gundlach & Murphy, 1993; Tax, Brown,
relationships & Chandrashekaran, 1998
Forbearing opportunistic behavior Linked to success of buyer/seller Morgan & Hunt, 1994
relationships
Ethical acts Linked to success of buyer/seller Boedecker, Morgan, & Stoltman, 1991;
relationships Lagace et al., 1991
Shared values Linked to success of buyer/seller Morgan & Hunt, 1994
relationships
Promotional investments Add customer value or perceived value Woodruff & Gardial, 1996
Relational investments Add customer value or perceived value Berry, 1995; De Wulf et al., 2001; Goff et al., 1997
Innovations Add customer value or perceived value Dickson, 1992; Ghemawat, 1986; Jacobson, 1992
Total quality management Add customer value or perceived value Buzzell & Wiersema, 1981; Capon, Farely, & Hoenig, 1990
Seeking sustainable competitive Lead to acceptance of customer value Christopher et al., 1991; Porter, 1985;
advantage as competitive advantage Treacy & Wiersema, 1995
Process efficiencies Lead to acceptance of customer value McKenna, 1991; Porter, 1985
as competitive advantage
Cost-cutting initiatives Lead to acceptance of customer value Cannon & Homburg, 2001
as competitive advantage
678 J.M. Spiteri, P.A. Dion / Industrial Marketing Management 33 (2004) 675–687

frameworks and tools have now been developed for manag- Table 2
Definitions for customer value
ing toward customer-focused competitive advantage (Clem-
ons & Woodruff, 1992; Day, 1990; Day & Wensley, 1988; Benefits Costs or sacrifices Researcher(s)
Slater & Narver, 1995). In traditional quality models, product Maximize rewards Minimize costs Bagozzi, 1974
quality precedes customer satisfaction (Parasuraman, Zei- Utility of what is received What is given Zeithaml, 1988,
p. 14
thaml, & Berry, 1988). Iacobucci et al. (1994) claim that the
Quality and benefits Relative to sacrifices Monroe, 1991,
traditional quality model needs a simple modification to p. 46
include financial factors so that the customer’s evaluation Five categories of value: Sheth, Newman,
can take into consideration ‘‘what they paid for what they functional, social, & Gross, 1991
got.’’ emotional, epistemic,
and conditional value
Product quality has also not been sufficient as a sustain-
Worth of set of economic, Exchanged for price Anderson, Jain,
able competitive advantage (Butz & Goodstein, 1996). technical, service, and of product & Chintagunta,
Organizations have earnestly pursued internal changes in social benefits 1993, p. 5
the process of quality control, such as downsizing, restruc- Product value, value-in-use, Burns, 1993
turing, and reengineering. However, although the way possession value, and
overall value
organizations work may change, these steps have not always
Perceived quality Relative prices Gale, 1994
had the desired impact on the bottom line (Hall, Rosenthal, Emotional bond between Butz & Goodstein,
& Wade, 1993). Consequently, the search for competitive customer and producer 1996
advantage goes on. One direction has been the call for
organizations to direct the pursuit of competitive advantage
toward delivering superior customer value (Band, 1991; customers, ‘‘considering what they want and believe that
Day, 1990; Gale, 1994; Naumann, 1995). they can get from buying and using a seller’s product’’
Ulaga and Chacour (2001) conceptualized customer (Woodruff, 1997). This research deals with this latter
value as being derived from the concept of product quality, concept.
as depicted in Fig. 2. There are many diverse meanings and perspectives for
Subsequent to this work, Ulaga and Eggert (2002) went the concept of customer value. These are summarized in
on to develop a scale for measuring customer value, which Table 2.
is discussed in the next section. The definitions have similarities and differences. Con-
sensus is seen in the linking of customer value to some
2.3. The concept of customer value product and/or service. Further, at the core of customer
value is the perspective of the customer and not of the seller.
The term value is used in many different contexts in Finally, the perception of customer value involves a tradeoff
marketing. One perspective in organizational strategy main- between what the customer receives (e.g., quality, benefits,
tains that creating and delivering superior customer value to worth, utilities) and what they give up (prices, sacrifices) to
high-value customers will increase the value of an organi- acquire the product. Divergence is seen in the reliance on
zation (Slywotzky, 1996). High-value customers have their other terms—utility, worth, benefits, and quality—to con-
monetary worth as individual customers to the organization struct the definitions, but these terms are not all well
quantified, whereas value of an organization quantifies the developed or defined. Hence, it becomes difficult to com-
worth to its owners (stakeholders). On the other hand, pare concepts: Is customer value as quality the same as
customer value takes the perspective of an organization’s customer value as utility?

Fig. 2. Components of customer perceived product value.

 J.M. Spiteri, P.A. Dion / Industrial Marketing Management 33 (2004) 675–687 679

2.4. Customer value scale

Recently, customer value has been defined and a psy-

chometrically sound scale for measuring the concept in
business-to-business markets has been developed (Ulaga
& Eggert, 2002). Four customer-value components emerged
from their study: product-related benefits, strategic benefits,
personal benefits, and relationship sacrifices. These were
aggregated into the higher-order construct of customer
value. As has been stressed earlier, the literature contains
a variety of definitions of customer-perceived value. In these
definitions, Eggert and Ulaga (2002) identified three com-
mon elements:

1. Customer-perceived value is a trade-off between bene-

fits and sacrifices perceived by the customer in a sup-
plier’s offering (Monroe, 1990, p. 46; Zeithaml, 1988,
p. 14).
2. Value perceptions are subjective (Kortege & Okonkwo,
1993).
3. Competition is important, because value is relative to
competition. Fig. 3. Customer value indicators and links to benefit.

Ulaga and Eggert (2002) define customer value in business

markets as the trade-off between the multiple benefits and
sacrifices of a supplier’s offering as perceived by key The concept of customer value has a strong relationship
decision-makers in the customer’s organization and taking to customer satisfaction. Both concepts describe evaluation
into consideration the available alternative suppliers’ offer- and judgments of products in use situations. In fact, per-
ings in a specific use situation. The perceived benefits are ceived value may lead directly to the formation of overall
some combination of physical attributes, service attributes, satisfaction feelings (Churchill & Surprenant, 1982). Alter-
and technical support available relative to the particular natively, they may be compared with one or more other
use of the product as well as the purchase price and other standards (Clemons & Woodruff, 1992; Woodruff, Schu-
indicators of perceived quality (Ravald & Gronroos, 1996). mann, Clemons, Burns, & Gardial, 1990).
Additionally, a complexity is that an individual may
evaluate the same product differently on different occa- 2.5.2. End-user loyalty
sions. The concepts of quality, brand and/or company equity,
Ulaga and Eggert’s (2002) scale covers the indicators of customer satisfaction, and customer value are interrelated.
the variables of perceived ‘‘relationship benefits’’ and ‘‘sac- Customer satisfaction is one of the most important criteria
rifices.’’ These indicators (in parentheses) and their con- for customer loyalty (Heskett, Sasser, & Schlesinger,
structs are shown in Fig. 3. 1997). Howard and Sheth (1969). Several studies discuss
and/or observe a strong link between customer satisfaction
2.5. Consequence of customer value and loyalty (Anderson & Sullivan, 1993; Bearden & Teal,
1983; Boulding, Staelin, Kaira, & Zeithaml, 1993; For-
2.5.1. Overall satisfaction nell, 1992; Oliver & Swan, 1989). Reichheld and Sasser
The concept of customer value is related to, but different (1990) discuss why increasing customer loyalty should
from, both concepts of satisfaction (Woodruff & Gardial, lead to higher profitability. However, recent studies have
1996). The literature identifies two types of satisfactions: demonstrated mixed results in analyzing the relationship
transactional and overall satisfaction (or cumulative satis- between satisfaction and loyalty. These studies suggest
faction). Transactional satisfaction is defined as the post- that satisfied customers may not be sufficient to create
choice evaluative judgment of a specific purchase occasion loyal customers (e.g., Cronin & Taylor, 1992; Fornell,
(Hunt, 1977; Oliver, 1980, 1993), whereas cumulative 1992).
customer satisfaction is an overall evaluation based on the
total experience (Fornell, 1992; Johnson & Fornell, 1991). 2.5.3. Market performance
Because relationship marketing is long term, the more An empirical study conducted in Europe’s pharmaceutical
appropriate variable to measure is ‘‘overall satisfaction’’ market (Scharitzer & Kollarits, 2000) showed physicians’
(Ravald & Gronroos, 1996). subjective customer value assessments of certain pharma-
680 J.M. Spiteri, P.A. Dion / Industrial Marketing Management 33 (2004) 675–687

ceutical companies lead to the market success of a product. 3.1. Sample selection
The construct of ‘‘customer loyalty’’ is at the interface
between subjectively observed and objectively measured The 623 sample respondents were chosen from 427
dimensions of enterprise performance. They also found a clinics in a monthly panel of 1200 American Veterinary
link between perceived service quality and economic suc- Medical Association (AVMA) members. Potential clinics
cess. In that study, the pharmaceutical companies used were randomly chosen and invited to the panel, if they
relationship-marketing activities, such as personal selling, refused; the clinic was replaced and was available for
for building long-term relationships. Rust and Zahorik another draw. To insure representativeness responders were
(1993) empirically demonstrated the relationship between statistically compared with nonresponders and the AVMA
customer satisfaction and profitability for a health care universal demographics universe.
organization and also found that although firms see an initial
strong relationship between satisfaction scores and perfor- 3.2. Statement of research hypotheses
mance, this relationship declines over time. Ravald and
Gonroos (1996) found that ‘‘loyalty’’ was linked to mutually The hypotheses were directed at testing the study’s
profitable relationships. In fact some studies have indicated theoretical framework outlined in Fig. 1. They are summa-
that satisfied clients can also be just as disloyal (Lowenstein, rized in Table 3.
1997).
Several additional industry-related researches include 3.3. Measurement of research variables
Scharitzer and Kollarits (2000), who used a satisfaction
model to support the link between perceived service The five-point low (1) and high (5) response scales for
quality and economic success. In that study, the pharma- each construct were evaluated using Cronbach’s alpha and
ceutical companies used relationship-marketing activities, exploratory and confirmatory factor analysis. After minor
such as personal selling, for building long-term relation- modifications for double-loading and nonloading items,
ships. Rust and Zahorik (1993) empirically demonstrated measures were only included after they demonstrate accept-
the relationship between customer satisfaction and profit- able levels validity and reliability. Table 4 presents both the
ability for a health care organization, and also found that
although firms see an initial strong relationship between
satisfaction scores and performance, this relationship
declines over time. Table 3
H1 – H3: There are positive relationships between perceived customer
value and overall buyers’ satisfaction (H1), end-user brand and
company loyalty (H2), and market performance (H3).
3. Study method H4 – H6: There are positive relationships between perceived
product benefits and overall buyers’ satisfaction (H4), end-user brand
The study was conducted in three phases, with data or company loyalty (H5) and market performance (H6).
drawn from a randomly selected panel of 1200 veterinar- H7 – H9: There are positive relationships between perceived
personal benefits and overall buyers’ satisfaction (H7), end-user brand
ians who participated in a monthly buyer panel and
or company loyalty (H8), and market performance (H9).
responded by mail. Telephone reminders were used to H10 – H12: There are positive relationships between perceived strategic
increase the response rate. benefits overall buyers’ satisfaction (H10), end-user brand or company
loyalty (H11), and market performance (H12).
1. In Phase I, 22 respondents checked the wording H13 – H15: There are negative relationships between perceived relationship
sacrifices and overall buyers’ satisfaction (H13), end-user brand or
changes to the original instrument of Ulaga and Eggert
company loyalty (H15), and market performance (H16).
to adapt it to the present research context. This panel H16 – H17: There are positive relationships between overall buyers’
consisted of five academicians, seven industry con- satisfaction and end-user or company loyalty (H16) and a positive
sultants, four sales professionals, and six market relationship between end-user loyalty and market performance (H17).
professionals. H18: There is a positive relationship between overall buyer satisfaction and
market performance (H18).
2. Phase II of the study tested the new wording and new
H19: There is a linear relationship between customer value and overall
items on a pilot sample of 100 that excluded the 623 used buyer’s satisfaction and end-user loyalty (H19).
for the full questionnaire. The purpose was to investigate H20: There is a linear relationship among perceived product, personal,
if the originally worded indicators in the customer value strategic benefits and perceived sacrifices, overall buyer’s satisfaction,
scale were important to the purchasing decision-makers and end-user loyalty (H20).
H21: There is a linear relationship customer value and overall buyer’s
in DII.
satisfaction, end-user loyalty, and market performance (H21).
3. Phase III implemented a complete rollout to 623 H22: There is a linear relationship among perceived product, personal,
qualified panelists (those who have been panelists for strategic benefits and perceived sacrifices, overall buyer’s satisfaction,
at least 3 years) whose purchase histories and profiles are end-user loyalty, and market performance (H22).
known and have established relationships with the H23a: That is, the main model in Fig. 1 is a theoretically better fit than any
other alternate models and is, therefore, more parsimonious (H23).
various suppliers.
 J.M. Spiteri, P.A. Dion / Industrial Marketing Management 33 (2004) 675–687 681

Table 4
Cronbach’s alpha values of present and previous studies
Perceived independent, Items Sample item Referenced/computed References
mediator, and dependent Cronbach’s alpha
variables
Product-related benefits 4 Please rate product quality. .82/.79 Ulaga & Eggert, 2001
Strategic-related benefits 4 Please rate product brand on .82/.93 Ulaga & Eggert, 2001
sharing their knowledge or
expertise with you.
Personal benefits 4 Please rate product brand on .84/.94 Ulaga & Eggert, 2001
how pleasant the working
relationship is with you.
Relationship sacrifices 3 Please rate the product brand .90/.65 Ulaga & Eggert, 2001
on how much time they cost you.
Buyer’s customer loyalty 3 Product, brand, or company loyalty .57/.78 Butaney & Wortzel, 1988
Overall satisfaction 3 Please rate how satisfied you or .87 – .94/.82 Crosby & Stephens, 1987
your clinic are with using the
product brand.

referenced and computed Cronbach’s alpha values for the 4. Study findings
scales.
The relationship benefits variable was computed using 4.1. Sample characteristics
factor weights from the product, strategic, and personal
benefits. The perceived customer value variable was com- A total of 220 usable responses were received for a
puted as a ratio of relationship benefits to relationship 35.5% response rate. The vets/clinic variable shows that the
sacrifices. Market performance was computed as the 3-year sample is almost identical to the universe (2.12 vs. 2.13),
average of the monthly doses dispensed. but there is a difference when compared with the non-

Table 5
Summary of tests of hypotheses
Hypotheses Independent variable(s) Dependent variable Pearson’s correlation Significance level ( P)
H1 Customer value Overall satisfaction .22 .004
H2 Customer value End-user loyalty .23 .000
H3 Customer value Market performance .04 .690
H4 Product benefits Overall satisfaction .62 .000
H5 Product benefits End-user loyalty .50 .000
H6 Product benefits Market performance .19 .016
H7 Personal benefits Overall satisfaction .42 .000
H8 Personal benefits End-user loyalty .37 .000
H9 Personal benefits Market performance .056 .469
H10 Strategic benefits Overall satisfaction .50 .000
H11 Strategic benefits End-user loyalty .45 .000
H12 Strategic benefits Market performance .01 .854
H13 Relationship sacrifices Overall satisfaction .09 .242
H14 Relationship sacrifices End-user loyalty .04 .591
H15 Relationship sacrifices Market performance .03 .723
H16 Overall satisfaction End-user loyalty .52 .000
H17 End-user loyalty Market performance .22 .005
H18 Overall satisfaction Market performance .04 .593
H19 Customer value and overall satisfaction End-user loyalty R2 = 29% F = 33.5, P = .000
H20 Perceived product, personal, strategic, End-user loyalty R2 = 36.1% F = 18.2, P = .000
benefits, perceived sacrifices, and overall
satisfaction
H21 Customer value, overall satisfaction, and Market performance R2 = 8.5% F = 5.05, P = .002
end-user loyalty
H22 Perceived product, personal, strategic, Market performance R2 = 13.5% F = 4.16, P = .001
benefits, perceived sacrifices, overall
satisfaction, and end-user loyalty
H23a Main model Model did not load on SEM.
H23b Alternate model GFI=.992 (a very good fit), v2 = 4.49 (6 df), P=.661
682 J.M. Spiteri, P.A. Dion / Industrial Marketing Management 33 (2004) 675–687

responders. Larger clinics seemed to respond more, most and indicators are ranked lowest. The mean ratings ranged
probably because they are better staffed and more likely from 4.97 for the highest individual items to the lowest
have office staff to help with the clinics purchase function; 2.98 out of a scale of 1 –5 for the lowest to 5 at the
they are thus better targets of the multi-informant method- highest, implying that even the lowest was only slightly
ology. All the other variables—practice types, geographic lower than the average in the scale. Confirmatory factor
spread, and dispensed/vet—all show good representation of analysis validated the Phase II scale and just over 68% of
the sample with the universe and nonresponders. the variance in the indicators was explained by four
factors, which were theoretically the three benefits and
4.2. Phase II findings on importance of scale indicators one sacrifices proposed in Ulaga and Eggert’s (2002)
scale.
The 100 multi-informant Phase II research tested the
importance of all the scales indicators to confirm that 4.3. Measurement properties
these benefits and sacrifices applied to DII setting. The
results showed that they were all important and meaning- The computed Cronbach’s scores are reported in Table 2
ful, of course, to varying degrees, and that an open-ended and all exceed .6 (Hair et al., 1995, p. 88). A confirmatory
request for more or different benefits and sacrifices factor analysis of the indicator variables loaded onto factors
provided no additional items. Weighting analysis showed representing relationship benefits, sacrifices, satisfaction,
that the highest weights were given to product benefits and loyalty.
(34%), followed by relationship sacrifices or costs (27%),
strategic benefits (21%), and finally personal benefits 4.4. Tests of research hypotheses
(18%). Ranking of the individual items ranged from
product reliability (item 1), the new/specific item product In Table 5, the results of tests of Hypotheses 1 –22 using
safety (item 2), product quality (item 3), and all the way correlation and regression are presented. Hypothesis 23 was
to personal value (item 14) and personal recognition (item tested using a structural equation modeling approach and is
15). It is noteworthy that the relationship building benefits presented below.

Fig. 4. Customer values’ alternate model.

 J.M. Spiteri, P.A. Dion / Industrial Marketing Management 33 (2004) 675–687 683

Hypothesis 23: The main model presented in the intro- impact on the overall satisfaction and end-user loyalty
duction was tested using a SEM (AMOS 4.01) approach. outcome measures than on market performance, most
While the main model did not load at all, the results for the likely because of the time dependent promotional lagged
alternate model (see Fig. 4) were acceptable. However, it effects. Moreover, the scale’s component influence on
was necessary to drop the customer value variable to market performance also appears to be direct and not
achieve an acceptable GFI. Because the customer value through the roles of mediating variables like the business
was a ratio of relationship benefits to sacrifices the relation- buyer’s overall satisfaction and end-user loyalty. There-
ship benefits variable was dropped because it was derived fore, the alternate model had higher goodness-of-fit than
from product benefits, strategic benefits, and personal ben- the main model. Nevertheless, the research did confirm
efits using factor weights and would introduce a latent the general structure proposed—antecedents, dimensions
variable into the model. The alternate model gave the of or higher order customer value construct, linked to the
following: GFI=.992, adjusted GFI=.964, v2 = 4.49 (6 df ), three outcome measures of overall satisfaction to end-user
and P =.661. These results imply that this is an acceptably loyalty to market performance.
fitting model. Thus, the alternate model is a more acceptable The research also confirmed both the validity and
and a better goodness-of-fitting model than the main model. reliability of all the measures used, including the new
While better fitting models were obtainable, each required scale of Ulaga and Eggert (2002). The language of the
the omission of model variables for only marginal increases items in the instrument was also sufficiently generic that
in better fit. However, these more parsimonious models they were used in a new DII setting with only minimal
were not as well supported by the theory that argued for the use of questionnaire instructions and directions. While
inclusion of the variables, such as market performance and the literature-supported conceptual antecedents to custom-
relationship costs (which provided the best fitting mode er value were not all empirically researched, key ante-
with the omission of these variables: GFI=.999). The results cedents arising from sales-force activities and efforts
for the specific relationships in the model are presented were found to be positively linked to both personal
below. They provide support for the alternate model in that and strategic benefits. Additionally, the results from the
all the relationships were found to be statistically significant. 100 multi-informant Phase II part of this research em-
pirically confirmed the importance of all of the scale’s
items.
Relationship Estimate S.E. Critical Significant
value
The research also confirmed the importance of end-
user loyalty, particularly in this hybrid marketing situa-
Personal p Product 0.236 0.073 3.232 yes
Strategic p Personal 0.770 0.045 17.127 extremely
tion—manufacturer selling to business buyer who, in turn,
Strategic p Product 0.183 0.044 4.156 yes sells to the consumer. This points to the influence of a
Satisfaction p Product 0.803 0.095 8.434 very brand message in business-to-business (e.g., the promo-
Loyal p Satisfaction 0.242 0.085 2.833 just tion of ‘‘Intel Inside’’). In these cases, the business buyer
Loyal p Strategic 0.366 0.114 3.211 yes must consider the influence of the ultimate consumer in
Loyal p Product 0.403 0.124 3.242 yes
Market performance 772.809 241.10 3.205 yes
making their product selection decisions between viable
p Loyal alternatives.
Market performance 904.159 273.94 3.301 yes The link between end-user loyalty and market perfor-
p Satisfaction mance was positive but weak. Most probably, this is due
Market performance 1169.40 404.34 2.892 just to the many promotional lagged effects that are present
p Product
when looking at present time cross-sectional data obtained
from current ratings when compared with the longitudi-
nally impacted product usage or market performance.
5. Conclusions and managerial implications While this was one of the limitations of the study, the
current results did set up a benchmark for future research
This study empirically tested and supported a large that may consider these lagged effects.
part of a postulated framework of antecedents, processes, It is highly desirable for a marketing scale to be
and outcomes of customer value. However, the SEM did sufficiently generic so that it is theoretically grounded
not support the use of a higher order construct of and generalizable. However, from a practitioner’s perspec-
customer value, as proposed in the earlier theory first tive, it is strategically desirable that the instrument used is
by Monroe (1991) and then by Ulaga and Eggert sufficiently specific so as to provide managerially action-
(2002). able insights. These two goals are often in conflict. In this
Most of the influence on the product selection process study, the results supported the general acceptance of the
is through the direct effects of the customer value generic language used in Ulaga and Eggert’s (2002)
components rather than through the indirect effect of customer value scale. Nevertheless, the practitioners re-
the higher order construct. Acting directly, these individ- quired specific language particularly for the strategic
ual customer value benefits and sacrifices also have more benefits components. For example, one strategic benefit
684 J.M. Spiteri, P.A. Dion / Industrial Marketing Management 33 (2004) 675–687

indicator was ‘‘Please rate Supplier 1 and Supplier 2 on customer value. It also found that overall satisfaction is a
helping you or your clinic to develop and maintain your better predictor of end-user loyalty than customer value.
strategic advantages.’’ Many of the respondents proffered However, in this research, there is an important difference
suggestions on language that made it clear what ‘‘strategic in the foci of these variables that was not investigated in
advantages’’ meant for medical practitioners. This re- previous research. The customer of the buyer (the pet
search balanced these two needs (theoretically generic owner, in this case) influences the perceptions of the buyer
vs. practitioner specific) by providing ‘‘explanations’’ or on the end-user loyalty ratings. However, the buyer makes
‘‘instructions’’ of what this general wording specifically the customer value and overall satisfaction ratings deci-
means to the responder. For example, to a clinic’s owner, sions. The theoretical implication of this is that industrial
strategic advantages ‘‘include attracting, maintaining, and purchase decisions must consider the responses of the
educating pet owners about animal health products and customers of the product. There are many buying busi-
veterinary professional.’’ ness-buying decisions that have brand loyalty implications.
Since the original responders of the customer value For example, the ‘‘Intel Inside’’ branding message cited
scale (Ulaga & Eggert, 2002) were French industrial before. Clearly, generic components in computers or auto-
buyers, it could not be assumed, without further testing, parts would not have applicable branding messages, but
that the items or indicators were important and hence many products in DII have been increasingly promoted as
meaningful to the DII setting. Accordingly, the 100 multi- DTC brand messages. The medical professional manager
informant Phase II research tested the importance of all now has to contend with a marketing manager directly
the scales indicators to confirm that these benefits and influencing the consumer. This implies that the consumer
sacrifices applied to DII setting. has become more powerful in brand selection. Interestingly,
The studies proposed antecedents have implications for this sets up potential conflict as to what the buyer may
managing toward customer-focused competitive advantages consider to be ‘‘value’’ (discounted price or deal) but on a
by assisting organizations to match internal quality man- product that the consumer does not prefer. Therefore, the
agement capabilities with an external strategic focus. This is measurements of customer value in business-to-business
accomplished by gathering customer value-oriented infor- must now consider what the consumer also ‘‘values.’’
mation necessary to make decisions that improve the cus- Ulaga and Eggert’s (2002) scale does not provide for the
tomer value delivery process. This leads to organizational influence of the consumer. However, in this research the
learning to align the internal organizations with what the combined scale includes the ‘‘pull power’’ of the end-user,
customers value. Management can, therefore, improve the leading to interesting insights. This has changed the dy-
customer value rating by increasing the effectiveness of the namics of marketing in DII industries, where the old theory
appropriate research confirmed antecedents. While a gener- that stressed the importance of the sales-force efforts, is
alizable instrument serves the purpose of establishing the now supplemented with the increased importance of end-
importance of customer value research in a particular user request for products spurred on by DTC adverting
setting, a specific instrument is then needed for strategic messages.
alignment and implementation. End-user loyalty was found to have a more direct link to
Support for the structure of the alternate research model market performance, and would probably be a stronger link
implies that increasing the overall customer value rating by if lagged effects are taken into account. Overall satisfaction
increasing specific components will improve all three-out- explains more of the variance of end-user loyalty than
come measures. Managerially, the insight is that suppliers customer value. So the question this raises is: ‘‘Do we
would do better by concentrating on measuring the effec- really need the customer value construct?’’ Iacobucci et al.
tiveness of improving the individual components rather than (1994) were the first to observe that all customer satisfaction
indexing them into the higher order construct, such as research needs is the inclusion of price. Certainly, the
customer value. A possible explanation for the ineffective- implication of this study is that traditional customer satis-
ness of the higher order construct is this: When reaching a faction could accomplish the same end as the components of
decision on optimal value, buyers are assumed to mentally customer value. However, the focus on what the customer
calculate a ratio of customer benefits to customer sacrifices values would need to guide what product attributes or uses
(this is termed the index method). As this is rather a are included in the scale.
sophisticated calculation, it may not be realistic in every In the DII hybrid marketing situation, this research on
industrial buying setting. But since the research on the actual the importance of the individual components demonstrated
mental model or decision-making mechanism is not yet that the product benefits are more important than personal
available, it is speculation at this point. or strategic benefits—the relationship-marketing dimen-
The roles of overall satisfaction and end-user loyalty sion, which loaded onto one variable. When they are
follow from the established theory and are necessary for combined, they aggregate to 39%, which is higher than
product selection or preference decisions. This study rein- the product benefits. The implication is that the compet-
forced the notion that end-user loyalty is a better predictor itive advantage has shifted to the strategic benefits, which
of market performance than either overall satisfaction or is where the impact for DTC is seen. This supports the
 J.M. Spiteri, P.A. Dion / Industrial Marketing Management 33 (2004) 675–687 685

notion that ‘‘despite the growing focus on customer References

service in business-to-business marketing, when it comes
to perception of customer perceived value, product quality Anderson, E., & Sullivan, M. (1993, Spring). The antecedents and con-
has a greater impact on customer’s perceived value than sequences of customer satisfaction for firms. Marketing Science, 12,
125 – 143.
service value.’’ Anderson, E., & Weitz, B. A. (1989, Fall). Determinants of continuity in
conventional industrial dyad. Marketing Science, 8, 310 – 323.
6. Limitations and recommendation for future research Anderson, E., Jain, C., & Chintagunta, P. K. (1993). Customer value as-
sessment in business markets. Journal of Business-to-Business Market-
ing, 1(1), 3 – 29.
The limitations of the study were the following: Anderson, E., Fornell, C., & Lehmann, D. R. (1994, July). Customer sat-
isfaction, market share, and profitability: Findings from Sweden. Jour-
1. Many of the antecedents have not been empirically nal of Marketing, 58, 53 – 66.
tested or confirmed and there is no consensus yet on Band, W. A. (1991). Creating value for customers. New York: Wiley.
Bagozzi, R. P. (1974, October). Marketing as an organized behavioral
their specification or conceptualization. This study has system of exchange. Journal of Marketing, 38(4), 77 – 81.
offered one possible set and researched several of them Bearden, W. O., & Teal, J. E. (1983, February). Selected determinants of
to provide a plausible basis for the indicators. Future consumer satisfaction and complaint reports. Journal of Marketing Re-
research should empirically investigate more. search, 20, 21 – 28.
2. Mediator variables, such as trust, commitment, and Berry, L. L. (1995). Relationship marketing. In L. L. Berry, L. Shostack,
& G. D. Upah (Eds.), Emerging Perspective on Services Marketing
satisfaction, have not been fully modeled or empiri- ( pp. 25 – 28). Chicago: American Marketing Association.
cally tested and so are also fruitful areas for future Boedecker, K. A., Morgan, F. W., & Stoltman, J. L. (1991, January). Legal
research. dimensions of salespersons’ statements: A review and managerial sug-
3. Within the model structure, the new customer value gestions. Journal of Marketing, 55, 70 – 80.
Boulding, W., Staelin, R., Kaira, A., & Zeithaml, V. (1993, February). A
scale has only been shown to apply to DII. Hence,
dynamic process model of service quality: From expectations to behav-
hypotheses testing and inferences are limited to DII ioral intentions. Journal of Marketing Research, 30, 7 – 27.
and may not be generalizable to other marketing Burns, M. J. (1993). Value in exchange: The customer perspective. Knox-
settings. Future research could also be directed at ville: The University of Tennessee.
proving the model and scale also apply in other market Butaney, G., & Wortzel, L. H. (1988, January). Distributor power versus
settings. manufacturer power: The customer role. Journal of Marketing, 52,
52 – 63.
4. No attempt has been made to fully include the impact of Butz Jr., H. E., & Goodstein, L. D. (1996, Winter). Measuring customer
time as a variable, although a 3-year average of market value: Gaining the strategic advantage. Organizational Dynamics, 24,
performance was used as a measure. The findings 63 – 77.
indicate that the low variance explained for the market Buzzell, R. D., & Wiersema, F. (1981, January – March). Modeling changes
performance outcome maybe due to the time-lagged in market research: A cross-sectional analysis. Strategic Management
Journal, 2, 27 – 42.
effects of promotions. Again, this is a fruitful area for Cannon, J. P., & Homburg, C. (2001, January). Buyer-supplier relationships
future research, where a longitudinal study can be and customer firm costs. Journal of Marketing, 65, 29 – 43.
performed with the results of this study used as a Capon, N., Farley, J. U., & Hoenig, S. (1990, October). Determinants of
benchmark. financial performance: A meta-analysis. Management Science, 36,
1143 – 1159.
5. Limitations with the definitions of customer value also
Carter, F. (1997). Effects of relationship marketing: The impact of
concern the lack of treatment of the time dimension. As personal selling on repeat-buying. Doctoral study, Carnegie Mellon
changes in the perception of value come from successive University.
purchase and use of the product, time will impact Christopher, M., Payne, A., & Ballantyne, D. (1991). Relationship market-
customers’ perception of value. Furthermore, time will ing: Bringing quality, customer service and marketing together. Oxford:
certainly influence the positive or negative development Butterworth-Heinemann.
Churchill, G. A., & Surprenant, C. (1982, November). An investigation into
of the relationship with the company and/or its the determinants of customer satisfaction. Journal of Marketing Re-
representatives. The time or circumstance dependent search, 19, 491 – 504.
information needed to deal with the above dynamic has Clemons, D. S., & Woodruff, R. B. (1992). Broadening the view of con-
not been considered. sumer (dis)satisfaction: A proposed means – end disconfirmation model
6. Finally, the last limitation is that by using panel cross- of CS/D. In C. T. Allen, T. J. Madden, T. A. Shimp, R. D. Howell, G.
M. Zinkhan, D. D. Heisley, R. J. Semenik, P. Dickson, V. Zeithaml, &
sectional data, this may result in bias due to common R. L. Jenk (Eds.), Marketing theory and applications ( pp. 209 – 216).
method variance and spurious cause –effect inferences. Chicago: American Marketing Association.
These are known to inflate correlation measures, Cronin Jr., J. J., & Taylor, S. A. (1992, July). Measuring service quality: A
resulting in overestimations of the influence of hypoth- reexamination and extension. Journal of Marketing, 56, 55 – 68.
esized predictors. Recognizing that drawing cause – Crosby, L. A., & Stephens, N. (1987, November). Effects of relationship
marketing on satisfaction, retention, and prices in the life insurance
effect inferences from cross-sectional data maybe industry. Journal of Marketing, 24, 404 – 411.
tenuous, a longitudinal study is recommended to confirm Day, G. S. (1990). Market driven strategy: Processes for creating value.
any established hypothesized sequence of effects. New York: Free Press.
686 J.M. Spiteri, P.A. Dion / Industrial Marketing Management 33 (2004) 675–687

Day, G. S. (1994a, October). The capabilities of market driven organiza- Alderson, & S. J. Shapiro (Eds.), Theory in marketing ( pp. 61 – 67).
tions. Journal of Marketing, 37 – 52. Homewood, IL: Richard D. Irwin.
Day, G. S. (1994b, Summer). Continuous learning about markets. Califor- McKenna, R. (1991). Relationship marketing—successful strategies for the
nia Management Review, 9 – 31. age of the customer. Reading, MA: Addison-Wesley.
Day, G. S., & Wensley, R. (1988, April). Assessing advantage: A frame- Monroe, K. B. (1990). Pricing: Making profitable decisions. New York:
work for diagnosing competitive superiority. Journal of Marketing, 52, McGraw-Hill.
1 – 20. Monroe, K. B. (1991). Pricing—making profitable decisions. New York:
De Wulf, K., Odekerken-Schroder, G., & Iacobucci, D. (2001, October). McGraw-Hill.
Investments in consumer relationships: A cross-country and cross- Morgan, R. M., & Hunt, S. D. (1994, July). The commitment-trust
industry exploration. Journal of Marketing, 65, 33 – 50. theory of relationship marketing. Journal of Marketing, 58, 20 – 38.
Dickson, P. R. (1992, January). Toward a general theory of competitive Naumann, E. (1995). Creating customer value. Cincinnati, OH: Thompson
rationality. Journal of Marketing, 56, 69 – 83. Executive Press.
Doney, P. M., & Cannon, J. P. (1997, April). An examination of the Oliver, R. L. (1980, November). A cognitive model of the antecedents and
nature of trust in buyer-seller relationships. Journal of Marketing, consequences of satisfaction decisions. Journal of Marketing Research,
61, 35 – 51. 17, 460 – 469.
Fornell, C. (1992, January). A national customer satisfaction barometer: Oliver, R. L. (1993). A conceptual model of service quality and service
The Swedish experience. Journal of Marketing, 55, 1 – 21. satisfaction. In T. A. Swartz, D. E. Bowen, & S. W. Brown (Eds.),
Gale, B. T. (1994). Managing customer value. New York: Free Press. Advances in services marketing and management ( pp. 65 – 86). Green-
Ghemawat, P. (1986, September – October). Sustainable advantage. Har- wich, CT: JAI.
vard Business Review, 64, 53 – 58. Oliver, R. L., & Swan, J. E. (1989, April). Consumer perceptions of inter-
Goff, B. G., Boles, J. S., Bellenger, D. N., & Stojack, C. (1997). The personal equity and satisfaction in transactions: A field survey ap-
influence of salesperson selling behaviors on customer satisfaction with proach. Journal of Marketing, 53, 21 – 35.
products. Journal of Retailing, 73(2), 171 – 183. Parasuraman, A., Zeithaml, V., & Berry, L. (1988). SERVQUAL: A mul-
Gronroos, C. (1990). Service management and marketing-managing the tiple item scale for measuring consumers’ perceptions of service quality.
moment of truth in service competition. Lexington, MA: Lexington Journal of Retailing, 64(1), 13 – 40.
Books. Porter, M. (1985). Competitive advantage-creating and sustaining superior
Gundlach, G. T., & Murphy, P. E. (1993, October). Ethical and legal foun- performance. New York, NY: Free Press.
dations of relational marketing exchange. Journal of Marketing, 57, Ravald, A., & Gonroos, C. (1996). The value concept and relationship
35 – 46. marketing. European Journal of Marketing, 30(4), 1 – 7.
Hair Jr., J., Anderson, R. E., Tatham, R. L., & Black, W. C. (1995). Mul- Reichheld, F. F., & Sasser, W. E. (1990, September/October). Zero
tivariate data analysis (4th ed.). Englewood Cliffs, New Jersey: Prentice defections: Quality comes to services. Harvard Business Review,
Hall. 68, 105 – 111.
Hall, G., Rosenthal, J., & Wade, J. (1993, November). How to make reen- Rust, R., & Zahorik, A. (1993). Customer satisfaction, customer retention,
gineering really work. Harvard Business Review, 71, 119 – 133. and market share. Journal of Retailing, 69, 145 – 156.
Heskett, J. L., Sasser Jr., W. E., & Schlesinger, L. A. (1997). The service Scharitzer, D., & Kollarits, H. C. (2000). Satisfied customers: Profitable
profit chain. New York: The Free Press. customer relationships: Pharmaceutical marketing: How pharmaceutical
Howard, J. A., & Sheth, J. N. (1969). The theory of buyer behavior. New sales representatives can achieve economic success through relationship
York: Wiley. management with settled physicians—an empirical study. Total Quality
Hunt, H. K. (1977). CS/D—overview and future research directions. In H. Management, 11, 7.
K. Hunt (Ed.), Conceptualization and measurement of customer satis- Sheth, J. N., Newman, B. I., & Gross, B. L. (1991). Consumption values
faction and dissatisfaction ( pp. 445 – 488). Cambridge, MA: Marketing and market choices: Theory and applications. Cincinnati, OH: South-
Science Institute. western Publishing.
Iacobucci, D., Grayson, K. A., & Ostrom, A. L. (1994). The calculus of Slater, S. F., & Narver, J. C. (1995, July). Market orientation and the
service quality and customer satisfaction: Theoretical and empirical learning organization. Journal of Marketing, 59, 63 – 74.
differentiation and integration. In T. A. Swartz, D. E. Bowen, & S. T. Slater, S. F., & Narver, J. C. (2000). Intelligence generation and superior
Brown (Eds.), Advances in services marketing and management, vol. 3 customer value. Journal of Academy of Marketing Science, 28(1),
(p. 1067). Greenwich, CT: JAI Press. 120 – 127.
Jacobson, R. (1992). The ‘Austrian’ school of strategy. Academy of Ma- Slywotzky, A. J. (1996). Value migration. Boston: Harvard Business School
nagement Review, 17(4), 782 – 807. Press.
Johnson, M. D., & Fornell, C. (1991). Framework for comparing customer Smith, M. C. (1991). Pharmaceutical marketing: Strategy and cases. Lon-
satisfaction across individuals and product categories. Journal of Eco- don: Academic Press (for Pharmaceutical Product Press).
nomic Psychology, 12(2), 267 – 286. Smith, J. B., & Barclay, D. W. (1997, January). The effects of organiza-
Kohli, A. K., & Jaworski, B. J. (1990, April). Market orientation: The tional differences and trust on the effectiveness of selling partner rela-
construct, research propositions, and managerial implications. Journal tionships. Journal of Marketing, 61, 3 – 21.
of Marketing, 54, 1 – 18. Tax, S. S., Brown, S. W., & Chandrashekaran, M. (1998, April). Customer
Kortege, G. D., & Okonkwo, P. A. (1993). Perceived value approach to evaluation of service complaint experiences: Implications for relation-
pricing. Industrial Marketing Management, 22, 133 – 140. ship marketing. Journal of Marketing, 62, 60 – 76.
Krishnamurthi, L., & Zoltners, A. (1994, Summer). Analysis of new Treacy, M., & Wiersema, F. (1995). The discipline of market leaders. Read-
product diffusion using a four-segment trial – repeat model. Marketing ing, MA: Addison-Wesley.
Science, 13(3), 224. Ulaga, W., & Chacour, S. (2001). Measuring customer-perceived value in
Lagace, R. R., Dahlstrom, R., & Gassenheimer, J. B. (1991, Fall). The business markets. Industrial Marketing Management, 30, 525 – 540.
relevance of ethical salesperson behavior on relationship quality: The Ulaga, W., & Eggert, A. (2002, March 20). Customer perceived value: A
pharmaceutical industry. Journal of Personal Selling & Sales Manage- substitute for satisfaction in business market. The Journal of Business
ment, XI(4), 39 – 47. and Industrial Marketing, 17(2 – 3), 10(12).
Lowenstein, M. W. (1997). The customer loyalty pyramid. Westport, Con- Woodruff, R. B. (1997). Customer value: The next source of competitive
necticut: Greenwood Publishing, Quorum Books. advantage. Journal of the Academy of Marketing Science, 25(2),
McInnes, W. (1964). A conceptual approach to marketing. In R. Cox, W. 139 – 153.
 J.M. Spiteri, P.A. Dion / Industrial Marketing Management 33 (2004) 675–687 687

Woodruff, R. B., & Gardial, F. S. (1996). Know your customer: New Joseph Spiteri, the president of a company that does research on markets for
approaches to customer value and satisfaction. Cambridge, MA: veterinary medical products, earned a DBA from the Wayne Huizenga
Blackwell. School of Business in Fort Lauderdale, FL.
Woodruff, R. B., Schumann, D. W., Clemons, S. D., Burns, M. J., &
Gardial, S. F. (1990). The meaning of satisfaction and dissatisfaction: Paul Dion, an associate professor of management at the Sigmund Weis
A themes analysis from the consumer’s perspective. Working paper, The School of Business at Susquehanna University, researches business-to-
University of Tennessee. business marketing and purchasing performance, marketing logistics, and
Zeithaml, V. A. (1988, July). Consumer perceptions of price, quality, and research methodology and statistics.
value: A means – end model and synthesis of evidence. Journal of Mar-
keting, 52, 2.