Journal of Business Logistics, 2011, 32(2): 167–179

 Council of Supply Chain Management Professionals

Electronic Logistics Service Quality (e-LSQ): Its Impact on the

Customer’s Purchase Satisfaction and Retention
Shashank Rao1, Thomas J. Goldsby2, Stanley E. Griffis3, and Deepak Iyengar2
1
Nova Southeastern University
2
University of Kentucky
3
Michigan State University

T he Internet retailing industry continues to grow rapidly. Several Internet retailers are, however, struggling to retain customers due
to the high level of competition among incumbents. We propose that customer satisfaction with the order fulfillment process is an
important determinant of overall customer satisfaction with the retailer, as well as with the extent of customer retention enjoyed by the
retailer. This paper offers a new concept for electronic logistics service quality and investigates the relationship between the quality of
online fulfillment and the ensuing retention of customers, using archival data on 260 online retailers. The structural equation model
results indicate that satisfaction with the physical distribution quality and cost are positively related with customer’s purchase satisfac-
tion and customer retention. Additionally, the results indicate that while purchase satisfaction is a strong indicator of customer reten-
tion, underlying drivers of purchase satisfaction do not have nearly the same impact on customer retention directly as they have on
purchase satisfaction.
Keywords: customer retention; e-commerce; online retailing; order fulfillment; secondary data

INTRODUCTION customer satisfaction and retention (Lee and Whang 2001).

Newton (2001) asserts the successful transfer of physical
Since progressing beyond the infancy stages of online mar- products between the online merchant and the customer ulti-
keting, Internet retailers have seen tremendous growth in mately determines whether the customer will place another
sales in recent years. By 2008, online customers accounted order. Such distribution effectiveness or efficient order fulfill-
for over $166 billion in total business-to-consumer (B2C) ment has been difficult to achieve though. Understanding the
sales in the United States (IR 500 2008). Though still small connection between online consumer behavior and shipping
as a percentage of total retail sales, the quarterly growth rate cost has become a kind of ‘‘Holy Grail’’ for online retailers
of 8.6% enjoyed by U.S. Internet retailers is substantially that most companies fail to achieve (Lewis et al. 2006). At
larger than the corresponding 1.3% growth of all U.S. retail the same time, others have emphasized the adverse impacts
(Agatz et al. 2008). Critical to this growth is the satisfaction of order fulfillment; that poor logistics capabilities and high
of consumers with the complete online purchasing experience shipping costs are key limitations of online retailing channels
(Montoya-Weiss et al. 2003). According to a recent study by (Grewal et al. 2004).
the National Retail Federation, over 55% of surveyed online The literature underscores that delivery performance and
retailers indicated that they were experiencing lower cus- shipping fees are relevant to the larger issue of how market-
tomer retention than expected, and that improving customer ing tactics impact online customer acquisition (Thomas 2001;
retention was essential for growth (Shop.org 2009). Hsu and Li 2006). Researchers are still cautioning that the
Many retailers at the dawn of e-commerce realized the order fulfillment process is one of the most critical opera-
importance of order fulfillment only after it was too late. tions for Internet-based sellers (De Koster 2003; Agatz et al.
While much attention was directed to establishing a strong 2008). Evidently then, research is needed on how order
web presence, an easy-to-navigate website, and directing pro- fulfillment impacts purchase satisfaction and retention in
spective consumers to the website, inattentiveness to servic- online retailing. The present article seeks to address this
ing the customer’s order doomed many online retailers issue. Our guiding research question is: How does perfor-
(Ricker and Kalakota 1999; Rabinovich and Bailey 2004). mance along the dimensions of order fulfillment satisfaction
This so-called ‘‘last mile’’ of the online transaction has gar- and its cost relate to purchase satisfaction and customer
nered much attention in recent years among retailers and the retention in online retailing? We investigate this central
research community (Esper et al. 2003; Rabinovich and research question using archival data on order fulfillment
Evers 2003; Boyer and Hult 2005). satisfaction, purchase satisfaction, and customer retention
Research has shown that the effectiveness with which across 260 online retailers.
online retailers fulfill orders is a significant determinant of For this analysis, we view order fulfillment in a manner
consistent with Pyke et al. (2001) as consisting of all activi-
Corresponding author: ties from the point of a customer’s purchase decision until
Thomas J. Goldsby, Gatton College of Business and Economics, the product is delivered to the customer. By building on
University of Kentucky, Lexington, KY 40506, USA; the established literature on physical distribution service
E-mail: [email protected] (PDS) quality (PDSQ) and logistics service quality (LSQ)
168 S. Rao et al.

typically applied in business-to-business (B2B) settings, we Mentzer et al. (1989) conceptualization, capturing the
introduce a scale for electronic LSQ (e-LSQ) in the B2C dimensions of availability, timeliness, and condition of the
environment. We demonstrate that high levels of e-LSQ products delivered.
(i.e., satisfaction with order fulfillment and the costs associ- Mentzer et al. (1999) build on the PDSQ construct in their
ated with it) are important in satisfying and retaining development of LSQ, deemed a broader view of the logistics
online consumers. This article will first present the underly- provision than that established in PDSQ. Through an exten-
ing drivers of order fulfillment satisfaction in an online con- sive data analysis they identify nine first-order constructs
text. A synthesis of the logistics, operations, and marketing that compose LSQ. These constructs include the attributes of
literature then leads to the conceptual model and hypothe- timeliness, availability (i.e., order accuracy), and order condi-
sis development. The data sources and methodology are tion found in the PDSQ, but also include information qual-
presented next, followed by a discussion of the data analy- ity, ordering procedures, order release quantities, order
sis. The results of this study are followed by the key impli- quality, order discrepancy handling, and personnel contact
cations for research and practice. Finally, directions for quality. In subsequent work, Mentzer et al. (2001) relate
future research are highlighted. these dimensions of LSQ to customer satisfaction through a
process model that captures the nine dimensions of LSQ as
aspects of order placement and order receipt.
THEORETICAL BACKGROUND Another derivation of PDSQ is found in Stank et al.
(1999) that conceptualizes logistics service performance as
Bowersox et al. (1968) suggest that the premise of trade is consisting of two factors: operational factors and relational
contingent on the supplier’s ability to physically distribute factors. In subsequent research, Stank et al. (2003) suggest
goods to a customer. The fundamental economic utility of that cost performance should be considered a factor separate
‘‘place’’—providing the right product at the right place at from operational performance, thus bringing the total
the right time in the right quantities and condition at the number of LSQ dimensions to three: operational factors,
right cost—constitutes the work of logistics. With notable relational factors, and the cost of the service. Operational
exceptions (Delfmann et al. 2002; Maltz et al. 2004; Rabino- performance factors can be understood as those activities
vich and Bailey 2004; Rabinovich et al. 2007), the prepon- performed by service providers that contribute to consistent
derance of research in the logistics field has focused on the quality, productivity, and efficiency. Operational elements
industrial provision of place utility in B2B environments, could include physical features of the service; for example,
examining how an industrial supplier can best accommodate characteristics of delivery that define and capture form, time,
industrial customers through effective and efficient logistics and place utilities of the service. Measurement of this con-
services. Though direct marketing and shopper-direct catalog struct calls for the customer’s subjective evaluation of the
sales have long been staples of commercial activity, the operational service performance delivered by the logistics
advent and widespread adoption of the Internet as a means provider (Stank et al. 1999).
for conducting commerce have substantially increased inten- Relational performance (RP) focuses on those activities
sity with which companies seek to reach end users of their that enhance service firms’ closeness to customers so that
goods and services, sometimes bypassing the conventional firms can understand customer needs and expectations and
brick-and-mortar channels of distribution altogether (Frazier develop processes to fulfill them (Stank et al. 1999). Several
1999; Lancioni et al. 2000). Interest then turns to whether studies have concluded that RP of logistics service has signif-
household consumers maintain similar expectations and act icant positive links to customer satisfaction and repurchase
in similar ways to industrial buyers when evaluating the pro- intentions (Innis and LaLonde 1994; Daugherty et al. 1998;
vision of order fulfillment services. More specifically, the Davis-Sramek et al. 2008). RP aspects have been examined
question arises as to whether the dimensions of order fulfill- in the literature in some detail, but the terminology used to
ment as outlined in the industrial logistics literature have operationalize this dimension has varied across studies.
similar impacts in the B2C domain, as they do in the B2B Within a B2B context, certain aspects of RP have all been
domain. operationalized through constructs such as relational behav-
The logistics discipline, coupled with marketing, has a ior, relationship quality, relationship marketing, and rela-
long tradition of customer service research. Among the tionship strength (Vickery et al. 2004).
seminal works is that of LaLonde and Zinszer (1976) who The aspect of logistics cost performance was operational-
define logistics customer service and distinguished its pre-, ized in Stank et al. (2003). Logistics cost satisfaction ele-
post-, and transactional elements. PDS focuses on the ments could be understood as those components of the
transaction aspects of the order fulfillment process, includ- logistics service that capture the satisfaction of a buyer of
ing those activities that support the flow of materials and logistics services with respect to the prices offered and the
related information from point of origin (stocking loca- costs imposed by the logistics service provider (Davis-Sramek
tion) to point of consumption (the consumer location) et al. 2008). Price plays a central role in the decision to
(Lambert and Stock 1993). Mentzer et al. (1989) examine maintain a relationship with a supplier because it is an
the outcomes of PDS found in the logistics literature dat- important driver of the perception of value, which, in turn,
ing back to the late 1960s, summating PDS as consisting has an effect on satisfaction (Varki and Colgate 2001). Over-
of availability, timeliness, and quality. Bienstock et al. all, then, operational elements related to product availability,
(1997) devise an extensive scale for PDSQ based on the product condition, delivery reliability, and delivery speed, in
Impact of e-LSQ on Customer’s Satisfaction 169

addition to relational elements such as communications, ent. However, the online B2C environment lacks many of
responsiveness, and cost elements have been found to influ- the interactional human elements of B2B channels (Yang
ence customer satisfaction and purchasing patterns positively and Jun 2002). For example, Rabinovich and Bailey (2004)
(Stank et al. 1999; Vickery et al. 2004; Davis-Sramek et al. suggest that physical distribution in online settings is ‘‘part
2008). of a context in which sellers and buyers are physically apart
The challenge with applying these logistics-based concep- and the service provided is aimed at goods and not at peo-
tualizations of service quality in the current analysis is ple’’ (p. 654). Similarly, in revising the SERVQUAL instru-
related to the B2B focus of these previous works. B2C ment for use in online environments (e-SERVQUAL),
service quality is much studied in the services marketing lit- Parasuraman et al. (2005) operationalize the order fulfillment
erature, primarily through the lens of the SERVQUAL construct using operational, but no relational items. Grewal
scale, originally devised by Parasuraman et al. (1985). et al. (2004) also suggest that lack of interpersonal trust and
SERVQUAL measures service quality using five underlying customer service issues were some of the limiters in an online
dimensions: reliability, responsiveness, empathy, assurance, context.
and tangibles (Parasuraman et al. 1985). Bienstock et al. Other researchers also seem to concur with conceptualiza-
(1997), however, argue that PDSQ is distinct from the tra- tions of online retail order fulfillment that are purely opera-
ditional service quality measures of SERVQUAL, since the tional in nature (Collier and Bienstock 2006). In the
traditional SERVQUAL ‘‘lays a very strong emphasis on marketing literature, RP is part of a broad umbrella of cus-
functional or process dimensions [which] may not tomer relationship management (CRM). CRM has been
adequately address the content validity of the [PDSQ] defined as integrating people, processes, technology, and sys-
construct’’ (p. 34). They further contend that while SERV- tems to increase customer satisfaction (Anton 1996; Anton
QUAL is best applied to consumer applications where con- and Hoeck 2002). In a B2C context, electronic-CRM
sumers receive intangible actions in the presence of the (e-CRM) was found to have a positive relationship with cus-
provider, people are replaced by ‘‘things’’ and there is phys- tomer satisfaction (Feinberg and Kadam 2002).
ical separation between provider and customer in industrial Due to the lack of personal contact between the buyer and
logistics services. the seller, RP has often been measured in an online context
Further, the SERVQUAL scale is not well adapted for the by asking the question on how well the merchant stands
B2C environment of electronic retailing. To address this behind their product or service after the sale of the product.
shortcoming, Parasuraman et al. (2005) devise e-SERV- Typical questions asked to validate the RP construct are
QUAL for the online retail context. The e-SERVQUAL evaluations of E-mail support, live chat, 1-800 support num-
scale is notable in that it does incorporate physical distribu- bers, complaining ability, ability to track the order, customer
tion (measured as order fulfillment effectiveness) as one service web page, and returns policy, to name a few (Boyer
dimension of overall service quality—the other dimensions et al. 2002; Feinberg and Kadam 2002; Kim et al. 2006;
being efficiency, system availability, and privacy. This one Heim and Field 2007). Hence, the two main limitations in
dimension does not, however, capture fully the richness of using RP in an online context are: (1) a physical barrier
the order fulfillment provision established in the industrial between the buyer and seller, and (2) the premise that RP
(i.e., B2B) logistics literature. typically occurs only after the sale in electronic retailing, and
Based on a detailed analysis of LSQ satisfaction across only then when a customer experiences a problem or diffi-
260 online retailers, we investigate the factor structure and culty. In sum, RP is more reactive in an online context com-
measurement of e-LSQ. We examine the relationships pared to a brick-and-mortar firm. Arguably, then, the RP
between aspects of online order fulfillment and customer’s aspects of physical distribution will be of less importance
purchase satisfaction (CPS). In addition, we examine the than the operational and cost aspects. Thus, we focus our
relationship between purchase satisfaction and customer hypothesis development on the two remaining dimensions
retention. Researchers have argued that the effectiveness of offered by Stank et al. (2003): operational performance and
order fulfillment in online retailing is an important driver of cost performance.
customer retention (Newton 2001), yet the presence and We concern ourselves with a customer’s satisfaction with
strength of this relationship remain untested. the operational factors of an online retailer’s PDSQ in
the online retailing environment and satisfaction with the
Conceptual model and research hypotheses economic consideration, or the PDS price (PDSP). Figure 1
presents the hypothesized model. Satisfaction with PDS
Researchers have noted that the relative importance of the quality (PDSQ) and satisfaction with PDS price (PDSP) are
underlying components of logistics service dimensions differ identified as separate factors, both predictors of the cus-
for B2C channels, as compared to B2B channels (Agatz et al. tomer’s purchase satisfaction (CPS), directly, and retention,
2008). For example, Vickery et al. (2004) suggest that of the indirectly. This is in line with the arguments proposed in the
three logistics service dimensions (operational, relational, literature for LSQ (Stank et al. 1999) and last-mile PDS in
and cost), RP would be of the least importance in B2C, e-tailing (Rabinovich and Bailey 2004). We discuss these
owing to its explicit focus on customer–supplier relationships ideas in more detail below. Theoretical foundations for the
in industrial settings. The original conceptualization of RP depicted relationships are subsequently highlighted, based on
in Stank et al. (1999) includes components of personal con- a detailed review of the literature, and lead to our research
tact between the logistics service provider and service recipi- hypotheses.
170 S. Rao et al.

Figure 1: Conceptual model and hypotheses. off between the quality, or benefits consumers perceive in a
product, relative to the perceived sacrifice expressed by the
price. With respect to PDS, in particular, researchers argue
PDS that sellers endeavor to offer service quality that is consistent
Quality H1 with the level of PDS utility (Kirmani and Rao 2000). Simi-
(PDSQ)
larly, Rabinovich and Bailey (2004, 655) reason that ‘‘Inter-
Customer’s
Purchase H3 Customer net retailers may increase the consumers’ time and place
Satisfaction Retention utility and afford higher levels of PDS quality when they can
(CPS) sustain higher PDS prices.’’ In H1, we hypothesize that
PDS H2
PDSQ should be positively associated with improved CPS.
Price
(PDSP)
Based on that hypothesis, and the preceding arguments in
this section, we may thus conclude that, ceteris paribus,
improved satisfaction with PDSP should be positively associ-
ated with CPS. Thus, our second hypothesis states:
PDSQ satisfaction and CPS
Several authors have addressed the issue of choosing an H2: PDSP positively affects CPS.
appropriate PDSQ level, in terms of time windows and lead
times, in order to satisfy online shoppers (for a detailed CPS and customer retention
review, see Agatz et al. 2008). It has been argued that Inter- It has been demonstrated in the literature that satisfaction
net retailers who perform favorably in terms of PDSQ are scores are not an end, in and of themselves; they are merely
able to generate customer value and, subsequently, attract a means to an end—customer retention (Jones and Sasser
loyal customers (Rao et al. 2000; Rabinovich 2004; Rabino- 1995; Reichheld 1996). Several organizations have allocated
vich and Bailey 2004; Boyer and Hult 2005). Similarly, other time and money in order to study how satisfaction ratings
researchers have argued that Internet retailers who outper- affect customer retention (Bolton 1998). Arguably, when
form the competition on PDSQ are able to develop a loyal retailers measure satisfaction with a purchase they are, in
customer base and charge higher prices (Chen and Hitt effect, trying to measure the likelihood that the customer will
2002). make a repeat purchase.
Additionally, Rohm and Swaminathan (2004) demonstrate Indeed, studying the relationship between CPS and cus-
that one of the primary reasons for buyers to shop online is tomer retention is not merely one of academic tautol-
convenience. Wolfinbarger and Gilly (2001, 41) argue that ogy—some researchers, finding no significant link between
among the conveniences considered before shopping online, the two, have even labeled satisfaction measurement a ‘‘trap’’
the convenience of ‘‘not having to go out’’ is one of the and argued for curtailing satisfaction measurement efforts
more important reasons for shopping online. As previously (Reichheld 1996). Thus, the relationship between purchase
noted, PDS consists of the activities that support the flow of satisfaction and actual customer retention is an important
materials and related information from point of origin issue worthy of research attention, particularly in online
(stocking location) to point of consumption (the consumer retailing where the number of options available to the cus-
location) (Lambert and Stock 1993). PDSQ should thus be tomer continues to grow (Srinivasan et al. 2002).
directly related to the CPS, since improved PDSQ would We posit that CPS should be associated with customer
suggest improved convenience associated with shopping retention. Some empirical examination of this link has been
online. Consequently, retailers who are able to provide better initiated by Nelson et al. (1992), who demonstrate the rela-
PDSQ should yield greater satisfaction with the purchase. tionship of customer satisfaction to customer retention in the
These arguments lead to our first hypothesis: health care sector. Subsequently, other researchers examined
the relationship between satisfaction and retention in retail
H1: PDSQ positively affects the CPS. banking (Rust and Zahorik 1993; Halowell 1996) and in
mobile telecommunications (Gerpott et al. 2001). Finally,
PDSP satisfaction and CPS Davis-Sramek et al. (2008) investigate this relationship in a
It is valuable to model online PDSP satisfaction as a sepa- B2B electronic goods setting. However, conclusive evidence
rate dimension of the online order fulfillment process, dis- of this link in B2C online retail setting is lacking. Building
tinct from PDSQ. Mentzer et al. (1989) surmise that two of on the finding of the previously cited works in brick-and-
the central questions in the PDS literature are: (1) identifica- mortar and B2B contexts, we propose our third hypothesis:
tion of the elements of PDS; and (2) determination of the
cost effects of providing PDS. The second hypothesis deals H3: CPS positively affects customer retention.
with the perceived economic effects associated with PDS,
from the customer’s perspective. In summary, our overall research model focuses on the
Economic theory suggests there should be a relationship impact of PDS performance factors on CPS in online retail-
between PDSP satisfaction and CPS. In the marketing litera- ing contexts, an emerging area of interest for logisticians.
ture, value for the customer is usually operationalized as a Additionally, we also study the impact of purchase satisfac-
trade-off between quality (benefit) and cost (price) (Bolton tion on customer retention, a critical issue in the marketing
and Drew 1991). As Monroe (1990) notes, value is the trade- literature. Thus, we investigate how PDS-related factors
Impact of e-LSQ on Customer’s Satisfaction 171

indirectly impact customer retention in online retailing—an each of the 261 online retailers were then manually tran-
issue that has heretofore received little attention in the litera- scribed by two of the researchers.
ture. In the larger scheme, we investigate how logistics helps BizRate.com, which is one of the most popular price com-
retain customers. This is an idea that has been espoused by parison engines on the web, routinely surveys online retail-
logisticians, but has yet to be examined thoroughly. ers’ customers and asks them to evaluate the online retailers’
services. The survey results are published on the
BizRate.com website and are freely available to users who
RESEARCH METHOD wish to engage in comparison shopping. Fifteen aspects of
online retailers’ services are evaluated using a 10-point scale,
This section reviews the research design. Discussion of the and an overall measure of the average of the 15 measures is
sample design is followed by a description of the measure- also provided. BizRate.com data have often been used for
ment sources and scales. Results of the measurement model purposes of academic research (Pan et al. 2002; Reibstein
assessment are presented as well. 2002; Jiang and Rosenbloom 2004; Srinivasan and Moorman
2005) and have high credibility. Reibstein (2002) shows that
Sample design the BizRate survey data were free of nonresponse bias.
Additionally, there is evidence to suggest that the retailers
The research model calls for data collection across a large themselves take the BizRate results quite seriously. The fact
cross-section of online retailers. Moreover, the data need to that thousands of online retailers volunteer to participate in
accurately reflect figures for the online channel of these the data collection supports this belief (Thirumalai and Sinha
retailers in order to enable appropriate comparisons. The 2005).
Internet Retailer Top 500 Guide (IR 500) for 2006 and The different aspects of the purchase process captured in
2007 served as our initial sampling frame. The IR 500, which the BizRate survey are consistent with aspects of the online
ranks online retailers based on the volume of their online purchasing process identified in the academic literature (Field
sales, is one of the only comprehensive rankings of U.S. et al. 2004). BizRate’s online customer survey has two parts.
online retailers (Brohan 2008). This database lists the top The first part of the survey is delivered immediately after a
500 online retailers in the United States, with total sales purchase is completed. When an online purchase is made
exceeding $166 billion, accounting for over 60% of the total from a BizRate.com associated website, BizRate.com inde-
online retailing market (Howlett 2008). The database is pendently invites buyers to participate and complete a sur-
updated annually, and published in print and online form by vey. The survey uses a 10-point scale (where 1 = poor,
Internetretailer.com—a monthly national business magazine. 6 = satisfactory, and 10 = outstanding) to measure levels
Internetretailer.com also organizes the leading e-commerce of satisfaction with: ease of finding the item; selection of
industry conference, and publishes four leading e-commerce products; clarity of product information; product prices;
directories. With a readership exceeding 42,000 senior look and design of the website; shipping and handling
executives primarily from retail chains, independent stores, charges; the variety of shipping options; and the clarity of all
catalogs, virtual merchants, and brand name manufacturers, charges before submission of the order. The second part of
the IR 500 is the leading standard of the online retailing mar- the survey is e-mailed shortly after the scheduled delivery
ketplace, thus adding veracity to its credibility (Reuters 2009). date, and customers are asked to rate their satisfaction levels
While the use of this data indeed limits our focus to online with: availability of the product, on-time delivery, order
retailers in the United States, the United States is by far the tracking, whether the product met expectations, and cus-
largest online retailing market in the world. According to a tomer support. The six measures in the BizRate survey that
study by Retail Week (2009), the size of the online retailing are of interest as independent variables in this study are vari-
market (by total dollar value of sales) in the United States is ety of shipping options, availability of the item, on-time
more than the size of the online retailing market in France, the delivery, order tracking, shipping and handling charges, and
United Kingdom, Germany, Poland, and Russia combined. clear statement of all charges prior to order submission. We
The IR 500 divides the U.S. online retailers into four cate- propose that the measures for PDSQ are available shipping
gories: web-only, store-based, catalog-based, and consumer options, item availability, on-time delivery, and order track-
brand manufacturers. Of these 500 online retailers, consumer ing. Alternatively, the measures of interest for PDSP are
brand manufacturers (51 in total) were dropped from the shipping and handling charges and all charges stated clearly
sample since they did not fit into our categorization of pure prior to order submission. The variables capturing CPS (first
retailers, thus leaving us with a potential sample pool of 449 dependent variable) are overall experience with the purchase,
cases. whether the product met expectations, and likelihood of buy-
ing from the store again.
Measurement sources and scales Customer retention (our second dependent variable) was
measured as the percentage of shoppers at an online retailer
Customer-based measures (PDSQ, PDSP, and CPS) for the who were repeat buyers. This information on customer reten-
449 firms in the sample were gathered using data from Biz- tion was obtained from the IR 500. To collect these data, IR
rate.com. We searched for the URLs of the 449 firms in the 500 researchers contact hundreds of retailers over a six-
BizRate.com data. We were able to obtain full data on 261 month period. In order to validate the quality of these data,
of the 449 retailers in this way. The customer ratings for we also obtained the percentage of repeat shoppers at the
172 S. Rao et al.

Table 1: Measurement items

Factor Measures Source Average SD Skewness

Satisfaction with PDS quality Satisfaction with the variety of shipping options BizRate 8.43 .61 )2.79
Satisfaction with item availability (In stock) 8.69 .73 )6.16
Satisfaction with the ability to track orders 8.63 .76 )4.95
Satisfaction with the on-time product delivery 8.74 .78 )5.13
Satisfaction with PDS price Satisfaction with shipping charges BizRate 7.80 .89 )0.26
All charges were presented clearly prior to order 6.87 .79 )0.24
submission
Customer purchase satisfaction Store satisfaction BizRate 8.48 .78 )4.90
Store loyalty 8.62 .78 )5.20
Expectations fulfilled 7.91 .69 )4.01
Customer retention Percentage of shoppers who are repeat buyers IR500 0.30 .12 0.55

Note: PDS, Physical distribution service.

online stores from Quantcast.com scores. Quantcast is a measurement items in the study (Table 1). The mean values
web-traffic monitoring service that rates websites by ranks. for the items representing PDSP, PDSQ, and purchase satis-
Participating websites voluntarily insert Quantcast HTML faction (CPS) were all somewhat skewed to the high end of
code inside the pages that they wish to monitor. This code the 10-point scales, ranging from a low of 6.87 for satisfac-
allows Quantcast to keep track of the traffic directed toward tion with shipping charges, to a high of 8.74 for on-time
those websites. The correlation between the data obtained delivery. As a result, it was considered appropriate to study
from these two sources was extremely high (r > .87, the Box-Cox plots for all the variables in order to normalize
p < .001). Thus, for tractability and parsimony, it was con- them. This was done in SPSS (v. 15.0; IBM SPSS, Chicago,
sidered prudent to retain only one set of these scores (IR IL), and logarithmic transformations were subsequently
500). applied to normalize the data. Subsequently, Shapiro–Wilk
Of the 449 e-tailers in the potential sample pool, all the tests indicated that the violation of normality had been suffi-
necessary data were available for 261 e-tailers. Of these, ciently corrected.
Amazon.com1 was removed since it was a substantial outlier, The correlations among the individual items are presented
leaving us with 260 e-tailers. The data, measures, and sources in Table 2. The BizRate scales demonstrated sound internal
are summarized in Table 1. Descriptive statistics of the data consistency after the transformation; Cronbach’s alpha for
are also provided in the table. In order to test for any biases PDSQ was .943, .98 for PDSP satisfaction, and .88 for CPS.
that may have arisen by dropping the data points with miss- All these values are well above the minimum recommended
ing observations, we compared the two groups (dropped vs. cut-off value of .70 (Nunnally 1978). Further validity assess-
retained) in terms of sales (t = 0.282, p = .77), number of ment of the measurement instrument was carried out
years experience in online retailing (t = 1.24, p = .21), cus- by using a confirmatory factor analysis (CFA) in AMOS
tomer retention percentage (t = 0.81, p = .41), and year- (v. 17.0; IBM, SPSS) and is presented in Table 3. It should
over-year sales growth (t = 0.39, p = .69). As can be seen, be noted that customer retention represents a single item
none of the differences between the groups was statistically construct and is thus dropped from the CFA.
significant at reasonable alpha levels, thus indicating that the The CFA provides a more rigorous method of assessing
dropped e-tailers were not statistically different to the ones measure unidimensionality as compared to Cronbach’s alpha
used for the analysis. or exploratory factor analysis (Anderson and Gerbing 1988).
The chi-square statistic obtained in the CFA was 84.9, with
Measurement model assessment 24 df (p < .001), for a normed chi-square statistic (the ratio
of the chi-square to the degrees of freedom) of 3.5:1. While
In order to use any scale in research reliably, it is important no consensus regarding a specific value for this normed chi-
to ensure that it has sufficient convergent and discriminant square statistic has been reported, ratios varying from 2:1 to
validity. Convergent validity measures the extent to which 5:1 have been offered as acceptable fit (Arbuckle 1997). The
scale items correlate strongly with other items used to mea- Bentler–Bonett Normed Fit Index (BBNFI), Bentler–Bonett
sure the same construct (Churchill 1979). Preliminary analy- Non-Normed Fit Index (BBNNFI), and Comparative Fit
sis of the data indicated univariate nonnormality for several Index (CFI) have also been proposed as model fit measures.
Values on these indices close to 1.0 are considered to be very
1
With $17 billion in online sales in 2007, Amazon.com alone good, and scores above 0.90 are generally considered accept-
accounted for over 15% of our overall online sales. Conse- able (Fornell and Larcker 1981; Bagozzi and Yi 1988;
quently, it was considered prudent to drop it as an unrepre- Hu and Bentler 1995). As can be seen in Table 3, all three
sentative data point. of these indices for the data are above the minimum
Impact of e-LSQ on Customer’s Satisfaction 173

Table 2: Correlation matrix

Charges
Shipping Item Order On-time Shipping presented Store Store Expectations Customer
options availability tracking delivery charges clearly satisfaction loyalty fulfilled retention

Shipping options 1.00

Item availability 0.74 1.00
Order tracking 0.74 0.82 1.00
On-time delivery 0.75 0.82 0.87 1.00
Shipping charges 0.38 0.37 0.30 0.30 1.00
Charges presented 0.47 0.37 0.30 0.30 0.89 1.00
clearly
Store satisfaction 0.64 0.69 0.62 0.63 0.46 0.47 1.00
Store loyalty 0.64 0.68 0.61 0.63 0.48 0.47 0.78 1.00
Expectations fulfilled 0.63 0.72 0.70 0.70 0.30 0.39 0.74 0.83 1.00
Customer retention 0.61 0.62 0.53 0.54 0.58 0.58 0.56 0.56 0.56 1.00

Note: All correlations are statistically significant at .05 level of significance.

Table 3: Measurement model results

Standardized model (maximum likelihood)

PDS quality PDS price Purchase satisfaction t-value

Shipping options .71 *

Item availability .89 14.76
Order tracking .98 15.88
On-time delivery .97 15.80
Shipping charges .98 *
Charges presented clearly .97 40.27
Store satisfaction .98 *
Store loyalty .91 44.33
Expectations fulfilled .79 29.35
Fit statistics
Chi-square statistic 84.9 (24 df), p < .01
Normed Fit Index 0.96
Non-Normed Fit Index 0.95
Comparative Fit Index 0.97
Root mean squared residual 0.035

Note: *Indicates that the paths were fixed for scaling and the t-statistic is not available.

threshold of 0.9, thus indicating that the measurement model test for discriminant validity. In this approach, comparisons
fit is acceptable. Additionally, the root mean squared resid- are made between the original measurement model and alter-
ual (RMSR) of 0.035 indicates a sound factor structure. native models by fixing the correlation among latent vari-
Anderson (1987) conveys that convergent validity is sup- ables to one. If any of the alternative models exhibits better
ported when factor loadings (lambdas) demonstrate that fit relative to the original model, then a lack of discriminant
measurement items load significantly on their designated validity can be implied (Bagozzi and Yi 1988). We tested the
latent variables. Table 3 suggests that this is indeed the case, discriminant validity, and all the constrained models had
as the lowest standardized loading was .71, which is accept- comparatively worse fits as compared to the unconstrained
able. Finally, high t-values for all the loadings provide fur- model (chi-squares of 159, 252, and 273, as compared to the
ther evidence of good factor loadings. original chi-squared value of 84.9). This suggests that the
Discriminant validity refers to the extent to which mea- measurement model has sufficient discriminant validity as
sures of different constructs are distinct from one another well as convergent validity. Additionally, composite reliabil-
(Churchill 1979). The nested model approach was used to ity was also assessed using the average variance extracted
174 S. Rao et al.

(AVE) procedure suggested by Fornell and Larcker (1981). Figure 2: Structural model results.
The AVE is the variance in the indicators explained by the
common factor, in relation to the variance due to random
measurement error. Dillon and Goldstein (1984) suggest that PDS H1: 0.74, t = 12.21,
Quality p < .001
an AVE of greater than .50 for each construct indicates that (PDSQ)
the validity of both the construct and the individual vari- Customer’s
H3: 0.44, t =
7.92, p < .001
ables is high. Additionally, Fornell and Larcker (1981) pro- Purchase Customer
Satisfaction Retention
pose that as long as the square root of the AVE exceeds the (CPS)
correlation between the latent constructs, there is sufficient PDS
evidence of discriminant validity in the model. As can be Price H2: 0.15, t = 3.73,
(PDSP)
seen from Table 4, these criteria for convergent and discrimi- p < .001
nant validity were met. Given the overall soundness of the
measurement model, our attention now turns to a discussion
of the structural model. price are likely to serve as prime differentiators among hosts
of retailers offering similar products at the click of a button.
One of the most interesting findings of this research is
RESULTS AND DISCUSSION obtained by combining H1 and H2, and looking at them
together along with the significant correlation between the
Given the support for sound internal structure of the two exogenous factors of PDSQ and PDSP (t = 0.518,
measurement model, analysis of the structural model allows p < .001). We demonstrate that both, PDSQ and PDSP
us to test the hypothesized relationships. As with the affect purchase satisfaction. Often, however, price is seen as
measurement model, the structural equation modeling was the limiting factor in providing services of exceedingly high
conducted in AMOS (v. 17.0). The structural model results quality, and firms are usually in a position to improve qual-
are presented in Figure 2 and discussed below. ity substantially, provided they are also able to increase the
The chi-square statistic (280.9, 32 df) is significant cost (Bolton and Drew 1991). Our results suggest that
(p < .001). However, the other fit indices, including the NFI improving shipping performance by charging higher shipping
(0.927), NNFI (0.909), CFI (0.935), and RMSR (0.044), all costs may not help online retailers improve purchase satisfac-
indicate sound model fit. Thus, we may reasonably conclude tion. The high correlation between the exogenous factors
that the model offers a good fit for the data (Fornell and indicates that PDSP and PDSQ are correlated. Improved
Larcker 1981; Bagozzi and Yi 1988; Hu and Bentler 1995). performance along the PDSQ dimension, at the expense of
Additionally, all the path coefficients between latent variables PDSP performance, may not result in improved customer
are significant at p < .001, which indicates that all the satisfaction to the extent desired. Thus, online retailers
hypotheses were supported. should seek ways to improve PDSQ while simultaneously
Significant path coefficients between PDSQ fi CPS and reducing associated costs. The emergence of drop-shipping
PDSP fi CPS provide strong support for H1 and H2, arrangements in online retail (Rabinovich et al. 2007, 2008)
indicating that there is a positive relationship between satis- might be attributed to this realization. Drop-shipping aims
faction with PDSQ ⁄ PDSP and purchase satisfaction. In addi- to reduce the costs of shipping products by centralizing
tion, the high path coefficients for these linkages indicate warehousing and storage through outsourcing these activities
that online retailers can improve their CPS by improving to specialist firms who would be better able to achieve the
performance along these dimensions. Though not tested in necessary scale economies.
the current analysis, PDSQ and PDSP may prove even more Additionally, we also went a step beyond previous
influential in online environments, where delivery service and research and examined the linkage between purchase satis-
faction and customer retention. The link between measures
Table 4: Variance extracted of purchase satisfaction and customer retention has been dif-
ficult to measure (Mittal and Kamakura 2001). One of the
Construct AVE Sqrt AVE most common reasons for this is the time lapse between a
customer’s satisfaction rating and the next purchase is often
Variance extracted quite long, and most survey research is not able to capture
PDSQ .82 .90 both phenomena. In this study, we show that the
PDSP .96 .98 relationship between purchase satisfaction and customer
Purchase satisfaction .93 .96 retention appears to hold quite strongly. This link is of
Estimate considerable importance for any firm, and especially so for
Correlations firms in an industry as dynamic as online retail. Fornell et al.
PDSQ, PDSP .367 (2006) demonstrate that firms with more loyal customers
PDSP, purchase satisfaction .455 tend to outperform the competition on key financial metrics
PDSQ, purchase satisfaction .647 such as earnings and return on assets, among others. It
appears that online retailers who are interested in outper-
Notes: AVE, average variance extracted; PDSQ, physical distribu- forming the competition should consider taking steps to
tion service quality; PDSP, physical distribution service price. improve their order fulfillment performance. Interestingly,
Impact of e-LSQ on Customer’s Satisfaction 175

the mean values for PDSQ items were quite high (8.43 and is a surprising finding and might indicate suppression in the
above on the 10-point scale), while the PDSP items (with model in the relationship between satisfaction and retention
means of 7.80 and 6.87, respectively) appear to offer remain- (Cohen and Cohen 1983). Testing a similar model in a B2B
ing room for improvement. bricks-and-mortar setting, Davis-Sramek et al. (2008) found
a similar result between the same constructs as those used in
Post hoc investigation this article. They posited that the nonsignificance was
because the link between satisfaction and retention was being
The strong linkage between purchase satisfaction and cus- mediated by some exogenous variable.
tomer retention suggests the need for further investigation. Of An alternative to Davis-Sramek et al. (2008) exogenous
particular interest is the relationship between PDSQ ⁄ PDSP variable explanation appears possible. Items purchased from
and customer retention. Based on the previous results that the Internet are often commodity-type products. It is possi-
showed linkages between PDSP ⁄ PDSQ and purchase satisfac- ble then that when considering purchase satisfaction simulta-
tion, and also the linkage between purchase satisfaction and neously with logistics performance (PDSQ and PDSP) the
customer retention, it is reasonable that one may observe a performance of the retailers’ fulfillment operations (on qual-
direct linkage between PDSP ⁄ PDSQ and eventual customer ity and price) dominate the commodity products’ characteris-
retention. Post hoc analysis was performed to examine this tics, and is the reason the more product-centric factor
possibility. The results for this re-specified model are shown in (purchase satisfaction) drops out. Essentially, for commod-
Figure 3. ity-like products sold online, the customer’s repurchase
While the chi-square was still significant (147.7, 30 df, behavior is more motivated by logistics performance than it
p < .001), the other four fit indices were good (NFI = is on the product that the retailer sells. Though conjecture at
0.9427, NNFI = 0.932, CFI = 0.955, RMSR = 0.034), this point, given the significance this would place upon the
thus indicating sound model fit. Additionally, the PDSP fi efficacy of logistics operations in the overall business of on-
Retention and PDSQ fi Retention paths were also signifi- line B2C, it appears worthy of further investigation.
cant at 99%. However, an interesting observation emerged: To sum up, this study has investigated a key research area
the path coefficient for PDSP fi Retention was, in fact, that is gaining the attention of logistics researchers: service
stronger than that for PDSQ fi Retention. This result is aspects of order delivery in online retailing. While several
different from what was earlier obtained for purchase satis- issues in this field, such as the importance of partnerships
faction (Figure 2), wherein we observed that PDSQ fi CPS (Esper et al. 2003), returns management (Mollenkopf et al.
had a higher path coefficient than PDSP fi CPS. We pro- 2007), and order fulfillment (Rabinovich and Bailey 2004;
pose a reason for this from the consumer behavior literature. Rabinovich et al. 2008), have been examined, an inquiry into
Researchers have suggested that the temporal distance the dynamics of physical distribution quality and price on
between the time a customer gives a satisfaction rating and the CPS and actual customer retention was absent. The pres-
the next purchase is often quite high (Mittal and Kamakura ent article has taken a step in this direction. We operational-
2001). Thus, consumers may tend to undervalue certain ize PDSP and PDSQ in a B2C context and demonstrated
aspects of the previous purchase and overvalue other aspects that both factors are important in explaining CPS. However,
(Hennig-Thurau and Klee 1997). It is possible that in the although as PDSQ seems to be more important than PDSP
short term (purchase satisfaction), consumers place a higher for explaining immediate purchase satisfaction, they appear
emphasis on PDSQ; but as time passes, PDSQ fades from almost equally important in actual retention of customers, as
memory, and consumers only remember the PDSP. This, per our post-hoc findings.
however, is mere conjecture on our part, but nevertheless Additionally, we demonstrated that purchase satisfaction
provides an avenue for future research. and customer retention are strongly related to each other.
Finally, on the addition of the direct paths from PDSQ Research linking satisfaction with actual retention has been
and PDSP to customer retention, the CPS fi Retention hard to come by. Mittal and Kamakura (2001) found some
link became nonsignificant. In fact, while satisfaction and support for the satisfaction-repurchase behavior linkage, but
retention are correlated, the results indicated a negative were confronted with differing results depending upon the
(albeit nonsignificant) path coefficient between the two. This demography of their samples. Paulssen and Birk (2007) were
similarly unable to demonstrate consistent support for the
Figure 3: Results for re-specified model. satisfaction-repurchase behavior connection. In light of these
past challenges, our results provide a clarity we believe lack-
ing in past samples. Using archival data from various
PDS 0.24, t = 3.36, p < sources, we demonstrate that satisfaction is strongly related
Quality .001
(PDSQ)
with retention, at least in online retailing. Thus, we demon-
0.75, t =
14.52, p -0.035, t =
strate empirically, a key argument that the logistics discipline
Customer’s
< .001 Purchase -6.35, p > .1 Customer has often espoused, but rarely demonstrated—that satisfac-
0.14, t = Satisfaction Retention tion is dependent not only on marketing activities but on the
4.79, p < (CPS)
.001 underlying logistics activities as well.
PDS
Price Finally, we operationalize the LSQ dimensions in an
(PDSP) 0.40, t = 7.76,
p < .001
online setting, in effect, measuring what we term e-LSQ. We
demonstrate that operational and cost factors of LSQ are
176 S. Rao et al.

indeed important influences on the CPS and retention, in a Moreover, our data do not represent a cross-sectional
manner similar to the study of logistics service performance panel and, in that sense, are not longitudinal. We have not
in traditional B2B contexts. measured PDS satisfaction, CPS, and retention for the same
customers over time. Our measures could be reflective of dif-
ferent customer groups. Future research could overcome this
LIMITATIONS AND FUTURE RESEARCH limitation by conducting a longitudinal study of the same set
of customers over time, to test the robustness of our results.
This study, while providing some key insights into satisfac- Additionally, we have operationalized two of the three sug-
tion in online retail, has limitations. These limitations are gested LSQ constructs in this study. While we have reason
presented below. to believe that RP aspects will be less important than opera-
Thirumalai and Sinha (2005) demonstrate that overall tional or cost measures in an online B2C setting, this belief
order fulfillment satisfaction in online retail varies with prod- has not yet been validated by research. Future research
uct type. However, they did not distinguish between our two should further examine the dimensionality of e-LSQ, and
components of order fulfillment (PDSP ⁄ PDSQ). It is possible evaluate the influence of RP on CPS, if indeed RP turns out
that PDSQ and PDSP may also vary based on product type. to be an integral aspect of e-LSQ. Further, the issue of
Thus, a natural extension of this study could be to examine repurchase behavior for commodity products being more
whether product type has any moderating influence on the related to logistics performance than to product characteris-
relationships between PDSP ⁄ PDSQ and CPS. Additionally, tics deserves substantial attention.
the relationships between PDSP ⁄ PDSQ and customer reten- Finally, it could be inferred that our sample set of firms
tion may also be moderated by product type, and this is an in this study does not represent a truly random sample.
area worthy of future investigation. There is an obvious selection bias in the sample of firms
Researchers have demonstrated that back-end order fulfill- itself. In this study, we have looked at data from the top
ment aspects of online retailers differ across online retailer 500 Internet retail firms in the United States, and it could
types (i.e., web-only, store-based, catalog-based retailers) be argued that our results are skewed toward larger firms.
(Rao et al. 2009). Linking the results of the current study While this argument may have some merit, we should reiter-
with the cited work on retailer types, it may become logical ate that our original 500 online retailers accounted for over
to argue that the relation between PDSP ⁄ PDSQ and pur- 60% of the total online retail sales in the United States,
chase satisfaction could be moderated by channel types. The and even the reduced sample of 260 retailers accounted for
results of this study suggest that improvements to the order over 35% of total online retail sales in the United States.
fulfillment process can yield benefits. However, whether the Thus, we believe that the data are fairly representative of
order fulfillment performance is equally important for all the U.S. market as a whole. As we have already pointed
types of retailers could prove interesting. Similarly, it has out, the U.S. online retailing market is larger than that of
been argued that the relationship between satisfaction and several of the world’s leading economies combined. Thus,
retention may be dependent on customer type (Mittal and we believe that these results should also be valid in other
Kamakura 2001). It has been demonstrated that factors such online retailing markets. Nevertheless, we do agree that
as education level of customers, and age, for example, future research should seek to overcome this limitation to
directly impact whether satisfaction translates into retention. see if our findings hold with a truly random sample of
Such moderating effects of customer demographics need to online retailers, and with retail operations founds in other
be considered in developing a more holistic model of order parts of the world.
fulfillment and customer retention.
Additionally, we did not consider the impact of product
performance on the relationship between satisfaction and REFERENCES
retention. Conceivably, how the product performs in actual
usage can have a very strong relationship on whether the Agatz, N.A.H., Fleischmann, M., and van Nunen, J.A.E.E.
customer will actually buy from the online retailer again. 2008. ‘‘E-Fulfillment and Multi-Channel Distribution—A
In this study, we did not consider the influence that such Review.’’ European Journal of Operations Research
product performance might have on customer retention, 187(2):339–56.
which might be another reason for our not having found a Anderson, J.C. 1987. ‘‘An Approach for Confirmatory Mea-
relationship between CPS and Retention in the post hoc surement and Structural Equation Modeling of Organiza-
model. This omission on our part, however, was deliberate; tional Properties.’’ Management Science 33(4):525–41.
as in this study, we were only interested in finding the Anderson, J.C., and Gerbing, D.W. 1988. ‘‘Structural Equa-
impact of the last-mile logistics elements on customer reten- tion Modeling in Practice: A Review of the Two-Step
tion. Nevertheless, future research should consider product Approach.’’ Psychological Bulletin 103(3):411–23.
performance while studying customer loyalty and retention Anton, J. 1996. Customer Relationship Management. New
in online retailing, keeping in mind the likely confounds York: Prentice-Hall.
that may be uncovered between two aspects of the product- Anton, J., and Hoeck, M. 2002. e-Business Customer Service.
service experience. Further, exploration of the influence of Santa Monica, CA: The Anton Press.
the customer’s returns experience on overall satisfaction is Arbuckle, J.L. 1997. Amos Users’ Guide, Version 3.6. Chi-
merited. cago, IL: Small Waters.
Impact of e-LSQ on Customer’s Satisfaction 177

Bagozzi, R.R., and Yi, Y. 1988. ‘‘On the Evaluation of Feinberg, R., and Kadam, R. 2002. ‘‘E-CRM Web Service
Structural Equation Models.’’ Journal of the Academy of Attributes as Determinants of Customer Satisfaction With
Marketing Science 16(1):74–94. Retail Web Sites.’’ International Journal of Service Indus-
Bienstock, C., Mentzer, J.T., and Bird, M.M. 1997. ‘‘Mea- try Management 13(5):432–51.
suring Physical Distribution Service Quality.’’ Journal of Field, J., Heim, G., and Sinha, K. 2004. ‘‘Managing Quality
the Academy of Marketing Science 25(1):31–44. in the E-Service System: Development and Application of
Bolton, R. 1998. ‘‘A Dynamic Model of the Duration of the a Process Model.’’ Production and Operations Manage-
Customer’s Relationship With a Continuous Service Pro- ment 13(4):291–306.
vider: The Role of Satisfaction.’’ Marketing Science Fornell, C., and Larcker, D. 1981. ‘‘Evaluating Structural
17(1):45–65. Equation Models With Unobservable Variables and Mea-
Bolton, R., and Drew, J. 1991. ‘‘A Multistage Model of Cus- surement Error.’’ Journal of Marketing Research 28(1):39–
tomers’ Assessments of Service Quality and Value.’’ Jour- 50.
nal of Consumer Research 17(4):375–84. Fornell, C., Mithas, S., Morgenson, F., and Krishnan, M.S.
Bowersox, D.J., Smykay, E.W., and LaLonde, B.J. 1968. 2006. ‘‘Customer Satisfaction and Stock Prices: High
Physical Distribution Management: Logistics Problems of Returns, Low Risk.’’ Journal of Marketing 70(1):3–14.
the Firm. New York: Macmillan. Frazier, G.L. 1999. ‘‘Organizing and Managing Channels of
Boyer, K.K., Hallowell, R., and Roth, A.V. 2002. ‘‘E-ser- Distribution.’’ Journal of the Academy of Marketing Sci-
vices: Operating Strategy—A Case Study and a Method ence 27(2):226–40.
for Analyzing Operational Benefits.’’ Journal of Opera- Gerpott, T., Rams, W., and Schindler, A. 2001. ‘‘Customer
tions Management 20(2):175–88. Retention, Loyalty and Satisfaction in the German
Boyer, K., and Hult, T. 2005. ‘‘Customer Behavior in an Mobile Cellular Telecommunications Market.’’ Telecom-
Online Ordering Application: A Decision Scoring munications Policy 25(4):249–69.
Model.’’ Decision Sciences 36(4):569–98. Grewal, D., Iyer, G.R., and Levy, M. 2004. ‘‘Internet Retail-
Brohan, M. 2008. ‘‘The Top 500 Guide.’’ (June). http:// ing: Enablers, Limiters, and Market Consequences.’’ Jour-
www.internetretailer.com/article.asp?id=22579. nal of Business Research 57(7):703–13.
Chen, P.-Y., and Hitt, L. 2002. ‘‘Measuring Switching Costs Halowell, R. 1996. ‘‘The Relationships of Customer Satisfac-
and the Determinants of Customers Retention in Inter- tion, Customer Loyalty, and Profitability: An Empirical
net-Enabled Businesses: A Study of the Online Brokerage Study.’’ International Journal of Service Industry Manage-
Industry.’’ Information Systems Research 13(3):255–74. ment 7(4):27–42.
Churchill, G. 1979. ‘‘A Paradigm for Developing Better Heim, G.R., and Field, J.M. 2007. ‘‘Process Drivers of
Measures of Marketing Constructs.’’ Journal of Marketing e-Service Quality: Analysis of Data From an Online
Research 16(1):64–73. Rating Site.’’ Journal of Operations Management 25(5):
Cohen, J., and Cohen, P. 1983. Applied Multiple Regres- 962–84.
sion ⁄ Correlation Analysis for the Behavioral Sciences. 2nd Hennig-Thurau, T., and Klee, A. 1997. ‘‘The Impact of Cus-
ed. Hillsdale, NJ: Lawrence Erlbaum Associates. tomer Satisfaction and Relationship Quality on Customer
Collier, J., and Bienstock, C. 2006. ‘‘Measuring Service Retention: A Critical Reassessment and Model Develop-
Quality in Online Retailing.’’ Journal of Service Research ment.’’ Psychology and Marketing 14(8):737–64.
8(3):260–75. Howlett, G. 2008. ‘‘Internet Retailer Top 500 Released.’’
Daugherty, P.J., Stank, T., and Ellinger, A. 1998. ‘‘Leverag- Marketing Pilgrim. http://www.marketingpilgrim.com/
ing Logistics ⁄ Distribution Capabilities: The Effect of 2008/06/internet-retailer-top-500-guide.html.
Logistics Service on Market Share.’’ Journal of Business Hsu, C.-I., and Li, H.-C. 2006. ‘‘Optimal Delivery Service
Logistics 19(2):35–52. Strategy for Internet Shopping With Time-Dependent
Davis-Sramek, B., Mentzer, J.T., and Stank, T.P. 2008. Consumer Demand.’’ Transportation Research, Part E
‘‘Creating Consumer Durable Retailer Customer Loyalty 42(6):473–97.
Through Order Fulfillment Service Operations.’’ Journal Hu, L.-T., and Bentler, P. 1995. ‘‘Evaluating Model Fit.’’ In
of Operations Management 26(6):781–97. Structural Equation Modeling: Concepts, Issues, and Appli-
De Koster, R. 2003. ‘‘Distribution Strategies for Online cations, edited by Hoyle, R.H., 76–99. Thousand Oaks,
Retailers.’’ IEEE Transactions on Engineering Manage- CA: Sage Publications, Inc.
ment 50(4):448–57. Innis, D.E., and La Londe, B.J. 1994. ‘‘Customer Service:
Delfmann, W., Albers, S., and Gehring, M. 2002. ‘‘The The Key to Customer Satisfaction, Customer Loyalty,
Impact of Electronic Commerce on Logistics Service Pro- and Market Share.’’ Journal of Business Logistics 15(1):1–
viders.’’ International Journal of Physical Distribution and 27.
Logistics Management 32(2):203–22. IR 500. 2008. ‘‘Editor’s Introduction,’’ Internet Retailer Top
Dillon, W.R., and Goldstein, M. 1984. Multivariate Analysis: 500 Guide. Chicago, IL: Vertical Web Media.
Methods and Applications. Hoboken, NJ: John Wiley. Jiang, P., and Rosenbloom, B. 2004. ‘‘Customer Intention to
Esper, T., Jensen, T., Turnipseed, F., and Burton, S. 2003. Return Online: Price Perception, Attribute-level Perfor-
‘‘The Last Mile: An Examination of Effects of Online mance, and Satisfaction Unfolding Over Time.’’ European
Retail Delivery Strategies on Consumers.’’ Journal of Journal of Marketing 39(1&2):150–74.
Business Logistics 24(2):177–203.
178 S. Rao et al.

Jones, T.O., and Sasser, E.W., Jr. 1995. ‘‘Why Satisfied Cus- Nunnally, J.C. 1978. Psychometric Theory. 2nd ed. New
tomers Defect.’’ Harvard Business Review 73(Novem- York: McGraw-Hill.
ber ⁄ December):88–99. Pan, X., Ratchford, B., and Shankar, V. 2002. ‘‘Can Price
Kim, M., Kim, J.-H., and Lennon, S.J. 2006. ‘‘Online Ser- Dispersion in Online Markets be Explained by Differences
vice Attributes Available on Apparel Retail Web Sites: in E-Tailer Service Quality?’’ Journal of the Academy of
An E-S-Qual Approach.’’ Managing Service Quality Marketing Science 30(4):433–45.
16(1):51–77. Parasuraman, A., Zeithaml, V., and Berry, L. 1985. ‘‘A Con-
Kirmani, A., and Rao, A. 2000. ‘‘No Pain, No Gain: A Crit- ceptual Model of Service Quality and Its Implications for
ical Review of the Literature on Signaling Unobservable Future Research.’’ Journal of Marketing 49(2):41–50.
Product Quality.’’ Journal of Marketing 64(2):66–79. Parasuraman, A., Zeithaml, V.A., and Malhotra, A. 2005.
LaLonde, B.J., and Zinszer, P.H. 1976. Customer Service: ‘‘E-S-Qual: A Multiple-Item Scale for Assessing Elec-
Meanings and Measurement. Chicago, IL: National Coun- tronic Service Quality.’’ Journal of Service Research
cil of Physical Distribution Management. 7(3):213–33.
Lambert, D.M., and Stock, J.R. 1993. Strategic Logistics Paulssen, M., and Birk, M.M. 2007. ‘‘Satisfaction and
Management. 3rd ed. Homewood, IL: Richard D. Irwin. Repurchase Behavior in a Business-to-business Setting:
Lancioni, R., Smith, M., and Oliva, T. 2000. ‘‘The Role of Investigating the Moderating Effect of Manufacturer,
the Internet in Supply Chain Management.’’ Industrial Company and Demographic Characteristics.’’ Industrial
Marketing Management 29(1):45–56. Marketing Management 36(7):983–97.
Lee, H.L., and Whang, S. 2001. ‘‘Winning the Last Mile of Pyke, D.F., Johnson, M.E., and Desmond, P. 2001. ‘‘E-ful-
e-Commerce.’’ Sloan Management Review 42(4):54–62. fillment: It’s Harder Than It Looks.’’ Supply Chain Man-
Lewis, M., Singh, V., and Scott, F. 2006. ‘‘An Empirical agement Review 5(1):26–32.
Study of the Impact of Nonlinear Shipping and Handling Rabinovich, E. 2004. ‘‘Internet Retailing Intermediation: A
Fees on Purchase Incidence and Expenditure Decisions.’’ Multi-Level Analysis of Inventory Liquidity and Fulfill-
Marketing Science 25(1):51–64. ment Guarantees.’’ Journal of Business Logistics
Maltz, A., Rabinovich, E., and Sinha, R. 2004. ‘‘Logis- 25(2):139–69.
tics—The Key to e-Tail Success.’’ Supply Chain Manage- Rabinovich, E., and Bailey, J. 2004. ‘‘Physical Distribution
ment Review 8(3):56–63. Service Quality in Internet Retailing: Service Pricing,
Mentzer, J.T., Flint, D.J., and Hult, G.T.M. 2001. ‘‘Logistics Transaction Attributes, and Firm Attributes.’’ Journal of
Service Quality as a Segment-Customized Process.’’ Operations Management 21(6):651–72.
Journal of Marketing 65(3):82–104. Rabinovich, E., and Evers, P.T. 2003. ‘‘Product Fulfillment
Mentzer, J.T., Flint, D.J., and Kent, J.L. 1999. ‘‘Developing in Supply Chains Supporting Internet-Retailing Opera-
a Logistics Service Quality Scale.’’ Journal of Business tions.’’ Journal of Business Logistics 24(2):205–36.
Logistics 20(1):9–32. Rabinovich, E., Knemeyer, M., and Mayer, C. 2007. ‘‘Why
Mentzer, J.T., Gomes, R., and Krapfel, R. 1989. ‘‘Physical do Internet Commerce Firms Incorporate Logistics Ser-
Distribution Service: A Fundamental Marketing Con- vice Providers in Their Distribution Channels? The Role
cept?’’ Journal of the Academy of Marketing Science of Transaction Costs and Network Strength.’’ Journal of
17(1):53–62. Operations Management 25(3):661–81.
Mittal, V., and Kamakura, W. 2001. ‘‘Satisfaction, Repur- Rabinovich, E., Rungtusanatham, M., and Laseter, T.M.
chase Intent, and Repurchase Behavior: Investigating the 2008. ‘‘Physical Distribution Service Performance and In-
Moderating Effect of Customer Characteristics.’’ Journal ternet Retailer Margins: The Drop-Shipping Context.’’
of Marketing Research 38(1):131–42. Journal of Operations Management 26(6):767–80.
Mollenkopf, D., Rabinovich, E., Laseter, T., and Boyer, K. Rao, A.R., Bergen, M.E., and Davis, S. 2000. ‘‘How to
2007. ‘‘Managing Internet Product Returns—A Focus on Fight a Price War.’’ Harvard Business Review 78(3):107–
Effective Service Operations.’’ Decision Science 38(2):215– 16.
50. Rao, S., Goldsby, T.J., and Iyengar, D. 2009. ‘‘The Market-
Monroe, K.B. 1990. Pricing: Making Profitable Decisions. ing and Logistics Efficacy of Online Sales Channels.’’
New York: McGraw-Hill ⁄ Irwin. International Journal of Physical Distribution and Logistics
Montoya-Weiss, M.M., Voss, G.B., and Grewal, D. 2003. Management 39(2):106–30.
‘‘Determinants of Online Channel Use and Overall Satis- Reibstein, D.J. 2002. ‘‘What Attracts Customers to Online
faction With a Relational, Multichannel Service Pro- Stores and What Keeps Them Coming Back?’’ Journal of
vider.’’ Journal of the Academy of Marketing Science the Academy of Marketing Science 30(4):465–73.
31(4):448–58. Reichheld, F.F. 1996. The Loyalty Effect: The Hidden Force
Nelson, E., Rust, R.T., Zahorik, A.J., Rose, R., Batalden, Behind Growth, Profits, and Lasting Value. Boston, MA:
P., and Siemanski, B.A. 1992. ‘‘Do Patient Perceptions of Harvard Business School Press.
Quality Relate to Hospital Financial Performance?’’ Jour- Retail Week. 2009. ‘‘Global Online Retail—The World at
nal of Health Care Marketing 12(4):42–49. Your Fingertips.’’ http://www.retail-week.com/multichannel/
Newton, C.J. 2001. ‘‘Home Delivery Can Make or Break online-retail/global-online-retail-the-world-at-your-finger
B2C.’’ Supply Chain Management Review 5(1):21–24. tips/5002140.article.
Impact of e-LSQ on Customer’s Satisfaction 179

Reuters. 2009. ‘‘Bidz.com Ranked Number 2 in Internet Yang, Z., and Jun, M. 2002. ‘‘Consumer Perception of
Retailer’s Online Jewelry Category.’’ http://www.reuters. E-Service Quality From Internet Purchaser and Non-Pur-
com/article/pressRelease/idUS177848+10-Jun-2009+BW chaser Perspectives.’’ Journal of Business Strategies
20090610. 19(1):19–41.
Ricker, F.R., and Kalakota, R. 1999. ‘‘Order Fulfillment:
The Hidden Key to e-Commerce Success.’’ Supply Chain SHORT BIOGRAPHIES
Management Review 11(3):60–70.
Rohm, A., and Swaminathan, V. 2004. ‘‘A Typology of Shashank Rao (PhD University of Kentucky) is an Assis-
Online Shoppers Based on Shopping Motivations.’’ Jour- tant Professor of Logistics, Operations, and Supply Chain
nal of Business Research 57(7):748–57. Management at the H. Wayne Huizenga School of Business,
Rust, R.T., and Zahorik, A.J. 1993. ‘‘Customer Satisfaction, Nova Southeastern University. He holds a B.S. (Environ-
Customer Retention, and Market Share.’’ Journal of mental Science), and a M.B.A. (Marketing) from India. His
Retailing 69(2):193–215. research interests lie in the areas of supply chain manage-
Shop.org. 2009. ‘‘The 2009 State of Online Retailing ment, online retailing, and environmentally responsible logis-
Marketing Report.’’ http://www.shop.org/web/guest/resear tics. He has published in The International Journal of
charchive. Logistics Management and the International Journal of Physi-
Srinivasan, R., and Moorman, C. 2005. ‘‘Strategic Firm cal Distribution and Logistics Management.
Commitments and Rewards for CRM in Online Retail-
ing.’’ Journal of Marketing 69(4):193–200. Thomas J. Goldsby (PhD Michigan State University) is an
Srinivasan, S.S., Anderson, R., and Ponnavolu, K. 2002. Associate Professor of Supply Chain Management at the
‘‘Consumer Loyalty in E-commerce: An Exploration of University of Kentucky. Dr. Goldsby’s research interests
Its Antecedents and Consequences.’’ Journal of Retailing focus on logistics customer service, supply chain integration,
78(1):41–50. and the theory and practice of lean and agile supply chain
Stank, T.P., Goldsby, T.J., and Vickery, S.K. 1999. ‘‘Effect strategies. He has published several articles in academic and
of Service Supplier Performance on Satisfaction and Loy- professional journals and serves as a frequent speaker at aca-
alty of Store Managers in the Fast Food Industry.’’ Jour- demic conferences, executive education seminars, and profes-
nal of Operations Management 17(4):429–47. sional meetings. He is coauthor of Lean Six Sigma Logistics:
Stank, T.P., Goldsby, T.J., Vickery, S.K., and Savitskie, K. Strategic Development to Operational Success (J. Ross Pub-
2003. ‘‘Logistics Service Performance: Estimating Its lishing, 2005) and a research associate of the Global Supply
Effect on Market Share.’’ Journal of Business Logistics Chain Forum at The Ohio State University.
24(1):27–55.
Thirumalai, S., and Sinha, K.K. 2005. ‘‘Customer Satisfaction Stanley E. Griffis (PhD The Ohio State University) is an
With Order Fulfillment in Retail Supply Chains: Implica- Associate Professor of Logistics Management in the Broad
tions of Product Type in Electronic B2C Transactions.’’ College of Business at Michigan State University. His cur-
Journal of Operations Management 23(3–4):291–303. rent research interests include social networks in supply
Thomas, J.S. 2001. ‘‘A Methodology for Linking Customer chains, the role of logistics in customer satisfaction and loy-
Acquisition to Customer Retention.’’ Journal of Market- alty, and the application of meta-heuristic techniques to
ing Research 38(2):262–68. logistics transportation networks. He has published in the
Varki, S., and Colgate, M. 2001. ‘‘The Role of Price Percep- Journal of Business Logistics, Transportation Journal, and the
tions in an Integrated Model of Behavioral Intentions.’’ Journal of Transportation Management.
Journal of Service Research 3(3):232–40.
Vickery, S., Dröge, C., Stank, T., Goldsby, T.J., and Mark- Deepak Iyengar (PhD University of Maryland) is Assistant
land, R. 2004. ‘‘The Performance Implications of Media Professor of Supply Chain Management at the University of
Richness in a B2B Service Environment: Direct Versus Kentucky. Dr. Iyengar’s research interests lie in the areas of
Indirect Effects.’’ Management Science 50(8):1106–19. logistics, distribution channels, operations management, and
Wolfinbarger, M., and Gilly, M. 2001. ‘‘Shopping Online for international trade. He has extensive research and consulting
Freedom Control and Fun.’’ California Management experience in lean and six sigma methodologies.
Review 43(2):34–55.