The development of service quality dimensions for internet serviceproviders: Retaining customers of different usage patterns

Paramaporn Thaichon n, Antonio Lobo 1, Catherine Prentice 2, Thu Nguyen Quach 3

Faculty of Business & Enterprise, Swinburne University of Technology, Melbourne, Victoria 3122, Australia

a r t i c l e i n f o

Article history:Received 1 April 2014Received in revised form18 June 2014Accepted 18 June 2014Available online 19 July 2014

Keywords:Customer commitmentValueTrustService qualityInternet usageInternet service provider (ISP)

a b s t r a c t

This study examines the relationships among relevant service quality dimensions of Internet serviceproviders (ISP) and their customers’ perceived value, trust and commitment. Data was collected fromresidential Internet users in Thailand. The final usable sample size was 1507. The analyses includesegmenting ISPs’ customers on the basis of their usage pattern and evaluating their perceptions ofInternet service quality dimensions. In addition, several alternatives models were compared usingstructural equation modelling to confirm the mediation effects. An ISP’s service quality is influenced bythe following four dimensions (a) network quality, (b) customer service and technical support,(c) information quality and (d) security and privacy. The findings reveal that while all dimensions havepositive effects on trust, only network quality, information support and privacy influence customer valuesignificantly and information support is the only dimension which is directly related to commitment.Additionally, the effects of customer service and information support on value vary across customers ofdifferent Internet usage patterns. The contribution of the present paper stems from the simultaneousmodelling of a range of mediation effects which can better help explain the impact of service qualitydimensions on customers’ cognitive and affective evaluations in high-tech service settings.

& 2014 Elsevier Ltd. All rights reserved.

1. Introduction

Service quality is an important differentiator in a competitivebusiness environment, and a driver of service-based businesses (Zhaoand Benedetto, 2013). By enhancing service quality, businesses caninfluence customers’ value (Lai et al., 2009), trust (Sabiote andRoman, 2009), and commitment (Fullerton, 2005). These are impor-tant for business success and long term customer loyalty (Prentice,2013). However, very few studies have assessed how different aspectsof Internet service providers’ (ISP) service quality would influencetheir customers’ value, trust, and commitment (Thaichon et al., 2014;Vlachos and Vrechopoulos, 2008). ISPs may benefit from obtainingaccurate information regarding their customers’ assessments of theirbrand’s delivered service quality; such information may enableservice brand managers to formulate appropriate marketing strate-gies in order to achieve competitive advantage and long termprofitability. This paper attempts to fill this important research gap

by investigating the effects of ISPs’ service quality on their customers’value, trust, and commitment in the high-tech Internet services.

Service quality measures how well the service deliveredmatches customer expectations (Zhao and Benedetto, 2013). Inaddition to SERVQUAL, E-S-QUAL has been developed byParasuraman et al. (2005) as an attempt to capture the measure-ment of service quality in the new information age. However,owing to the very special nature of the services offered by ISPs,their service quality cannot be effectively measured by SERVQUALor E-S-QUAL (He and Li, 2010; Thaichon et al., 2014). SERVQUALand E-S-QUAL focus on service providers who operate via theInternet platform (Vlachos and Vrechopoulos, 2008) but not thosewho actually provide the Internet connection and platform activ-ities. Numerous studies have been done in the telecommunica-tions industry, especially in the mobile telephony market (He andLi, 2010). However, several basic differences exist between Internetservices and other telecommunications services. For example,mobile phone service quality includes value-added services (e.g.SMS, MMS, WAP, GPRS) or mobile devices (Santouridis andTrivellas, 2010), which are not applicable in the case of ISPs.

In addition, as the nature of home Internet services is Internetrelated, privacy and security are more prominent when assessingan ISP’s service quality as compared to assessing service quality ofother telecommunications services such as mobile and televisionservices. An ISP’s server contains account information of many

Contents lists available at ScienceDirect

journal homepage: www.elsevier.com/locate/jretconser

Journal of Retailing and Consumer Services

http://dx.doi.org/10.1016/j.jretconser.2014.06.0060969-6989/& 2014 Elsevier Ltd. All rights reserved.

n Corresponding author. Tel.: þ61 3 92145266; fax: þ61 3 9214 5293.E-mail addresses: pthaichon@swin.edu.au (P. Thaichon),

alobo@swin.edu.au (A. Lobo), cathyjournalarticles@gmail.com (C. Prentice),squach@swin.edu.au (T.N. Quach).

1 Tel.: þ61 3 92148535.2 Tel.: þ61 406 627622.3 Tel.: þ61 3 92145266.

Journal of Retailing and Consumer Services 21 (2014) 1047–1058

users, which might put customers’ personal data at risk ifunauthorised access is granted (Rowe et al., 2011). Moreover,being more active than other telecommunications service provi-ders regarding online activities, ISPs observe and monitor trafficflowing through their networks; hence, they are able to detectsuspicious traffic spikes, and can either stop malicious traffic orprovide timely warnings to their customers (Rowe et al., 2011). Astudy in 2004 reports that 66 per cent of consumers would switchto other ISPs who offered more secured Internet service(Streamshield, 2004). Therefore, it can be concluded that custo-mers perceive ISPs protection from privacy invasion and cyber-crime as necessary and important.

A recent consumer study demonstrates that the more regularlycustomers access the Internet, the more they need and appreciateonline help (Oracle, 2012). Additionally, not every customer in thetelecommunications services, other than Internet services, hasaccess to online information support, especially in developingcountries. In other words, customers of other telecommunicationsservices might not perceive information support as important ascustomers of an ISP do. For example, in the mobile telephonyservices context, there are more than 84 million mobile subscri-bers in Thailand and 134 million subscribers in Vietnam (CIA,2013). However, only 31.9 per cent of the Thai population (NBTC,2013) and 40 per cent of Vietnamese users (Freedom, 2012) useonline services via their mobile phones. On the other hand, thenumber of residential Internet users account for 26 per cent(approximately 20 million users) and 36 per cent (approximately30 million users) of the population in Thailand and Vietnamrespectively (WorldBank, 2014). As such, in contrast to ISP userswho generally take advantage of online information support, themajority of mobile phone service customers would most likely ignorethe online information support. Hence, it can be assumed thatcustomers of an ISP are more likely to access the company’s websiteto look for information support as compared to customers of othertelecommunication services. On the basis of above discussion, thisstudy aims to provide a more holistic picture on the unique dimen-sions of an ISP’s service quality.

Several researchers (Ringle et al., 2013) suggest that studying asingle homogenous population in path models is insufficient tounderstand the path relationships as customers’ characteristic and

the nature of their demand for services differ (Mazzoni et al.,2007; Ringle et al., 2013). Segmentation is the process of sub-dividing a heterogeneous market into homogeneous groups ofcustomers who have similar characteristics or who respond tomarketing activities in the same way (Ko et al., 2012). Never-theless, there is hardly any evidence of how effective segmentationis operationalised for an ISP’s customers. This study segmentscustomers of ISPs based on their usage pattern, which is one of themost logical basis of segmentation in similar types of services(Mazzoni et al., 2007; Wedel and Kamakura, 2003). Segmentingmarkets by consumption patterns is a relatively intuitive steptoward comprehending customers (Weinstein, 2002). By categor-ising customers into usage groups, service providers can createsuitable marketing strategies for each segment. Furthermore, seg-mentation by usage is helpful in assessing the profitability ofcustomer retention, as well as developing retention strategies(McDougall, 2001). In a similar vein, Weinstein (2002) concludesthat usage analysis can support customer retention accomplishments.

Based on the foregoing discussion, the objectives of thisresearch study are threefold: first, to identify the relevant servicequality dimensions for an ISP; second, to evaluate their effects onan ISP’s customer’s value, trust, and commitment; and third, toinvestigate service perceptions of different market segments. Inorder to achieve the research objectives stated above, a model isproposed as depicted in Fig. 1. Section 2 reviews the literature anddevelops hypotheses. Next, data collection and analysis using thestructural equation modelling technique of comparing alternativemediation models are reported including the testing of hypotheses.The paper concludes with a discussion of the results, implications ofthe research as well as limitations and future research direction.

2. Literature review and development of hypotheses

Customer commitment is influenced by customers’ value (Tai,2011), trust (Wu et al., 2010) and service quality (Thaichon et al.,2012). Customer commitment has been defined as a customer’sconviction to maintain a relationship that might produce func-tional and emotional benefits (Tuškej et al., 2013). Lin and Wu(2011) consider customer commitment as a customer’s persistent

Fig. 1. The proposed conceptual model.

P. Thaichon et al. / Journal of Retailing and Consumer Services 21 (2014) 1047–10581048

wish and attempt to retain a relationship with a service provider.Trust is known to be a foundation of a long-term relationship,possibly building an advanced exchange relationship betweenbuyers and sellers (Hong and Cho, 2011). Customer trust refersto the customers’ perceptions of attributes of service providers,including the ability, integrity, and benevolence of the providers(Deng et al., 2010). Additionally, customer trust relates to theperception of customers relating to the ability of a brand to fulfilits promise while expertise refers to a brand’s capability ofrealising its promises (Ou et al., 2011).

Customer value has often been described as an exchangebetween what customers receive and what customers have togive to purchase a service (Lai et al., 2009; Shirin and Puth, 2011;Tam, 2012). Sweeney and Soutar (2001) introduce four types ofvalues based on the work by Sheth et al. (1991). These arefunctional value involving performance quality, cost/sacrifice valueinvolving price/value for money, emotional value referring tofeelings or affective states generated by a service, and social valuerelating to an enhancement of social self-concept. This study dealsonly with functional value (performance quality) and cost/sacrificevalue (price/value for money) which are directly related to an ISP’sservice quality and are considered to be important with respect tocustomers’ usage intentions and behaviour in the telecommunica-tions sector (Kim, 2012; Vlachos and Vrechopoulos, 2008).

Previous research reveals that overall service quality in thetelecommunications industry is associated with customers’ per-ceptions of a stable and strong network quality (Lai et al., 2009),ready-to-serve customer support team (Aydin and Özer, 2005),informative quality (Thaichon et al., 2014) and a high level ofsecurity and privacy that is trusted by customers (Roca et al.,2009). This study intends to identify this unique service qualitymeasurement model using the abbreviated acronym “NCISQuality Model” and each of the dimensions of this model arenow discussed.

In the telecommunications market, network quality is one ofthe most important drivers of overall service quality (Vlachos andVrechopoulos, 2008). In the Internet service industry, networkquality includes the quality and strength of the network signal,number of errors, downloading and uploading speed (Thaichonet al., 2012). Breaks in Internet connectivity can lead to poorperceptions of network quality in the customer’s perspective. Inthis respect, timely recovery of network connectivity is essential.Factors contributing to benefits or sacrifices in the relationshipbetween customers and service providers lead to different percep-tions of customer value (Wang and Lo, 2002). He and Li (2010)suggest that network quality is a positive driver of customer valuein the mobile phone services sector in Taiwan. In addition, aservice provider whose core performance meets or exceeds theexpectations of customers is likely to develop more trustingrelationships with its customers (Eisingerich and Bell, 2008).Hence, it can be posited that network quality is positively relatedto customer trust. On the other hand, attachment to the companycan be built through cognitive evaluation of service performance(Fullerton, 2005). Although scant evidence has been found in thedirect relationship between network quality and commitment inthe Internet services market, Fullerton (2005) concludes thatservice quality is a direct antecedent of customer commitmentwhich is also supported by Thaichon, Lobo and Mitsis (2014).Consistent with the foregoing discussion, the following hypothesesare postulated:

H1a. Network quality is positively related to customer value.

H1b. Network quality is positively related to customer trust.

H1c. Network quality is positively related to customer commitment.

Telecommunication companies can offer additional serviceattributes (Wang and Wu, 2012) to enhance the quality of theirservices (Tam, 2012) such as superior customer service and after-sales support. Customer service and technical support providetouch points between the company and their customers, and is acritical dimension of service quality in the Turkish telecommuni-cations industry (Aydin and Özer, 2005). Customers pursue cordialrelationships with the company which gives due importance totheir thoughts, emotions and concerns (Eisingerich and Bell,2008). These authors suggest that considerate, caring and respon-sive customer service can stimulate confidence in customers.Previous research reveals that customer service influences custo-mer commitment in UK retail banks (Malhotra et al., 2013). Also,customer service has an impact on customer trust in financialservices and on value perceptions of customers in the commu-nication services in the United States (Blocker, 2011). Hence, thefollowing hypotheses have been developed:

H2a. Customer service and technical support are positively relatedto customer value.

H2b. Customer service and technical support are positively relatedto customer trust.

H2c. Customer service and technical support are positively relatedto customer commitment.

The combination of information and communication technol-ogy generates massive impact on society by connecting businessesand their customers via the Internet (Asmussen et al., 2013). Infact, many businesses rely on the Internet as a main communica-tion channel (Lee et al., 2012). Information quality refers to theaccuracy, completeness, presentation and format of the informa-tion given by service providers (Elliot et al., 2013), and has beenconsidered as an important component of service quality (Yanget al., 2005). Information quality positively influences customertrust in the Chinese online travel agency business (Elliot et al.,2013). In addition, Kim and Niehm (2009) reveal that perceivedinformation quality is significantly related to perceived value inthe apparel retailing sector. Customers can evaluate value directlyor indirectly via information provided by the websites (Grewal etal., 2003).Moreover, information quality reflects quality of theservice (Kim and Niehm, 2009), and service quality has aninfluence on perceived value (Parasuraman and Grewal, 2000;Tam, 2004), as well as, customer commitment (Fullerton, 2005;Morgan and Hunt, 1994). Hence, it is reasonable to assume that thequality of information is related to perceived value and customercommitment. The foregoing discussion supports the followinghypotheses:

H3a. Information quality is positively related to customer value.

H3b. Information quality is positively related to customer trust.

H3c. Information quality is positively related to customer commitment.

Security and privacy are associated with customers’ feelings ofprotection and safety during their transactions and usage (Vlachosand Vrechopoulos, 2008). Security of payments and privacy ofpersonal information are positively related to service quality ine-commerce (Ha and Stoel, 2012). A trustworthy e-commerceservice provider is often associated with fewer privacy concerns(Cases et al., 2010). Wu et al. (2010) report that website privacypolicies enhance trust among virtual community members. More-over, customers tend to believe that it is safe to purchase servicesfrom providers who possess good reputation with regards to theirsecurity practice (Roca et al., 2009). In other words, a transparentand reliable security and privacy policy is likely to generatefavourable perceptions of overall ISP’s service quality. Hence,

P. Thaichon et al. / Journal of Retailing and Consumer Services 21 (2014) 1047–1058 1049

security and privacy are identified as dimensions of service quality(Ladhari, 2010; White and Nteli, 2004). By utilising the existingevidence on the relationship between service quality, and value(Tam, 2012; Wang and Wu, 2012) and commitment (Fullerton,2005), this study proposes that security and privacy are alsorelated to value and commitment. Based on the foregoing discus-sion the following hypotheses have been developed:

H4a. Security and privacy are positively related to customer value.

H4b. Security and privacy are positively related to customer trust.

H4c. Security and privacy are positively related to customercommitment.

The proposed conceptual model showing the relationships ofthe various constructs discussed so far is depicted in Fig. 1.

The second set of constructs in Fig. 1 explores the underlyingrelationships between customers’ evaluations of value, trust andcommitment. Previous research has shown that customer com-mitment is positively related to repeat purchase, and propensityto stay in the relationship (Fullerton, 2005). Relationship com-mitment is driven by functional and emotional benefits (Tuškejet al., 2013). In addition, consumer buying behaviour theorysuggests that customers consider both the loss and gain of acertain decision (Hauser and Wernerfelt, 1990; Ratchford, 1982).Furthermore, customer value involves perceived trade-offbetween benefits and sacrifices in relationships (Blocker, 2011).Therefore, the value obtained by remaining with the companymay enhance motivations for customer commitment (Lacey,2007). Musa et al. (2005) support this view by asserting thatthe perception of value is derived from the direct sales consump-tion experience and has a positive effect on relational commit-ment. Tai (2011) confirms that relationship commitment ispositively influenced by the functional and relational value inthe information sharing services.

Previous research postulates that there is a positive relation-ship between perceived value and customer trust, as perceivedvalue can enhance customers’ perceptions of their service provi-ders’ ability, reliability and benevolence, thereby increasing theirconfidence in purchasing the service (Chen and Chang, 2012). Inline with this thinking, Chen and Chang (2012) report thatperceived value has a positive effect on trust in a green marketingcontext as well as in mobile telecommunication services(Karjaluoto et al., 2012). Hence, the literature review informs thefollowing hypotheses:

H5. Perceived value is positively related to customers’ commitment.

H6. Perceived value is positively related to customers’ trust.

Customer trust is a primary element of long-term relationshipmarketing and an essential antecedent of purchase behavior(Benedicktus, 2011). Customer trust can be evaluated by exploringhow customers feel about their service provider in terms of thecompany’s honesty, responsibility and professional manners, andif the customers think that the firm understands and cares aboutthem (Chiou, 2004). Trust together with commitment is importantin establishing a long-term business relationship (Morgan andHunt, 1994). Recent research demonstrates that the more acustomer trusts the service provider, the more affectively com-mitted he or she becomes (Perry et al., 2004). In fact, trust shows apositive and significant effect on the customers’ commitment invirtual community services (Wu et al., 2010), and in the Greekbusiness-to-business services (Perry et al., 2004). Rutherford(2012) states that customer commitment increases as trust inthe salesperson increases. This study considers trust as a precursorof commitment, which involves vulnerability and sacrifice. Com-mitment will most likely develop in relationships in which trust is

present (Wu et al., 2010). Based on the above discussions, thefollowing has been hypothesised:

H7. Customers’ trust is positively related to their commitment.

Different customers have distinctive needs and require tailoredapproaches (Mazzoni et al., 2007; Ringle et al., 2013). Based ontheir usage pattern, ISP customers are generally segmented to beheavy, medium and light users. On average, an Internet userspends from 9 h to as much as 20 h weekly (ACMA, 2012). Heavyusers are those who spend more than 29 h on the Internet everyweek, whilst light users are those who use the Internet for lessthan 9 h per week (Assael, 2005). A study by Electronic Transac-tions Development Agency (ETDA, 2013) reveals that in generalThai Internet users who spend less than 11 h per week onlineaccount for 35.7 per cent; those who spend between 11 and 20 hper week online make up 25.8 per cent; 10.7 per cent of Thai usersspend 21–41 h on the Internet weekly; and 27.8 per cent spendmore than 41 h weekly. This study adapting usage segmentationfrom previous research, categorises three main groups of Internetusers which are: light (i.e. less than 9 h per week), medium (i.e. 9–29 h per week), and heavy users (i.e. more than 29 h per week).Furthermore, it is most likely that each specific service qualitydimension distinctively impacts customers’ value, trust and commit-ment, depending on differently segmented groups of customers(Ringle et al., 2013). Hence, the following has been hypothesised:

H8. The relationships between service quality dimensions andvalue, trust and commitment differ across different segments.

3. Method

3.1. The study sample

To test the hypotheses, an online survey was designed andconducted in all regions of Thailand. Thailand is ranked third inSouth East Asia by way of residential Internet usage with anestimated 17,483,000 Internet users in 2009 (CIA, 2013) and over24 million Internet users in 2012 (IWS, 2013). The number in 2012represented over one-third of the Thai population. The competi-tion among residential Internet service providers in Thailand isintense. Currently there are three majors ISPs and sixteen smallerones across the country (Thaichon and Quach, 2013). In this highlycompetitive market, the churn rate of Internet users was approxi-mately 12 per cent in 2009 (Thaichon et al., 2012). This scenario,therefore, poses huge challenges to ISPs especially in the area ofcustomers’ repurchase intention.

3.2. Measures

Vlachos and Vrechopoulos (2008) connection quality scale wasused to measure network quality. This scale examines connectionquality for mobile phone services which is very similar to thenature of an ISP’s network quality. The customer service scale wassourced from Wolfinbarge and Gilly’s (2003) scale whichaddresses both customer service and technical support in an ISP’sofferings. Four different scales relating to information quality fromChae et al. (2002), Lin (2007), Kim and Niehm (2009) and Vlachosand Vrechopoulos (2008) were considered. After a thoughtfulanalysis, the Kim and Niehm’s (2009) information quality scalewas selected as this scale has higher factor loadings (.80 � .83),and Cronbach’s alpha (α¼ .96). Vlachos and Vrechopoulos (2008)privacy scale was selected. The scale’s measurement itemsinvestigate whether customers feel safe parting with informationduring transactions, and also seek their opinions on securityfeatures of an ISP.

P. Thaichon et al. / Journal of Retailing and Consumer Services 21 (2014) 1047–10581050

To assess customer’s trust, Aydin and Özer (2005) scale wasselected. Items in this scale examine customers’ perceptions of acompany’s honesty, responsibility and professional dealings. Cus-tomer commitment was operationalised using Eisingerich andRubera’s (2010) scale which has relatively strong factor loadings(.716–.852) and reasonable Cronbach’s alpha (α¼ .865). Eisingerichand Rubera’s (2010) scale examines into customers’ feelings andsense of belonging to their service provider. Finally, Kim andNiehm’s (2009) perceived values scale was selected. This scaleinvestigates customers’ evaluations toward the service offerings interms of value for money.

Respondents were required to rate their perceptions forevery item using a Likert scale which was anchored at 1 forstrongly disagree and 5 for strongly agree. These statementsoriginally in English were translated into Thai language by aprofessional translator. The translated versions were then cross-checked by three other bilingual researchers to ensure face andcontent validity. The items of the survey are depicted in Table 1.

3.3. Data collection

Data was collected from residential Internet users in Thai-land. A selective customer database of a well-established majorISP in Thailand was utilised as the sampling frame. This databaseincluded customers throughout Thailand who were not lockedinto any fixed term contract with the ISP. It was a requirementthat the participants were over 18 years of age and they shouldhave used home Internet services. The web link of the onlinesurvey was relayed by the chosen ISPs to households in thesampling frame. The university’s Opinion platform was kept livefor a period of three months. The final usable sample size was1507. Approximately 55.1 per cent of the respondents weremale, and 44.9 per cent were female. In the age profile, 22 percent of the respondents were between 18 and 28 years, 38.8 percent between 29 and 39 years, 24.5 per cent between 39 and 49years and 14.7 per cent were 50 years or older. In terms ofinternet usage, 27.7 per cent of the respondents were lightusers, 23 per cent were medium users and 49.3 per cent wereheavy users.

4. Data analysis

4.1. Factor analysis and validity testing

Table 1 demonstrates the measurement items of each con-struct. Confirmatory factor analysis (CFA) using AMOS Version 20(Analysis of Moment Structures) confirmed that the first-orderfour-factor model of ISP’s service quality (λ2(50)¼233.11,po .0005, CFI¼ .98, TLI¼ .98, SRMR¼ .027, and RMSEA¼ .05) pro-duced a better fit than the second order service quality model(λ2(48)¼238.77, po .0005, CFI¼ .98, TLI¼ .97, SRMR¼ .03, andRMSEA¼ .05). Hence the first-order model was used for thesubsequent analysis.

Table 1 illustrates the 7 constructs and their Cronbach’s alpha,construct reliability and average variance extracted (AVE). It can beseen that factor loadings of all measurement items well exceededthe recommended .4 cut-off (Nunnally, 1978) and were statisticallysignificant. The AVE for each factor was greater than .50, showingsufficient convergent validity (Fornell and Larcker, 1981). Discri-minant validity was examined by calculating squared correlationscoefficients between each pair of constructs and comparing themwith the corresponding average variances extracted (AVE) of eachconstruct. As all squared correlations coefficients were belowAVEs, discriminant validity was confirmed (Fornell and Larcker,1981). The composite reliability was also satisfactory. Further-more, the correlations among all variables presented in Table 2ranging from .53 to .89 were below the .90 cut-off (Tabachinickand Fidell, 2001), showing neither redundancy nor violation ofmulti-collinearity.

4.2. Structural modelling

Structuring alternate and competitive models are recom-mended by Cronin et al. (2000) and McKenzie (1998), with a viewto obtaining better explanations for the relationships amongstconstructs. In addition to the Main Research Model (MRM), thisstudy tests four other alternative models with service qualitydimensions acting as initiators in the development of commit-ment. The MRM is the proposed model of this study and isillustrated in Fig. 1. The alternative models (Fig. 2) are Direct

Table 1Instrument items and reliability indices.

Items FL Α CR AVE

NQ I do not experience any Internet disconnection from this ISP .67 .82 .82 .60The Internet downloading and uploading speed meet my expectations .83The Internet speed does not reduce regardless peak or off-peak hours .82

CS Customer service personnel are knowledgeable .89 .89 .90 .74Customer service personnel are willing to respond to my enquiries .89My technical problems are solved promptly .80

IW This ISP provides sufficient information .84 .86 .86 .68This ISP provides up-to-date information .82This ISP provides relevant information .82

SP I feel that my personal information is protected at this ISP .74 .83 .85 .65I feel that my financial information is protected at this ISP .76I feel that the transactions with this ISP are secured .90

TRU I trust this ISP .92 .90 .92 .78I feel that I can rely on this ISP service .93I feel that this ISP will not deceive me in any way .80

COM I feel involved with this ISP .86 .86 .87 .69I am very proud to have this company as my service provider .74I feel attached to this ISP .89

VAL This Internet package is worth my money .96 .96 .96 .89I would consider this Internet package to be a good buy .96I feel that I purchase a good Internet package with a reasonable price .95

Notes: FL¼factor loadings, α¼Cronbach’s alpha, CR¼Construct reliability, AVE¼Average variance extracted, CS¼Customer Service; NQ¼Network Quality; IW¼ InformationQuality; SP¼Security and Privacy; COM¼Commitment; VAL¼Value; TRU¼Trust.

P. Thaichon et al. / Journal of Retailing and Consumer Services 21 (2014) 1047–1058 1051

Effects Model (DEM), Simple Mediation Model (SMM), DoubleMediation Model (DMM) and Real Mediation Model (RMM).

The DEM presents one-way direct effects of the independentconstructs (i.e. service quality dimensions, value, and trust) on thedependent construct (i.e. commitment). The second model, SMM,introduces the effect of service quality dimensions on commit-ment through value as a mediator. The third alternate viewdemonstrated in the DMM proposes that trust mediates the effectof value on commitment, whereas value is a mediator of therelationship between service quality dimensions and commit-ment. Finally, the RMM posits direct effects of service qualitydimensions on value and trust, direct effects of value on trust, anddirect effects of both value and trust on customer commitment. Asthe research model involves several latent constructs and differentlevels of mediating relationships, structural equation modelling(SEM) is the appropriate method of data analysis. The proposedmain research model and other alternatives were tested using SEMwith maximum likelihood estimation. Bias correct bootstrappingwas also conducted to assist the mediation test as mentioned byPreacher and Kelley (2011).

Both the DEM and the SMM resulted in a poor fit (λ2(177)¼3716.30, GFI¼ .83, TLI¼ .84, CFI¼ .87, RMSEA¼ .12, AIC¼3824.30;and λ2(177)¼2721.51, GFI¼ .90, TLI¼ .89, CFI¼ .90, RMSEA¼ .10,AIC¼2829.51 respectively). Although having slightly improved fitindices, the DMM (λ2(176)¼1657.26, GFI¼ .91, TLI¼ .93, CFI¼ .94,RMSEA¼ .08, AIC¼1767.26) was also not satisfactory in represent-ing the influence of service quality dimensions on customercommitment. This conclusion is further confirmed by better fitindices in the RMM as the chi square statistics decreased toλ2(172)¼988.78, and other fit indices were much improved:GFI¼ .94, TLI¼ .93, CFI¼ .97, RMSEA¼ .06, AIC¼1106.78.

The fit indices in the RMM and the main research model (MRM)were very similar. Because they were nested models, a chi squaredifference test was conducted to compare these models. Theresults show that the difference in Chi square test was significantwhen comparing the MRM with RMM (λ2(4)¼15.91; p¼ .00). As aresult, the more complicated model, which is MRM, is preferred. Inother words, the importance of the comprehensive direct andindirect effects of network quality, customer service and technicalsupport, information quality, and privacy and security on customercommitment is established.

For the MRM, although the Chi square statistics was significant(p¼ .000), which can be explained owing to the relatively largesample size (41000), other fit indices (CMIN/DF¼5.79, GFI¼ .94,TLI¼ .96, CFI¼ .97, RMSEA¼ .06, AIC¼1098.87) indicate that thestructural model was a good fit to the data. The model explained76.6 per cent of the variance in trust (R2¼ .77), 55 per cent in value(R2¼ .55) and 73.3 per cent in commitment (R2¼ .73). Whereas allservice quality dimensions directly influenced customer trust, onlythe effects of network quality, information quality, and privacy andsecurity on value were found to be significant. Also, the directimpact on customer commitment was confirmed only for informa-tion quality. The results are depicted in Table 3.

Table 4 shows the standardised total, direct, and indirect effectsof the predictors on their criterions based on the results of biascorrected bootstrapping. As Tables 3 and 4 demonstrate, mediating

Table 2Correlations among all variables.

SP IW CS NQ VAL TRU COM

SP .60a

IW .77a .68a

CS .61a .75a .74a

NQ .58a .69a .60a .60a

VAL .64a .66a .53a .67a .89a

TRU .78a .80a .67a .69a .74a .78a

COM .65a .70a .54a .58a .66a .82a .69a

Notes: CS¼Customer Service; NQ¼Network Quality; IW¼ Information Quality;SP¼Security and Privacy; COM¼Commitment; VAL¼Value; TRU¼Trust.

a All correlations are significant at the .01 level (2-tailed).

Double Mediation Model Real Mediation Model

Direct Effect Models Simple Mediation Model

Fig. 2. Alternative models.

P. Thaichon et al. / Journal of Retailing and Consumer Services 21 (2014) 1047–10581052

conditions set up by Baron and Kenny (1986) were satisfied.Except for customer service and technical support, the relationshipof all the other service quality dimensions with trust wasmediated by value. While network quality and security andprivacy fully manifested their effects on commitment throughtrust and value, trust and value partly mediated the relationshipbetween information quality, and commitment. Apart from that,trust was also a partial mediator in the relationship between valueand commitment.

Drawing upon the above findings, network quality directlyinfluenced customer trust and value, and indirectly affectedcustomer commitment, supporting H1a and H1b, and partiallyconfirming H1c. The influence of customer service and technicalsupport was only found significant on trust; thus, H2a and H2c

were rejected whereas H2b received support. The direct effects ofinformation quality on trust, value, and commitment were sig-nificant, providing evidence for H3a, H3b, and H3c. Security andprivacy was positively and directly related to trust and value whileonly the indirect effect of this dimension on commitment wasevidenced. Therefore H4a, and H4b were supported, and H4c waspartially confirmed. In addition, direct relationships between trustand commitment, between value and commitment, and between

Table 3Regression weights in the main research model.

Path Estimates S.E. C.R. p

VAL o— NQ .43 .04 11.59 nnn

VAL o— CS � .02 .05 � .36 .72VAL o— IW .26 .07 3.79 nnn

VAL o— SP .34 .05 7.37 nnn

TRU o— NQ .12 .03 4.37 nnn

TRU o— CS .11 .03 3.54 nnn

TRU o— IW .27 .05 5.64 nnn

TRU o— SP .33 .03 9.77 nnn

TRU o— VAL .22 .02 10.35 nnn

COM o— NQ � .05 .03 �1.60 .11COM o— CS � .10 .04 �2.78 .01COM o— IW .17 .06 3.12 .00COM o— SP � .09 .04 �2.30 .02COM o— TRU .80 .05 15.62 nnn

COM o— VAL .07 .03 2.85 .00

Notes: CS¼Customer Service; NQ¼Network Quality; IW¼ Information Quality;SP¼Security and Privacy; COM¼Commitment; VAL¼Value; TRU¼Trust. Fullmodel: λ2(168)¼972.87, CMIN/DF¼5.79, GFI¼ .94, AGFI¼ .92, TLI¼ .96, CFI¼ .97,RMSEA¼ .06, 90% CI¼ .05:.06; SRMR¼ .03.

nnn pr .001.

Table 4Standardized total, direct and indirect effects in the main research model.

VAL TRU COM

Std. DE Std. IE Std. TE Std. DE Std. IE Std. TE Std. DE Std. IE Std. TE

NQ .38nnn .00 .38nnn .12nnn .10nnn .22nnn � .05 .22nnn .16nnn

CS � .01 .00 � .01 .10nnn � .01 .09nn � .09 .08nnn � .01IW .19nn .00 .19nn .23nnn .05nn .28nnn .15n .25nnn .40nnn

SP .28nnn .00 .28nnn .31nnn .07nnn .38nnn � .09 .34nnn .25nnn

VAL – – – .26nnn .00 .26nnn .09 .21nnn .30nnn

TRU – – – – – – .83 .00 .83nnn

Notes: CS¼Customer Service; NQ¼Network Quality; IW¼ Information Quality; SP¼Security and Privacy; COM¼Commitment; VAL¼Value; TRU¼Trust; Std. DE¼Standar-dized Direct Effect; Std. IE¼Standardized Indirect Effect; Std. TE¼Standardized Total Effect.

n pr .05.nn pr .01.nnn pr .001.

Table 5Structural results for different internet user groups in the main research model.

Path Light users Medium users Heavy users

VAL o— NQ .36nnn .53nnn .43nnn

VAL o— CS .09 .14 � .14n

VAL o— IW .22 � .09 .40nnn

VAL o— SP .37nnn .40nnn .30nnn

TRU o— NQ .07nnn .07 .16nnn

TRU o— CS .15n .05 .13nn

TRU o— IW .29nnn .31nn .24nnn

TRU o— SP .34nnn .32nnn .33nnn

TRU o— VAL .20nnn .24nnn .24nnn

COM o— NQ .04 � .13 � .06COM o— CS � .11 � .06 � .10COM o— IW .26nn .17 .11COM o— SP � .09 � .04 � .10COM o— VAL .02 .13n .08n

COM o— TRU .77nnn .89nnn .74nnn

Goodness of fitindices

λ2(168)¼422.10, CMIN/DF¼2.51, GFI¼ .91,AGFI¼ .88, TLI¼ .96, CFI¼ .97, RMSEA¼ .06,90% CI¼ .05:.07

λ2(168)¼359.17, CMIN/DF¼2.14, GFI¼ .91,AGFI¼ .88, TLI¼ .96, CFI¼ .97, RMSEA¼ .06,90% CI¼ .05:.07

λ(168)¼588.98, CMIN/DF¼3.51, GFI¼ .93,AGFI¼ .90, TLI¼ .96, CFI¼ .97, RMSEA¼ .06,90% CI¼ .05:.06

Chi Squaredifference test

Δλ2(28)¼46.14, p¼ .02

Notes: CS¼Customer Service; NQ¼Network Quality; IW¼ Information Quality; SP¼Security and Privacy; COM¼Commitment; VAL¼Value; TRU¼Trust.n pr .05.nn pr .01.nnn pr .001.

P. Thaichon et al. / Journal of Retailing and Consumer Services 21 (2014) 1047–1058 1053

trust and value were also established as shown in Table 3,confirming H5, H6, and H7.

Additionally, H8 postulated the moderating effect of Internetusage on the paths from service quality elements towards value,trust and commitment. To examine the interaction effect, thisstudy split the sample into three groups based on their usage level(light users: less than 9 h a week, medium user: 9–29 h a week;and, heavy users: more than 29 h a week) and determined paths atdifferent levels of the moderating variable. As indicated in Table 5,the main research models of Internet usage show reasonable fit tothe data. In addition, the difference of overall Chi square test wassignificant.

To further confirm the differences in each structural pathbetween the three groups of users, the structural models wereseparated for the three subsamples. The moderating effect wastested by constraining the twelve paths (from four quality dimen-sions to trust, value and commitment) to be equal, using the chi-square difference test for the effect of Internet usage. An uncon-strained model that simultaneously fit all three usage groups wasrun and the paths of interest were fixed to be invariant in allgroups to arrive at a constrained model (Cunningham, 2010). Inthe twelve models, only two models testing the paths fromcustomer service, and information quality towards value resultedin significant difference in the chi-square test.

This result suggests that the effects of customer service, andinformation quality on value were not the same for people fromdifferent Internet usage groups (Δχ2(2)¼7.84, p¼ .02 in the pathof customer service and technical support to value; Δχ2(2)¼7.30,p¼ .03 in the path of information quality to value). These findingsdemonstrate that the two paths were not invariant amongcustomers from different Internet usage groups, indicating astrong moderating effect of Internet usage. As illustrated inTable 6, Figs. 3 and 4, heavy users were distinctively differentfrom the others. Specifically, customer service had a negativeinfluence on value among heavy users. Meanwhile, the positiveimpact of information quality on value perceived by heavy userswas more significant as compared to that manifested by light andmedium users. On the basis of these findings, H8 was partiallysupported.

5. Discussion

This study aims to examine the individual influence of eachservice quality dimension on an ISP’s customers’ affective andcognitive evaluation. The results confirm that network quality,customer service and technical support, information quality, andprivacy and security have different influences on perceived value,trust and commitment. In addition, it was revealed that onlyinformation quality directly influenced customer commitment.The effect of customer service and technical support was foundsignificant only on trust. The findings are considered reasonablyrobust as the model explained a considerable portion of variancein customer perceived value, trust and commitment.

5.1. The relationship between service quality dimensionsand perceived value

As expected network quality was the strongest predictor ofperceived value. An ISP’s service package is considered to be ofhigh value when it includes good network quality. This is likely asnetwork quality is the core element of an ISP’s service offerings.The strength and stability of the network is usually the primaryconcern of customers when evaluating an ISP’s service. Thisfinding echoes Tam’s (2012) research which highlights that net-work quality is an antecedent of perceived value. Similarly,security and privacy as well as information quality were alsosignificant antecedents to value. This conclusion is not surprisingconsidering the information era today. Customers are well awareof the importance and integrity of the personal information givento their service providers. They also assess the value of the servicethrough the quality and quantity of the information they acquirefrom the providers. High-quality information tailored to thecustomer’s needs and wants allows customers to diminish thecosts of seeking and handling information.

Table 6Results of regressing value against customer service and information quality.

Usagegroup

Adjusted Rsquare

FUnstandardizedcoefficients

Customerservice

Light users .24 156.28 (Constant) .96Customerservice

.63

Mediumusers

.25 127.80 (Constant) 1.06

Customerservice

.60

Heavyusers

.27 204.03 (Constant) 3.96

Customerservice

� .52

Informationquality

Light users .35 229.22 (Constant) .69Informationquality

.74

Mediumusers

.32 165.40 (Constant) 3.13

Informationquality

� .25

Heavyusers

.37 442.17 (Constant) .66

Informationquality

.75

0

1

2

3

4

5

1 2 3 4 5

Val

ue

Customer Service

Light users

Medium users

Heavy users

Fig. 3. The relationship between customer service and perceived value amonglight, medium and heavy users.

0

1

2

3

4

5

1 2 3 4 5

Val

ue

Information Quality

Light users

Medium users

Heavy users

Fig. 4. The relationship between information quality, and perceived value amonglight, medium and heavy users.

P. Thaichon et al. / Journal of Retailing and Consumer Services 21 (2014) 1047–10581054

Surprisingly, the direct effect of customer service and technicalsupport on value was not evident from the results. This findingcontrasts the work of Blocker (2011) who found that customerservice was directly related to perceived value. This can beexplained because of the fact that in Asian cultures customerstypically tend to avoid in-person interaction with the company’spersonnel, especially when they experience lack of resources andsupport from the service provider. This is typical in Asian countriesand is generally attributed to customers’ ego and value orienta-tions (Neuliep, 2012).

5.2. The relationship between service quality dimensions and trust

Although all the service quality dimensions influenced custo-mers’ trust towards the service provider, security and privacy hadthe greatest influence. Internet services are characterised by aunique combination of online and offline transactions. Customersare able to choose to conduct their business with the ISPs online orusing brick-and-mortar retailers. However, the role of Internet hasbecome considerably important, especially to time-poor custo-mers, and this has resulted in numerous cases of illegal disclosureof personal data as well as exposure of transaction data. Customersare highly concerned about the integrity of their ISP and themanner in which their privacy and security are guaranteed. In fact,millions of dollars were stated lost due to cybercrime in 2012 (IC3,2012). In addition, according to a report by the Financial Times,basic personal information of customers is available at a cheap cost(Steel, 2013). As a result, the guarantee of privacy and security canenhance trust of an ISP’s customers which is similar to the findingsof Cases et al. (2010).

Moreover, an ISP’s customers are unable to see the company’tangibles, making it harder for them to decide if a service provideris trustworthy. Hence, customers establish trust in their serviceprovider on the basis of the network performance together withusability and clarity of the available information. Customersusually believe that a company which delivers stable and strongnetwork quality is reliable. Similarly, as the information is ade-quate and transparent, customers will most likely perceive thecompany as being trustworthy. In addition, customer service andtechnical support have an impact on customer trust. In otherwords, customer trust increases through superior customer ser-vices. This is evident as customer service personnel are animportant conduit between users and service providers. Whencustomers experience superior service and technical support inhandling their enquiries and complaints, they perceive the ISP tobe reliable and sincere, which then translates into trust.

5.3. The relationship between service quality dimensions andcommitment

Interestingly, among four ISP service quality dimensions, onlythe direct effect of information quality on commitment was foundsignificant. Customers develop a sense of belonging and commit-ment to an ISP based on perceived quality and quantity of theinformation made available to them by the company. This conclu-sion, once again, confirms the importance of information inservicing customers. Lack of support for the direct effects of otherservice quality dimensions on commitment may be possible ascustomers in Thailand consider network quality, customer serviceand privacy and security to be similar among major ISPs. Thisstems from compulsory industry standards and limited number ofservice providers. On the other hand, although only the effect ofcustomer service on trust is evident, it is critical to remember thatperceptions of customer service might influence other affectiveresponses excluded in this study (for example, satisfaction), whichwould have an impact on commitment.

The absence of direct effects of network quality, and privacyand security on commitment was compromised by other indirecteffects. The results reveal that value and trust were the precursorsto commitment, thereby being the mediators in the relationshipbetween network quality, and privacy and security and commit-ment. In fact the relationship between information quality andcommitment was also partly mediated by customer trust andvalue. Furthermore, whilst both of the direct effect of value ontrust and the effect of trust on commitment were significant, thedirect effect of value on commitment also remained significant,signaling the presence of a partially mediated effect rather than analternative fully mediated relationship. Hence, value had both adirect and an indirect effect on commitment. Table 4 indicates thedirect and indirect effects of each predictor on the criteriaconstructs. Service quality dimensions influences cognitive eva-luations (i.e. value), and affective responses (i.e. trust and commit-ment) which are important antecedents of customer loyalty.

5.4. Segmentation analysis

The invariance tests for different Internet usage groups revealsome interesting findings. Internet usage moderated the effects ofcustomer service and technical support, and information quality onvalue. Customer service and technical support had a negative effecton perceived value of heavy users. Also, information qualitydetermined how heavy users perceive the value of their servicepackage while this effect was absent in the other two groups. Onepossible explanation is that heavy users are usually conversant withtechnology and might consider the help from customer servicepersonnel unnecessary. Moreover, they tend to spend considerabletime on the Internet and prefer online tools, for example websites,to more conservative means of communications, for example face toface customer service. In addition, as mentioned earlier, Asiancustomers are more likely to avoid direct communications withan ISP’s staff in the interest of their ego. Heavy users usually takepride in their Internet expertise and hesitate to admit that they lackknowledge. As a result, customer service might not add any value tothe service package provided for heavy users.

6. Theoretical and managerial implications

6.1. Theoretical implications

Though widely discussed in the relevant literature, servicequality is rarely researched in the ISP context. The findings ofthe study demonstrate several theoretical contributions. Drawingupon previous studies (e.g. Aydin and Özer, 2005; Lai et al., 2009;Roca et al., 2009) and taking the unique characteristics of ISPservices into account, the current study identifies ISP’s servicequality dimensions, namely network quality, customer service andtechnical support, information quality, and privacy and security,and examines the relationships among ISP service quality dimen-sions, perceived value, customer trust and commitment. Finding ofthe significant effects on customer cognitive and affectiveresponses exerted by identified service quality factors has implica-tions for customer satisfaction and loyalty research. Establishmentof the mediation model enriches service quality literature andprovides insights into customer commitment research by incor-porating customer value and trust into the quality–commitmentrelationship. This finding presents a new perspective on andchallenge to the traditional chain relationship of service quality,customer satisfaction and commitment (loyalty or retention)proposed by Heskett et al. (1994).

Furthermore, this study extends the service quality literatureby including customer segmentation in the analysis. Consistent

P. Thaichon et al. / Journal of Retailing and Consumer Services 21 (2014) 1047–1058 1055

with the findings of Homburg and Giering (2001) this studyconfirms the mediating role of customer characteristics in therelationship among psychological factors. Internet usage mediatesthe link between customer service and technical support, as wellas information quality and perceived value. The effects of customerservice and technical support, and information quality on per-ceived value differ across different Internet usage groups.

6.2. Managerial implications

This research has developed an understanding of customer beha-viour relating to home Internet services. The results emphasise thecritical role of service quality and its dimensions, and highlight thesignificance of ISPs dedicating resources in improving the servicequality dimensions. Nowadays customers are geared with availableinformation and tools to make comparisons between services andcompanies. Hence it is crucial that companies invest in improvingservice quality in order to induce positive cognitive and affectiveresponses of customers and eventually increase their loyalty.

The findings suggest that management should guarantee net-work consistency and reliability, and satisfactory Internet speed.As network quality becomes more reliable among service provi-ders, a transparent policy of privacy protection and ensuredtransaction security can provide ISPs with a strong point ofdifferentiation. Findings from this study also highlight the needfor management to establish easy access and informative channels,especially websites, in order to provide adequate informationsuited to customer needs and wants. The company should keepin mind that these channels are means of information provisionand communications, hence they should have user friendly inter-faces and reliable functions. Furthermore, ISPs should attend to theaspects pertaining to customer service and technical support.Customer service and technical support personnel should also bereadily available. Staff with appropriate skills and competencemust demonstrate sincere interest in dealing with problems orissues in their responses to customers’ enquiries and complaints.

This study suggests that ISPs delve deeper into customersegmentation in order to effectively and efficiently understandthe market and develop appropriate marketing strategies fordifferent segments. In particular, ISPs should distinguish heavyInternet users from other segments. On the basis of the findingthat customer service and information support made exclusiveand significant contribution to the heavy users’ perception ofvalue, ISPs should offer different service package for these seg-ments with a view to differentiating service offering and max-imizing use of organizational resources. For example, informationpackage could be designed in accordance with different levels ofInternet usage and knowledge, such as novice, intermediate andexpert. Communications material should be tailored to suit eachsegment on the basis of their characteristics. Similarly, customercare could be varied by offering customers with a wide range ofalternatives, for instance, face to face consultation, online support,and one-off or periodical follow-ups relating to technical issues.These would eventually result in significant long-term profitability.

Noticeably, service providers should be aware of the negativeeffect of customer service on value among heavy users. Many ISPsattempt to create rapport with their customers through follow-upcalls or emails which might have adverse effects on heavy users dueto the orientations of value and face (Neuliep, 2012). Consequently,communications with heavy users should be succinct and covert. It isposited that heavy Internet users tend to be less comfortablecommunicating and establishing relationships offline (Thayer andRay, 2006). Therefore, instead of face to face consultation, onlinesupport could be a more appropriate alternative for heavy users.While being convenient, interactions through website environmentmay be more favourable for this group of Internet users as heavy

Internet users spend considerable time on the Internet and are mostlikely to prefer Internet related activities to non-Internet ones(Assael, 2005). Hence, it is advisable for ISPs to focus on the designof the information support platform in both online and offlineenvironment in order to increase the perceived value of heavy users.

In contrast to heavy users, light users should be approacheddifferently. Light users tend to emphasise security and privacy in theirevaluation of service value. Therefore, a thorough explanation ofsecurity and privacy policy are recommended for light users. Inaddition, information support demonstrates a significant effect oncommitment among light users. ISPs could provide informationpackage tailored to the needs and wants of customers in this Internetusage group. Light users are usually novices, and more likely to prefersimple information than complicated descriptions. Hence, any com-munications material intended for this segment should be specialisedat the beginner level. Moreover, although it is not directly related tovalue and commitment, customer service is an important factor oftrust, a precursor to customer commitment. Since light users spendthe least time online among the three usage segments, they probablychoose face to face support over online support. Follow-up calls andemails are also a good idea to improve customer experience with anISP service.

Finally, whereas network quality is the predominant factor indetermining value, privacy and security has the strongest influenceon trust of medium users. In addition to clarify the transparency andreliability of the company’ security and privacy practice, an ISP couldalso promote high network quality via their communications tomedium users, for example, brochures or flyers. Medium users areexpected to have better knowledge of Internet than light users albeitless knowledgeable than heavy users. Therefore, intermediate level ofinformation is considered appropriate. On the other hand, the role ofcustomer support appears to be less significant in the developmentof medium users’ commitment towards their ISP. Although servicequality dimensions demonstrate various effects on the outcomevariables among the different segments, ISPs should endeavor toensure delivering optimal services to these segments and avoidperceived prejudice by customers which may affect their perceptionof the firm’s service quality and subsequently their commitment.

7. Limitations and future research direction

This study has a few inherent limitations. First, the choice ofdestination where this study took place (i.e. Thailand) might limitthe generalisability of the findings due to cultural differences inother countries. Replications of this study should be considered totest the generalisation of this model in other contexts such as inVietnam, Cambodia and Burma. Second, the model in this studyonly includes value, trust and commitment as customers’ cognitiveand affective responses. An investigation on other factors such assatisfaction and loyalty could generate more insights. Third, thisstudy segmented ISP customers only based on their Internet usage.It might be desirable to include other characteristics of customersin order to understand thoroughly the importance of segmenta-tion in the ISP market. Finally, a longitudinal study of ISP’scustomers might be worthwhile to capture the changing patternsof customers’ psychology and behaviours over time.

References

ACMA 2012. Internet use and Social Networking by Young People, AustralianCommunications and Media Authority, available at: ⟨http://www.acma.gov.au/webwr/_assets/main/lib310665/no1_internet_use_social_networking_young_people.pdf⟩. (accessed 1/04/2014.

Asmussen, B., Harridge-March, S., Occhiocupo, N., Farquhar, J., 2013. The multi-layered nature of the internet-based democratization of brand management.J. Bus. Res. 66, 1473–1483.

P. Thaichon et al. / Journal of Retailing and Consumer Services 21 (2014) 1047–10581056

Assael, H., 2005. A demographic and psychographic profile of heavy internet usersand users by type of internet usage. J. Advertising Res. 45 (1), 93–123.

Aydin, S., Özer, G., 2005. The analysis of antecedents of customer loyalty in theTurkish mobile telecommunication market. Eur. J. Marketing 39 (7/8), 910–925.

Baron, R.M., Kenny, D.A., 1986. The moderator-mediator variable distinction insocial psychological research: conceptual, strategic, and statistical considera-tions. J. Pers. Soc. Psychol. 51 (6), 1173–1182.

Benedicktus, R.L., 2011. The effects of 3rd party consensus information on serviceexpectations and online trust. J. Bus. Res. 64, 846–853.

Blocker, C.P., 2011. Modeling customer value perceptions in cross-cultural businessmarkets. J. Bus. Res. 64, 533–540.

Cases, A., Fournier, C., Dubois, P., Tanner, J.F., 2010. Web Site spill over to emailcampaigns: the role of privacy, trust and shoppers’ attitudes. J. Bus. Res. 63,993–999.

Chae, M., Kim, J., Kim, H., Ryu, H., 2002. Information quality for mobile internetservices: a theoretical model with empirical validation. Electron. Markets 12 (1),38–46.

Chen, Y., Chang, C., 2012. Enhance green purchase intentions. The roles of greenperceived value, green perceived risk, and green trust. Manage. Decis. 50 (3),502–520.

Chiou, S., 2004. The antecedents of consumers’ loyalty toward Internet ServiceProviders. Manage. Sci. 41, 685–695.

CIA 2013. The World Factbook, Central Intelligence Agency, available at: ⟨https://www.cia.gov/library/publications/the-world-factbook/geos/th.html⟩ (accessed1st August 2013).

Cronin, J.J., Brady, M.K., Hult, G.T., 2000. Assessing the effects of quality, value, andcustomer satisfaction on consumer behavioral intentions in service environ-ments. J. Retailing 76 (2), 193–218.

Cunningham, E. 2010. A Practical Guide to Structural Equation Modelling UsingAmos, first ed., Statsline, Melbourne.

Deng, Z., Lu, Y., Wei, K., Zhang, J., 2010. Understanding customer satisfaction andloyalty: an empirical study of mobile instant messages in China. Int. J. Inf.Manage. 30, 289–300.

Eisingerich, A., Rubera, G., 2010. Drivers of brand commitment: a cross-nationalinvestigation. J. Int. Marketing 18 (2), 64–79.

Eisingerich, A.B., Bell, S.J., 2008. Perceived service quality and customer trust doesenhancing customers’ service knowledge matter? J. Serv. Res. 10 (3), 256–268.

Elliot, S., Li, G., Choi, C., 2013. Understanding service quality in a virtual travelcommunity environment. J. Bus. Res. 66, 1153–1160.

ETDA 2013. Thailand Internet User Profile 2013, Electronic Transactions Develop-ment Agency, available at: ⟨http://www.etda.or.th/etda_website/files/system/InternetUserProfile_2013_14.11.13.pdf⟩ 15/05/2014).

Fornell, C., Larcker, D.F., 1981. Evaluating structural equation models with unob-servable variables and measurement error. J. Marketing Res. 18 (1), 39–50.

Freedom 2012. Safety on the Line: Exposing the Myth of Mobile CommunicationSecurity—Socialist Republic of Vietnam, Freedom House, available at: ⟨http://www.refworld.org/docid/502a0c550.html⟩ 20/05/2014).

Fullerton, G., 2005. How commitment both enables and undermines marketingrelationships. Eur. J. Marketing 39 (11), 1372–1388.

Grewal, D., I., G.R., Krishnan, R., Sharma, A., 2003. The internet and the price-value-loyalty Chain. J. Bus. Res. 56 (5), 391–398.

Ha, S., Stoel, L., 2012. Online apparel retailing: roles of e-shopping quality andexperiential e-shopping motives. J. Serv. Manage. 23 (2), 197–215.

Hauser, J.H., Wernerfelt, B., 1990. An evaluation cost model of consideration sets. J.Consum. Res. 16 (4), 393–408.

He, H., Li, Y., 2010. Key service drivers for high-tech service brand equity: themediating role of overall service quality and perceived value. J. MarketingManage. 27 (1/2), 77–99.

Heskett, J.L., Jones, T., Loveman, G.W., Sasser Jr, W.E., Schlesinger, L.A., 1994. Puttingthe service profit chain to work. Harv. Bus. Rev. 72 (2), 164–174.

Homburg, C., Giering, A., 2001. Personal characteristics as moderators of therelationship between customer satisfaction and loyalty—an empirical analysis.Psychol. Marketing 18, 43–66.

Hong, P., Cho, H., 2011. The impact of consumer trust on attitudinal loyalty andpurchase intentions in B2C e-marketplaces: intermediary trust vs. seller trust.Int. J. Inf. Manage. 31 (5), 469–479.

IC3 2012. Internet Crime Report 2012, Internet Crime Complain Center, available at:⟨http://www.ic3.gov/media/annualreport/2012_IC3Report.pdf⟩.

IWS 2013. Asia Marketing Research, Internet Usage, Population Statistics andFacebook Information, Internet World Stats, available at: ⟨http://www.internetworldstats.com/asia.htm⟩ (accessed 4th November 2013).

Karjaluoto, H., Jayawardhena, C., Leppaniemi, M., Pihlstrom, M., 2012. How valueand trust influence loyalty in wireless telecommunications industry. Telecom-mun. Policy 36, 636–649.

Kim, B., 2012. The diffusion of mobile data services and applications: exploring therole of habit and its antecedents. Telecommun. Policy 36 (1), 69–81.

Kim, H., Niehm, S., 2009. The impact of website quality on information quality,value, and loyalty intentions in apparel retailing. J. Interact. Marketing 23 (3),221–233.

Ko, E., Taylor, C.R., Sung, H., Lee, J., Wagner, U., Navarro, D.M., Wang, F., 2012. Globalmarketing segmentation usefulness in the sportswear industry. J. Bus. Res. 65,1565–1575.

Lacey, R., 2007. Relationship drivers of customer commitment. J. Marketing TheoryPract. 15 (4), 315–333.

Ladhari, R., 2010. Developing e-service quality scales: a literature review. J. RetailingConsum. Serv. 17 (6), 464–477.

Lai, F., Griffin, M., Babin, B., 2009. How quality, value, image, and satisfaction createloyalty at a Chinese telecom. J. Bus. Res. 62, 980–986.

Lee, Y., Park, J., Chung, N., Blakeney, A., 2012. A unified perspective on the factorsinfluencing usage intention toward mobile financial services. J. Bus. Res. 65,1590–1599.

Lin, J., 2007. Online stickiness: its antecedents and effect on purchasing intention.Behav. Inf. Technol. 26 (6), 507–516.

Lin, J., Wu, C., 2011. The role of expected future use in relationship-based serviceretention. Manag. Serv. Qual. 21 (5), 535–551.

Malhotra, N., Mavondo, F., Mukherjee, A., Hooley, G., 2013. Service quality offrontline employees: a profile deviation analysis. J. Bus. Res. 66, 1338–1344.

Mazzoni, C., Castaldi, L., Addeo, F., 2007. Consumer behavior in the Italian mobiletelecommunication market. Telecommun. Policy 31 (10/11), 632–647.

McDougall, G., 2001. Customer retention strategies: when do they pay off? Serv.Marketing Q. 22 (1), 39–55.

McKenzie, C.R. M., 1998. Taking into account the strength of an alternativehypothesis. J. Exp. Psychol. 24 (3), 771–792.

Morgan, M., Hunt, D., 1994. The commitment-trust theory of relationship market-ing. J. Marketing 58 (3), 20–38.

Musa, R., Pallister, J., Robson, M., 2005. The roles of perceived value, perceivedequity and relational commitment in a disconfirmation paradigm framework:an initial assessment in a ‘relationship-rich’ consumption environment. Adv.Consum. Res. 32 (349-357).

NBTC 2013. Information about Telecommunications Industry, The National Broad-casting and Telecommunications Commission, available at: ⟨http://www.nbtc.go.th/wps/portal/NTC/!ut/p/c5/04_SB8K8xLLM9MSSzPy8xBz9CP0os3hnd0cPE3MfAwMDP09TAyOXMBc_T1dLAwMnU_1wkA48Kowg8gY4gKOBvp9Hfm6qfkF2dpqjo6IiAEc6d5w!/dl3/d3/L2dJQSEvUUt3QS9ZQnZ3LzZfNE83RDVQRzVJQkFWNDBJQ0JDOEo1MTJTQjQ!/?WCM_GLOBAL_CONTEXT=/wps/wcm/connect/libraryþntc/internetsite/04newsactivi/0402newspaper/040201press/040201press_detail/0fbec200421ef7aba010e28b91da655b⟩.

Neuliep, J.W., 2012. Intercultural Communication: A Contextual Approach, fifth ed.SAGE Publications, Inc, St. Norbert College.

Nunnally, J., 1978. Psychometric Theory. McGraw-Hill, New York.Oracle 2012. Consumer Views of Live Help Online 2012: A Global Perspective,

Oracle, available at: ⟨http://www.oracle.com/us/products/applications/commerce/live-help-on-demand/oracle-live-help-wp-aamf-1624138.pdf 20/05/2014⟩).

Ou, W., Shih, C., Chen, C., Wang, K., 2011. Relationships among customer loyaltyprograms, service quality, relationship quality and loyalty: an empirical study.Chin. Manage. Stud. 5 (2), 194–206.

Parasuraman, A., Grewal, D., 2000. The impact of technology on the quality-value-loyalty chain: a research agenda. J. Acad. Marketing Sci. 28 (1), 168–174.

Parasuraman, A., Zeithaml, V.A., Malhotra, A., 2005. E-S-QUAL. A multiple-itemscale for assessing electronic service quality. J. Serv. Res. 7 (10), 1–21.

Perry, M.L., Sengupta, S., Krapfel, R., 2004. Effectiveness of horizontal strategicalliances in technologically uncertain environments: are trust and commitmentenough? J. Bus. Res. 57, 951–956.

Preacher, K.J., Kelley, K., 2011. Effect size measures for mediation models:quantitative strategies for communicating indirect effects. Psychol. Methods16 (2), 93–115.

Prentice, C. (2013), Who stays, who walks, and why in high-intensity servicecontexts, J. Bus. Res., 10.1016/j.jbusres.2013.02.044.

Ratchford, B.T., 1982. Cost-benefit models for explaining consumer choice andinformation seeking behavior. Manage. Sci. 28 (2), 197–212.

Ringle, C.M., Sarstedt, M., Schlittgen, R., Taylor, C.R., 2013. PLS path modeling andevolutionary segmentation. J. Bus. Res. 66, 1318–1324.

Roca, J., García, J., Vega, J., 2009. The importance of perceived trust, security andprivacy in online trading systems. Inf. Manage. Comput. Secur. 17 (2), 96–113.

Rowe, B., Wood, D., Reeves, D., Braun, F. 2011. The Role of Internet Service Providersin Cyber Security, Institute for Homeland Security Solutions, available at:⟨http://sites.duke.edu/ihss/files/2011/12/ISP-Provided_Security-Research-Brief_Rowe.pdf⟩ 20/05/2014).

Rutherford, B., 2012. Building buyer commitment to the salesperson. J. Bus. Res. 65,960–967.

Sabiote, E., Roman, S., 2009. The influence of social regard on the customer-servicefirm relationship: the moderating role of length of relationship. J. Bus. Psychol.24 (4), 441–453.

Santouridis, I., Trivellas, P., 2010. Investigating the impact of service quality andcustomer satisfaction on customer loyalty in mobile telephony in Greece. TotalQ. Manage. J. 22 (3), 330–343.

Sheth, J.N., Newman, B.I., Gross, B.L., 1991. Why we buy what we buy: a theory ofconsumption values. J. Bus. Res. 22 (2), 159–170.

Shirin, A., Puth, G., 2011. Customer satisfaction, brand trust and variety seeking asdeterminants of brand loyalty. Afr. J. Bus. Manage. 5, 11899–11915.

Steel, E. 2013. Companies Scramble for Consumer Data, Financial Times, availableat: ⟨http://www.ft.com/intl/cms/s/0/f0b6edc0-d342-11e2-b3ff-00144feab7de.html#axzz2W5lJhSoC⟩ (accessed 1st August 2013).

Streamshield 2004. Consumers Prepared to Pay Extra for Clean and Safe InternetService, StreamShield Networks, available at: ⟨http://www.streamshield.com/index.php?option=com_content&task=view&id=59&Itemid=130 20/05/2014⟩).

Sweeney, J.C., Soutar, G.N., 2001. Customer perceived value: the development of amultiple item scale. J. Retailing 77 (2), 203–220.

Tabachinick, B.G., Fidell, L.S., 2001. Using Multivariate Statistics. Allyn and Bacon,Boston.

P. Thaichon et al. / Journal of Retailing and Consumer Services 21 (2014) 1047–1058 1057

Tai, Y., 2011. Perceived value for customers in information sharing services. Ind.Manage. Data Syst. 111 (4), 551–569.

Tam, J., 2004. Customer satisfaction, service quality and perceived value: anintegrative model. J. Marketing Manage. 20 (7/8), 897–917.

Tam, J., 2012. Linking perceived service quality to relational outcomes in a chinesecontext. J. Int. Consum. Marketing 24 (1), 7–23.

Thayer, S.E., Ray, S., 2006. Online communication preferences across age, gender,and duration of Internet use. Cyberpsychol. Behav. 9 (4), 432–440.

Thaichon, P., Lobo, A., Mitsis, A. 2012. Investigating the Antecedents to Loyalty ofInternet Service Providers in Thailand: Developing a Conceptual Model, PaperPresented to ANZMAC 2012, Adelaide, Australia.

Thaichon, P., Lobo, A., Mitsis, A., 2014. An empirical model of home internet servicesquality in Thailand. Asia Pac. J. Marketing Logist. 26 (2), 190–210.

Thaichon, P., Quach, T.N., 2013. Integrated marketing communications and theirinfluences on brand loyalty: a Thai perspective. Int. J. Stud. Thai Bus., Soc. Cult.2 (1), 55–79.

Tuškej, U., Golob, U., Podnar, K., 2013. The role of consumer-brand identification inbuilding brand relationships. J. Bus. Res. 66, 53–59.

Vlachos, A., Vrechopoulos, P., 2008. Determinants of behavioral intentions in themobile internet services market. J. Serv. Marketing 22 (4), 280–291.

Wang, C., Wu, L., 2012. Customer loyalty and the role of relationship length.Managing Serv. Qual. 22 (1), 58–74.

Wang, Y., Lo, H., 2002. Service quality, customer satisfaction and behavior inten-tions: evidence from China’s telecommunication industry. info 4 (5), 50–60.

Wedel, M., Kamakura, W., 2003. Market Segmentation. Conceptual and Methodo-logical Foundations. Kluwer Academic Publishers, Norwell, MA.

Weinstein, A., 2002. Customer retention: a usage segmentation and customer valueapproach. J. Target., Meas. Anal. Marketing 10 (3), 259–268.

White, H., Nteli, F., 2004. Internet banking in the UK: why are there not morecustomers? J. Finan. Serv. Marketing 9 (1), 49–56.

Wolfinbarge, M., Gilly, C., 2003. eTailQ: dimensionalizing, measuring and predictingetail quality. J. Retailing 79 (3), 183–198.

WorldBank 2014. Internet Users, WorldBank, available at: http://data.worldbank.org/indicator/IT.NET.USER.P2 20/05/2014).

Wu, J., Chen, Y., Chung, Y., 2010. Trust factors influencing virtual communitymembers: a study of transaction communities. J. Bus. Res. 63, 1025–1032.

Yang, Z., Bi, Z., Zhou, N., 2005. The double jeopardy phenomenon and the mediatingeffect of brand penetration between advertising and brand loyalty. J. Advertis-ing Res. 45 (2), 221.

Zhao, Y.L., Benedetto, C.A. D., 2013. Designing service quality to survive: empiricalevidence from Chinese new ventures. J. Bus. Res. 66, 1098–1107.

P. Thaichon et al. / Journal of Retailing and Consumer Services 21 (2014) 1047–10581058