- Email: [email protected]
Website usability, consumer satisfaction and the intention to use a website: The moderating effect of perceived risk
Journal of Retailing and Consumer Services 19 (2012) 124–132
Contents lists available at SciVerse ScienceDirect
Journal of Retailing and Consumer Services journal homepage: www.elsevier.com/locate/jretconser
Website usability, consumer satisfaction and the intention to use a website: The moderating effect of perceived risk Daniel Belanche a,1, Luis V. Casalo´ b, 2, Miguel Guinalı´u a,n a b
´n de Marketing e Investigacio ´n de Mercados, Facultad de Economı´a y Empresa, University of Zaragoza, Gran Vı´a 2, 50005 Zaragoza, Spain Departamento de Direccio ´n de Marketing e Investigacio ´n de Mercados, Facultad de Empresa y Gestio ´n Pu ´ blica, University of Zaragoza, Ronda Misericordia 1, 22001 Huesca, Spain Departamento de Direccio
a r t i c l e i n f o
a b s t r a c t
Available online 12 November 2011
This paper analyzes the influence of website usability on both consumers’ satisfaction and intention to use a website, as well as the impact of satisfaction on usage intentions. Additionally, we study the moderating effect that consumer risk perceptions may have on the influence of website usability. Results show that website usability affects satisfaction which in turn affects intention to use. Contrary to expected, usability does not directly affect intention to use but has an indirect effect through consumer satisfaction. Finally, the usability effect on consumer satisfaction is moderated by perceived risk. & 2011 Elsevier Ltd. All rights reserved.
Keywords: Usability Satisfaction Intention to use Perceived risk Online consumer behavior
1. Introduction According to Oneupweb (2010), in the context of online transactions, online users’ mainly expect websites that facilitate their purchases. In this line, 95.5% of users expect pricing and shipping information to be clearly presented, so usability must be an essential component of e-retail strategy. Website usability can be defined considering the following aspects (Flavia´n et al., 2006): (a) the ease of understanding the structure of a website, its functions, interface and the contents that can be observed by the user; (b) simplicity of use of the website in its initial stages; (c) the speed with which the users can find what they are looking for; (d) the perceived ease of site navigation in terms of time required and action necessary in order to obtain the desired results; and (e) the ability of the user to control what they are doing, and where they are, at any given moment. Analogously to merchandising in offline stores, when a customer accesses to an online store, usability issues may affect customer’s perceptions and behaviors. Thus, more usable websites tend to create more positive attitudes toward online stores and increase conversion rates, whereas less usable websites have the opposite effect (Becker and Mottay, 2001). Considering the influence of usability on website performance, online retailers require a full understanding of this variable. Previous
n
Corresponding author. Tel.: þ34 976761000x4695; fax: þ 34 976761667. E-mail addresses: [email protected] (D. Belanche), [email protected] (L.V. Casalo´), [email protected] (M. Guinalı´u). 1 Tel.: þ34 976761000x4636; fax: þ 34 976761667. 2 Tel.: þ34 976761000x4695; fax: þ 34 976761667. 0969-6989/$ - see front matter & 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.jretconser.2011.11.001
literature in the field of retailing has studied the concept of usability, especially as a component of quality of service (Ladhari, 2010). However, the integration of usability management in the e-business strategy and the role of perceived risk are outstanding issues. This research allows us to understand the relevance of usability on strategic indicators of companies’ success analyzing the influence of perceived usability on the levels of consumer satisfaction and intention to use a website. Moreover, although the Internet shopping has been proposed to involve more risk than traditional shopping (Lee and Turban, 2001), different consumers may perceive different levels of risk in the same online environment. Thus, this research contributes to the literature analyzing whether risk moderates the influence of perceived usability on consumer satisfaction and intention to use a website. Bearing these considerations in mind, we structure the remainder of this article as follows. In the following section we formalize the working hypotheses. Next, data collection and measure validation processes are explained, followed by the results of the analyses. Finally, we discuss the main conclusions, managerial implications, and limitations of the study, as well as some possibilities for future research.
2. Hypotheses formulation Traditionally, several authors have stated that satisfaction – defined as an affective condition that results from a global evaluation of all the aspects that make up a relationship – is a crucial antecedent of re-purchase and re-usage intentions of a product or service (e.g. Oliver, 1980; Anderson and Sullivan, 1993). Focusing on the new technologies context, satisfaction
D. Belanche et al. / Journal of Retailing and Consumer Services 19 (2012) 124–132
with prior use is also found to be the strongest predictor of users’ continuance intentions (Bhattacherjee, 2001). In the same line, Devaraj et al. (2002) measured customer satisfaction in the e-commerce context and supported empirically that satisfaction is a key determinant of customer channel preference. More recently, researchers have found that satisfaction positively affects loyalty intentions toward online shopping (Chiu et al., 2009) and boosts the use of Internet portals (Lin et al., 2005), e-services (Liao et al., 2007) and online communities as well (Liu et al., 2010). Broadly speaking, literature supports that satisfied consumers exhibit a greater intention to use firm’s products, have a greater re-purchase intention, favor positive word-of-mouth and have a lower to look for alternative providers (Oliver, 1999; Kim et al., 2009). Consequently, it is expected that once users achieve certain levels of satisfaction with prior use of a website, the perceptions of satisfaction will influence their re-usage intentions. Thus, we propose our first hypothesis: H1. Consumer satisfaction has a positive effect on consumer intention to use a website. In one of the first works focusing on the determinants of satisfaction in online exchanges, Szymanski and Hise (2000) noted the ability of website design to promote satisfactory purchase experiences. This result has been supported by several authors afterwards. For example, Kim and Eom (2002) concluded that usability is of critical importance in achieving user satisfaction. More recently, Maditinos and Theodoridis (2010) pointed out that both quality of the interface and of the information provided to consumers (two key aspects of website usability) have a significant effect on the levels of users’ satisfaction. We must also note that making the purchase easier is one of the main motivations of consumers’ online purchasing (Bridges and Florsheim, 2008). Therefore, when visiting a website, consumers expect to find a channel whose features facilitate search, selection, payment and post-purchase actions. Consequently, it is reasonable to say that satisfaction with the purchase experience depend on website ease of use (Shankar et al., 2003). Complex and not intuitive interfaces, long purchasing processes, non-updated information or non-relevant information create an interaction atmosphere that negatively affects consumer satisfaction. According to all these ideas, we propose the following hypothesis: H2. Perceived website usability has a positive effect on consumer satisfaction. However, literature on online consumer behavior found that usability impact is even more relevant since it not only affects satisfaction but also favor future purchase intentions. In this respect, Flavia´n et al. (2006) found that usability positively affects consumer loyalty to a website. As well, Abdeldayem (2010) noted that attitudes toward online shopping and the intention to shop online are affected by ease of use. Therefore, we propose in our third hypothesis that: H3. Perceived website usability has a positive effect on consumer intention to use the website. 2.1. Moderating effect of perceived risk Focusing on the online context, the importance of risk perceptions is evident when conducting an e-purchase. Compared to traditional shopping, online commercial establishments are less known to consumers (Lee and Turban, 2001), and the absence of face-to-face interaction has introduced more uncertainty and risk (Wu and Chen, 2005). Apart from risk associated with intangibility and time lag, customers have to release personal and
125
financial data to not always well known providers (De Ruyter et al., 2001). Moreover, the outcome of this process depends on the behavior of the e-marketer and this is not within the consumer’s control (Lee and Turban, 2001). Therefore, perceived risk is a relevant parameter, especially in the early stages of a process of customer adoption of online purchase of products and services (De Ruyter et al., 2001; Hsu and Chiu, 2004). As a result of numerous online frauds and piracy crimes in the network, users need more and more signs to make sure that financial and personal information is under their control and that privacy and security are guaranteed. In doing so, usability could be understood as a sign of competence, ability and honesty by the organization hosting the website. Firms often use signals to communicate the level of some unobservable quality, especially to reduce the information asymmetry between buyers and sellers (Schlosser et al., 2006). Since an online transaction is usually unobservable by consumers before purchase (Schlosser et al., 2006), website usability might serve to demonstrate that the company is able to provide high-quality services and understands and is sensitive to online consumers’ fears. Besides, consumer perceptions about their own control of the online transaction process could be increased by perceived usability (Casalo´ et al., 2007), since greater usability is associated to low levels of difficulty to manage a system (Davis, 1989). That is, users perceptions of their own skills depend not only on their experience but also on the usability of the website (Flavia´n et al., 2006). In those cases in which consumers perceive high risk in an online transaction, website usability may help reduce this uncertainty and thus its effect in forming consumer satisfaction and intentions to use a website may be reinforced. To be precise, consumers perceiving high risk would more likely try to eliminate any suspicion when interacting with the website. Usability communicates information about performance, thus consumers might infer that a company that has invested in usability and website design can successfully handle online transactions (Schlosser et al., 2006). Thus, in a high-risk situation consumers will be more prone to carefully analyze all the information about the website and, as a result, website usability may help overcome these fears and form a more favorable opinion of using a website. In turn, when perceived risk is low, consumers will be less influenced by usability perceptions since they do not need these reinforcements to overcome worry about the possibility of nondesired outcomes. Taking into account the previous considerations we propose in the following hypothesis a moderating role of risk in the effect of usability on both satisfaction and intention to use. H4. If perceived risk increases, the relationship between perceived website usability and: (a) consumer satisfaction and (b) consumer intention to use the website will be strengthened. To sum up, the research model can be seen in Fig. 1.
3. Data collection Data were obtained though a web survey targeted to the users of a Spanish online retailer, which helped us to recruit participants. This practice is consistent with common online market research (e.g. Steenkamp and Geyskens, 2006). To be precise, data was collected among the users of a bus ticket e-selling service offered by one of the most important transport companies in Spain. It is important to note that only few companies operate online in this sector in Spain. In order to measure the variables, a structured questionnaire containing closed questions was developed. This questionnaire was designed to gather information about the studied constructs
126
D. Belanche et al. / Journal of Retailing and Consumer Services 19 (2012) 124–132
Consumer Satisfaction
H2: + Perceived Usability
H1: +
H3: + H4a
Intention to Use H4b
Perceived Risk Note: “- - >” reflects hypotheses affected by moderation Fig. 1. Research model.
Table 1 Representative nature of the data collected.
Sample size Age (years) o24 25–34 35–44 444 Gender Males Females
Current research project
AIMC (2009)
RED.ES (2009)
214
30,705
19,131
28.5% 27.6% 18.2% 25.7%
24.4% 28.5% 21.8% 25.2%
20.3% 30.5% 24.7% 24.5%
52.3% 47.7%
54.6% 45.4%
52.7% 47.3%
and other data in order to provide a more detailed insight of the socio-demographic characteristics of the user. Specifically, the consumer was asked to indicate his/her level of agreement or disagreement with a series of statements about the perceived usability and risk3 in the retailer’s website as well as their levels of satisfaction and intention to use the website (see Appendix). All questions related to construct used 7-point Likert scale. Finally, this method of collecting data, which relies on volunteer sampling, generated 214 valid questionnaires (atypical cases, repeated responses and incomplete questionnaires were removed). However, because we cannot statistically assess the reliability or possible bias associated with this non-random sample, we compare our sample characteristics with available information about a wider population, that is, the socio-demographic characteristics of the respondents to the largest studies of the online Spanish-speaking population (AIMC, 2009; RED.ES, 2009). The results appear very similar in terms of the age and sex of the respondents (see Table 1), which supports the representative nature of our sample.
4. Measures validation 4.1. Content and face validity An initial set of items was proposed once the relevant literature was revised. This review guaranteed the content validity of the 3 In order to measure risk perceptions, Jacoby and Kaplan (1972) distinguished between five risk dimensions in the overall construct – financial, performance, psychological, physical and social risk – although not every transaction involve all kind of risks. Consequently, in this work we decide to adapt a general measure of overall risk (Stone and Grønhaug, 1993).
measurements instruments, or the degree to which items correctly represent the theoretical content of the construct. To be precise, our measures were inspired by previous scales regarding perceived website usability (Flavia´n et al., 2006; Roy et al., 2001), customer satisfaction (Brockman, 1998; Severt, 2002; Janda et al., 2002; Smith and Barclay, 1997; Guinalı´u, 2005), intention to use (Wu and Chang, 2005; Cronin et al., 2000) and overall perceived risk (Stone and Grønhaug, 1993). These previous scales had been already proposed on either a service or online context, so that adaption to our specific context becomes natural. More efforts were needed to adapt the overall perceived risk scale to the online context since the original scale was tested in offline shopping. Due to this adaptation, face validity (the degree that respondents judge that the items are appropriate to the targeted construct) was also tested through a variation of the Zaichkowsky method (1985), whereby each item is qualified by a panel of experts (a total of 8 people who are professionals in different disciplines such as marketing, sociology, new technologies and Internet) as ‘‘clearly representative’’, ‘‘somewhat representative’’ or ‘‘not representative’’ of the construct of interest. In line with Lichtenstein et al. (1990) an item was retained if a high level of consensus was observed among the experts.
4.2. Exploratory analysis of reliability and dimensionality The validation process started with an initial exploratory analysis of reliability and dimensionality (Anderson and Gerbing, 1988). Cronbach’s alpha indicator was used to assess the initial reliability of the scales, considering a minimum value of .7 (e.g. Cronbach, 1970). The item-total correlation was used to improve the levels of Cronbach’s alpha, considering a minimum value of .3 (De Vaus, 2001). For these initial tasks we used statistical software SPSS v.14.0. All items were adjusted to the required levels. Second, we proceeded to evaluate the unidimensionality of the scales proposed by carrying out a principal components analysis. Factor extraction was based on the existence of eigenvalues higher than 1. Moreover, it was required that factorial loadings were higher than .5 points and a significant total explained variance (Hair et al., 1998). Construct validity was supported by principal component analysis with varimax rotation (see Table 2). After conducting the exploratory factor analysis using all items together, only one factor
Table 2 Principal components analysis with varimax rotation. Item
USAB1 USAB2 USAB3 USAB4 USAB5 USAB6 USAB7 SAT1 SAT2 SAT3 SAT4 IUSE1 IUSE2 IUSE3 RISK1 RISK2 RISK3
Construct loadings Factor 1 Usability
Factor 2 Int. to use
Factor 3 Satisfaction
Factor 4 Risk
.768 .830 .824 .864 .839 .761 .733 .411 .547 .513 .506 .076 .157 .240 .066 .125 .121
.115 .172 .160 .085 .120 .066 .086 .333 .269 .230 .252 .896 .914 .752 .071 .158 .147
.337 .132 .199 .186 .274 .226 .304 .699 .664 .667 .755 .124 .165 .336 .203 .078 .050
.070 .060 .035 .130 .117 .181 .203 .133 .196 .222 .138 .092 .140 .219 .788 .889 .918
D. Belanche et al. / Journal of Retailing and Consumer Services 19 (2012) 124–132
was extracted from each scale corresponding to: usability, satisfaction, perceived risk and intention to use.
127
Table 4 Convergent validity. Item
Loading
t-value
Item
Loading
t-value
USAB1 USAB2 USAB3 USAB4 USAB5 USAB6 USAB7
.826a .790a .822a .872a .890a .782a .786a
14.051 15.308 14.993 14.778 15.800 12.054 13.363
SAT1 SAT2 SAT3 SAT4 IUSE1 IUSE2 IUSE3 RISK2 RISK3
.817a .912a .882a .941a .819a .956a .823a .933a .884a
12.361 15.696 14.563 16.334 13.086 12.964 8.147 12.352 10.752
4.3. Confirmatory analysis of dimensionality With the aim of confirming the dimensional structure of the scales, we used the Confirmatory Factor Analysis. That is, we included all individual-level constructs in a single confirmatory factor model. In addition, this process also allows for a stringent test of convergent and discriminatory validity (Steenkamp and Geyskens, 2006). We employed the statistical software EQS v.6.1. As an estimation method we chose Robust Maximum Likelihood, since it affords more security in samples which might not present ¨ multivariate normality. The criteria proposed by Joreskog and ¨ Sorbom (1993) were followed:
The weak convergence criterion, which supposes eliminating
indicators that do not show significant factor regression coefficients. The strong convergence criterion, which involves eliminating non-substantial indicators (those indicators whose standardized coefficients are lower than .5). ¨ ¨ According to the suggestion of Joreskog and Sorbom (1993), we also eliminated the indicators that least contributes to the explanation of the model, taking R2 o.3 as a cut-off point.
a
Table 5 Discriminant validity. Pair of constructs
Correlation
Standard deviation
95% confidence interval
USAB–SAT USAB–IUSE USAB–RISK SAT–IUSO SAT–RISK IUSE–RISK
.824a .378a .286a .566a .381a .349a
.031 .067 .077 .065 .084 .080
.76324 .24668 .43692 .4386 .54564 .5058
a
Following these recommendations, we eliminated the item RISK1 (see Appendix) and we finally obtained acceptable levels of convergence, R2 (see Table 4) and model fit (Chi-square¼245.789, 98 d.f., po.001; Satorra–Bentler scaled Chi-square¼160.565, 98 d.f., p ¼.00007; Bentler–Bonett Normed Fit Index¼.912; Bentler– Bonett Nonnormed Fit Index¼.955; Comparative Fit Index (CFI)¼ .963; Bollen (IFI) Fit Index¼.964; Root Mean Sq. Error of App. (RMSEA) ¼.055; 90% Confidence Interval of RMSEA (.039; .069); normed Chi-square¼2.5081). 4.4. Composite reliability Additionally, we used the composite reliability indicator to asses construct reliability. Although Cronbach’s alpha indicator is the most frequent test to assess reliability, some authors consider that it underestimates reliability (e.g. Smith, 1974), so that the use of ¨ composite reliability (rc) has been suggested (Joreskog, 1971). Table 3 Descriptive statistics, construct reliability and convergent validity. Construct
Mean
S.D.
rc
AVE
Usability USAB1 USAB2 USAB3 USAB4 USAB5 USAB6 USAB7 Satisfaction SAT1 SAT2 SAT3 SAT4 Intention to use IUSE1 IUSE2 IUSE3 Risk RISK2 RISK3
5.01 5.13 4.82 4.98 5.16 5.14 4.99 4.88 5.40 5.58 5.35 5.24 5.45 6.08 5.82 6.14 6.29 1.99 2.05 1.93
1.16 1.19 1.43 1.42 1.33 1.38 1.34 1.41 1.17 1.23 1.32 1.32 1.27 1.10 1.32 1.18 1.14 1.40 1.48 1.47
.937
.681
.938
.791
.901
.754
Expresses that coefficients are significant at the level of .01.
.88476 .50932 .13508 .6934 .21636 .1922
Expresses that correlations are significant at the level of .01.
Table 3 shows that all values exceeded the .65 benchmark that literature suggests as acceptable (Steenkamp and Geyskens, 2006).
4.5. Construct validity Construct validity was assessed by considering two types of criteria: convergent and discriminant validity: a. Convergent validity. This shows if the items that compose a determined scale converge on only one construct. We used the Average Variance Extracted (AVE) to contrast the convergent validity and obtained acceptable values greater than .5, which implies that items that compose a determined scale contain less than 50% error variance and converge on only one construct (Fornell and Larcker, 1981), as Table 3 shows. We also checked that the factor loadings of the confirmatory model were statistically significant (level of .01) and higher than .5 points (Steenkamp and Geyskens, 2006). Results showed that all the indicators loaded significantly (po.001) and substantively (all factor loadings went beyond .5) on their proposed constructs, providing evidence of convergent validity of the measures (see Table 4). b. Discriminant validity. This verifies if a determined construct is significantly distinct from other constructs that are not theoretically related to it. To guarantee discriminant validity, correlations between the constructs must differ significantly at the .05 level from 1 (Bagozzi et al., 1991). In this sense, we checked that the value 1 did not appear in the 95% confidence interval of the correlations between the different constructs in any pair of constructs. Results showed an acceptable level of discrimination since all pairs satisfied this criterion (see Table 5). 5. Results
.904
.826
To test Hypotheses 1–3 we developed a structural equation model. Fig. 2 shows the results corresponding to these hypotheses. Results reveal the acceptance of hypotheses 1 and 2 to a level of .01,
128
D. Belanche et al. / Journal of Retailing and Consumer Services 19 (2012) 124–132
intention to use a website is positively affected by customer satisfaction (b ¼.785; po.01; t¼4.516); however the influence of website usability is slightly negative but non-significant (l ¼ .271; p4.05; t¼ 1.719). As a result, hypothesis 1 is confirmed whereas hypothesis 3 is not supported. The non-significant effect of website usability on intention to use a website is especially interesting since in previous works, the role of usability in forming consumer behavioral intentions has been confirmed (e.g. Flavia´n et al., 2006). A first explanation for this non-significant direct effect might be found in multicollinearity problems. To check this we calculated the VIF (variance inflation factor) and contrasted that its value (2.544) was lower than the acceptable limit of 10 points (Neter et al., 1990), so that we have to think in additional explanations rather than multicollinearity. In this respect, this non-significant effect might be due to the fact that usability may not guarantee future intentions to use the website if there are better alternatives available or, following attribution and equity theories (Oliver and DeSarbo, 1988), if the consumer perceive that the website is obtaining unfair benefits from the relationship (i.e. the website is easy to use but the e-retailer is obtaining a great market share because of this). As well, if there are no alternatives available (as it might be in the sector we used to collect data, which is very concentrated), using a given e-service might be the only possibility for consumers even if it exhibits low levels of usability. However, usability could positively affect intention to use a website indirectly through customer satisfaction. In other words, intention to use a website might depend on the development of consumer satisfaction, which emerges as a result of the proper management of usability on the Web site. Thus, although we posit that usability relates to future intention to use the website, its influence might be mediated by consumer satisfaction. Therefore, to examine the possible mediating role of customer satisfaction, we develop four models using structural equation modeling. We first analyze the direct effects of website usability (Fig. 3, Panel A) and satisfaction (Fig. 3, Panel B) on intention to use a website. According to Bloemer and de Ruyter (1998), an initial condition to support a possible mediating effect is the fact
while Hypothesis 3 was not supported. Besides, the model fit showed acceptable values (Chi-square¼ 199.095, 74 d.f., po.001; Satorra–Bentler scaled Chi-square¼129.3661, 74 d.f., p¼.00007; Bentler–Bonett Normed Fit Index¼ .920; Bentler–Bonett Nonnormed Fit Index¼ .955; Comparative Fit Index (CFI)¼.964; Bollen (IFI) Fit Index¼.964; Root Mean Sq. Error of App. (RMSEA)¼.059; 90% Confidence Interval of RMSEA (.042, .076); normed Chi-square¼2.6905). It was also notable that this simple model allow us to partially explain both intention to use a website (R2 ¼.340) and satisfaction (R2 ¼.678), which are two key factors to guarantee website success. To be precise, according to the standardized estimations, we may say first that customer satisfaction with a website is positively influenced by the perceived website usability (l ¼.823; po.01; t¼10.206), confirming hypotheses 2. At the same time, consumer
R2 = .678 .823* t = 10.206
Consumer Satisfaction
Perceived Usability
.785* t = 4.516
-.271(n.s.) t = -1.719
Intention to Use R2 = .340
Note: (*) expresses that coefficients are significant at the level of .01. (n.s.) expresses that coefficients are non-significant.
Fig. 2. Structural equation model: standardized solution.
A
B
(Direct Effect)
(Direct Effect)
.359* t = 4.595 USAB
.558* t = 6.417 I.USE
SAT
I.USE
R2 = .129
R2 = .311
χ2 = 99.799, 34 d.f., p < .05; NFI = .932; NNFI = .951; CFI = .963; IFI = .964; RMSEA = .071
C
χ2 = 58.662, 13 d.f., p < .05; NFI = .944; NNFI = .945; CFI = .966; IFI = .967; RMSEA = .082
(Full Mediating Effect)
.819* t = 10.148 USAB
.551* t = 6.937 SAT
I.USE
R2 = .671 R2 = .304 χ2 = 204.017, 75 d.f., p < .05; NFI = .918; NNFI = .954; CFI = .962; IFI = .962; RMSEA = .060
Note: (*) expresses that coefficients are significant at the level of .01. Fig. 3. Structural equation models.
D. Belanche et al. / Journal of Retailing and Consumer Services 19 (2012) 124–132
129
Table 6 Multisample analysis: perceived risk. Constraints
Estimated coefficients (lower perceived risk)
Estimated coefficients (higher perceived risk)
D.F.
w2 differences
Probability
USAB–SAT
.581a t ¼6.934
.868a t ¼11.053
1
4.047
.044
USAB–I.USE
.199 t ¼ 1.155
.272 t ¼ 1.082
1
.008
.928
FIT INDICES
NFI ¼.871, NNFI¼ .947, CFI¼ .957; IFI ¼.958; RMSEA ¼ .063
NFI ¼ .879, NNFI ¼.952, CFI¼ .961; IFI ¼ .962; RMSEA ¼ .068
a
Expresses that coefficients are significant at the level of .01.
that these two direct effects must be significant. Specifically, results show a positive effect of both usability (l ¼.359; p o.01; t ¼4.595) and satisfaction (l ¼.558; po.01; t¼ 6.417) on intention to use a website. Thus, in the absence of any other determinant factor, the direct effect of usability on intention to use a website is positive. The next step would consist on including satisfaction as a partial mediator; that is, allowing for both direct and indirect effects (i.e., mediated through consumer satisfaction) of usability on intention to use a website (note that this is the model already shown in Fig. 2). In this situation, the direct effect of usability on intention to use a website becomes non-significant (l ¼ .271; p 4.05; t ¼ 1.719) suggesting that, at least, a partial mediation of consumer satisfaction exists. Finally, the model shown in Panel C (Fig. 3) depicts the indirect effect when satisfaction fully mediates the relationship between usability and intention to use a website. Then, this last model was nested within the previous one, and a chi-square (w2) difference test was performed to determine whether consumer satisfaction fully or only partially mediates the effect of usability on intention to use a website (Kulviwat et al., 2009). The test indicates that the partial mediation model provides the best fit to the data (Dw2(1) ¼4.922, p o.05). Thus, we may state that usability exerts an indirect effect on intention to use a website through consumer satisfaction.
Chi-squared when the constraint of equalizing one of the coefficients is eliminated. Thus LM-Test assesses if the elimination of this constraint supposes a significant change in the Chi-squared, and as a consequence a significant improvement in the model fit. Table 6 suggests that there is a significant difference between the groups to a level of .05 for the relationship between perceived usability and consumer satisfaction. To be precise, the effect of perceived usability on satisfaction is strengthened when perceived risk is high, which supports hypothesis 4a. On the contrary, there is no significant difference on the other constraint considered, so that hypothesis 4b is not supported. Indeed, the direct effect of usability on intention to use is found to be nonsignificant in both cases. However, taking into account that: (1) usability also exerts an indirect effect on intention to use through consumer satisfaction and (2) the direct effect of usability on satisfaction is strengthen when perceived risk is high, we may conclude that the indirect effect of usability on intention to use a website may be reinforced for high-risk perceivers. Thus, when perceived risk is high, results suggest that the role of usability in forming behavioral intentions becomes more relevant.
5.1. Multisample analysis
Online stores usually expend significant sums of money to set up and maintain their websites. It is not a trivial question since important performance ratios may depend on users’ judgment about the online establishment and their intention to use this more times. However, nowadays consumers have different skills and abilities regarding the Internet use, so that companies must focus not only on complex and attractive website design but also on developing easy-to-use websites. Although previous literature has noted the importance of website usability, it is necessary to understand the real influence of usability on key variables that may benefit the host-firm and to find under what conditions usability affects in a greater extent to customers’ perceptions. The results of this work suggest that usability has a significant impact on consumer behavior. More specifically, website usability has been first found to directly influence customer satisfaction. Although some authors argue that usability will not necessarily favor a positive impression among users (Chen and Yen, 2004, Lowry et al., 2006), this finding is consistent with most of previous studies (e.g. Kim and Eom, 2002, Casalo´ et al., 2008). Second, this work suggests that website usability also influences the consumer intention to use a website indirectly through consumer satisfaction. This mediating role of consumer satisfaction represents a contribution to the literature since previous authors have proposed either a direct influence of usability on consumer loyalty to a website (e.g. Flavia´n et al., 2006) or a
In order to assess the moderating role of perceived risk (Hypothesis 4) a multisample analysis was performed. To do that, we employed the statistical software EQS version 6.1. More specifically, we divided the total sample into two groups according to their perceived risk in the selected website. Following Garcı´a et al. (2008), total sample was divided into two groups according to the arithmetic mean of the moderating variable. Around this mean we eliminated some cases (71/2 standard deviation). The first group was formed by 107 cases representing consumers that showed a low perceived risk in the selected website. The second group was formed by 90 cases representing consumers with a high perceived risk. The risk mean for the first group (M¼1.07) is significantly lower (t¼ 17.103; po.01) than the mean for the second group (M¼3.26), which supports the creation of these groups. First, multisample analysis generates an individual structural solution for each group. Fit indices for the whole multisample analysis shows acceptable fit levels (NFI¼.859, NNFI¼.943, CFI¼ .950); also fit indices for each subsample (reported in Table 5) are close to the recommended values in spite of the sample size reduction. Second, multisample analysis offers information about the significance of the differences between the coefficients of the two models. To assess these differences we use the LM-Test. This contrast analyzes the variation of the
6. Discussion and managerial implications
130
D. Belanche et al. / Journal of Retailing and Consumer Services 19 (2012) 124–132
non-significant relationship between usability and the attraction of viewers to the website (Chen and Yen, 2004, Lowry et al., 2006). In other words, a logical sequence among the study variables has appeared. This logical sequence implies that intention to use a website might depend on the development of consumer satisfaction, which emerges in part as a result of the proper management of usability on the website. As a consequence, although usability may not be one of the most relevant direct antecedents of intention to use a website, managers should not forget usability when designing their websites since it still influences consumer intentions by means of other variables such as satisfaction. In this situation, the smaller direct effect of usability may be compensated by the indirect effect through satisfaction. In addition, in this work we have analyzed under which conditions the role of usability is especially relevant. In this respect, due to the fact that perceived risk plays an important role in affecting users’ decision of e-services adoption (Hsu and Chiu, 2004), we have concentrated on how risk perceptions might moderate the influence of usability. Indeed, because consumers have different skills in managing and controlling a website, they may perceive different levels of risk when interacting with a website. Our results suggest that the influence of usability is strengthened in those consumers that perceive a great risk, probably because usability helps overcome these consumer fears and form a more favorable opinion of using a website. All these findings suggest that website usability still plays a relevant role in the e-commerce context. 6.1. Managerial implications Thus, these findings allow us to offer some alternatives to improve the levels of customer intention to use a website and, as a result, the retention-rate and profits of the e-business, which are key objectives of most organizations:
First, due to the influence of usability on consumer behavior,
management should not give priority to the design of complex websites full of multi-media effects but concentrate instead on designs and structures that are simple and easy for the user to understand. That is, the most effective website may not be the most sophisticated one, but the most easy to use (Casalo´ et al., 2007). Especially interesting is the role of usability when perceived risk is high. Indeed, companies must ensure great levels of website usability in those cases that consumers perceived as especially uncertain. Among others, when a company is not wellknown or when credit card is required to pay an online purchase, managers should prioritize usability in website design. For this goal, companies should take into account aspects such as: anticipation (i.e. websites must be designed according to the potential visitors’ needs and requirements), consistency (websites must be consistent with consumers’ knowledge and skills), reversibility (i.e. websites must allow consumers to undo all the tasks), legibility, efficiency (i.e. simplifying processes and minimizing errors), visibility (i.e. hidden menus are not recommended), adequacy of information (i.e. too much information might be counterproductive) or provision of feedback (this will reduce consumers’ uncertainty regarding the security of the purchasing processes) among others. Second, due to the relevance of consumer satisfaction in forming behavioral intentions, companies should try to maximize the satisfaction of their customers during their interactions through the companies’ websites. To be precise, customer satisfaction will be generated if the customer’s expectations about the relationship are met (e.g. Oliver, 1980). Therefore, companies should try to identify the needs of their online customers (not only in terms of website usability and design but also in terms of services offered, etc.) in order to offer them what they require in an efficient way.
6.2. Limitations and future research Despite the contributions of this work, we acknowledge several limitations of this research that also offer some possibilities for further research. First, in this work we have only considered consumer perceptions regarding a specific website (the one of the retailer that collaborated in the research), which means that extrapolating the findings to other websites requires great care. As we have mentioned in the data collection section, data was collected among the users of a bus ticket e-selling service offered by a wellknown transport company in Spain, which might reduce risk perceptions (the mean of perceived risk is quite low as can be seen in Table 3). The election of a well-known retailer was based on data collection convenience; however, this might represent a limitation of the study that opens new research avenues. For instance, it would be interesting to investigate if the role of perceived risk is different when retailers are unknown, or analyze possible differences in the role of risk when considering multichannel retailers versus those operating exclusively online. As well, the sample consists of only Spanish-speaking consumers, so that an interesting route to extend this research would be to repeat the study using a wider sample of customers (for example in terms of different cultures). This effort might help firms develop websites more efficiently depending on where their potential consumers’ come from. In addition, it would be useful to extend this work by analyzing the influence of usability on other key variables that determine consumer behavior depending on the product/service category provided. For instance, it is possible that greater levels of usability may be also associated to greater levels of perceived security in the website, increasing consumer trust as well. This analysis would help to clarify in more detail the relevance of website usability for managers in different sectors. Finally, it would be also a good idea to analyze the effects of website usability on consumer behavior when the consumer accesses the Internet by new methods (e.g. mobile phone, PDA, etc.). This would serve to understand the role of usability when individuals and companies interact by means of these new technologies.
Acknowledgments The authors are grateful for the financial support received from Spanish Ministry of Education (ECO2009-10157).
Appendix
USABILITY (USAB) USAB1 In this website everything is easy to understand USAB2 This website is simple to use, even when using it for the first time USAB3 It is easy to find the information I need from this website USAB4 The structure and contents of this website are easy to understand USAB5 It is easy to move within this website USAB6 The organization of the contents of this site makes it easy for me to know where I am when navigating it USAB7 When I am navigating this site, I feel that I am in control of what I can do SATISFACTION (SAT) SAT1 I think I made the correct decision to use this website SAT2 The experience that I have had with this website has been satisfactory
D. Belanche et al. / Journal of Retailing and Consumer Services 19 (2012) 124–132
SAT3
In general, I am satisfied with the way that this website has managed interactions SAT4 In general, I am satisfied with the service I have received from the website INTENTION TO USE (I_USE) I_USE1 I have the intention to use this website again in the near future I_USE2 The likelihood of using this website again is high I_USE3 If I had to use this website again, I will use it without any doubt PERCEIVED RISK (RISK) RISK1 Using this website causes me to be concerned with experiencing some kind of loss in the future RISK2 I am vulnerable to the actions conducted by this website RISK3 The actions conducted by this website may cause problems and uncertain consequences for me. Notes: Item in italics were rejected during the scale validation process. The scales were presented in Spanish due to respondent’s nationality.
References Abdeldayem, M.M., 2010. A study of customer satisfaction with online shopping: evidence from the UAE. International Journal of Advanced Media and Communication 4 (3), 235–257. AIMC—Asociacio´n para la investigacio´n de los medios de comunicacio´n, 2009. Audiencia en Internet segu´n el EGM febrero/marzo de 2009. Available: /http://www.aimc.esS (Accessed 19 October 2010). Anderson, E., Sullivan, M., 1993. The antecedents and consequences of customer satisfaction for firms. Marketing Science 12 (2), 125–143. Anderson, J., Gerbing, D., 1988. Structural equation modeling in practice: a review and recommended two-step approach. Psychological Bulletin 103 (3), 411–423. Bagozzi, R., Yi, Y., Phillips, L., 1991. Assessing construct validity in organizational research. Administrative Science Quarterly 36 (3), 421–458. Becker, S., Mottay, F., 2001. A global perspective on web site usability. IEEE Software 18 (1), 54–61. Bhattacherjee, A., 2001. Understanding information systems continuance: an expectation-confirmation model. MIS Quarterly 25 (3), 351–370. Bloemer, J., de Ruyter, K., 1998. On the relationship between store image, store satisfaction and store loyalty. European Journal of Marketing 32 (5/6), 499–513. Bridges, E., Florsheim, R., 2008. Hedonic and utilitarian shopping goals: the online experience. Journal of Business Research 61 (4), 309–314. Brockman, B., 1998. The influence of affective state on satisfaction ratings. Journal of Consumer Satisfaction, Dissatisfaction and Complaining Behavior 11, 40–50. Casalo´, L., Flavia´n, C., Guinalı´u, M., 2007. The role of usability and satisfaction in the consumer’s commitment to a financial services website. International Journal of Electronic Finance 2 (1), 31–49. Casalo´, L., Flavia´n, C., Guinalı´u, M., 2008. The role of satisfaction and website usability in developing customer loyalty and positive word-of-mouth in the e-banking services. International Journal of Bank Marketing 26 (6), 399–417. Cronbach, L., 1970. Essentials of Psychological Testing. Harper and Row, New York, NY. Cronin, J., Brady, M., Hult, G., 2000. Assessing the effects of quality, value, and customer satisfaction on consumer behavioral intentions in service environments. Journal of Retailing 76 (2), 193–218. Chen, K., Yen, D., 2004. Improving the quality of online presence through interactivity. Information & Management 42 (1), 217–226. Chiu, C.M., Lin, H.Y., Sun, S.Y., Hsu, M.Y., 2009. Understanding customers’ loyalty intentions toward online shopping: An integration of Technology Acceptance Model and Fairness Theory. Behavior and Information Technology 28 (4), 347–360. Davis, F., 1989. Perceived usefulness, perceived ease of use, and user acceptance of information technology. MIS Quarterly 13 (3), 319–340. De Ruyter, D., Wetzels, M., Kleijnen, M., 2001. Customer adoption of e-service: an experimental study. International Journal of Service Industry Management 12 (2), 184–207. De Vaus, D., 2001. Surveys in Social Research, 5th edition Social Research Today. Devaraj, S., Fan, M., Kohli, R., 2002. Antecedents of B2C channel satisfaction and preference: validating e-commerce metrics. Informaton Systems Research 13 (3), 316–333. Flavia´n, C., Guinalı´u, M., Gurrea, R., 2006. The role played by perceived usability, satisfaction and consumer trust on website loyalty. Information & Management 43 (1), 1–14. Fornell, C., Larcker, D., 1981. Evaluating structural equation models with unobservable variables and measurement error. Journal of Marketing Research 18, 39–50.
131
Garcı´a, N., Sanzo, M., Trespalacios, J., 2008. Can a good organizational climate compensate for a lack of top management commitment to new product development? Journal of Business Research 61 (2), 118–131. Guinalı´u, M., 2005. La gestio´n de la confianza en Internet. Un factor clave para el desarrollo de la economı´a digital. Doctoral Thesis. University of Zaragoza, Spain. Hair, J., Anderson, R., Tatham, R., Black, W., 1998. Multivariate analysis. Prentice Hall International, Englewood. Hsu, M., Chiu, C., 2004. Predicting electronic service continuance with a decomposed theory of planned behaviour. Behaviour and information Technology 23 (5), 359–373. Jacoby, J., Kaplan, L., 1972. The components of perceived risk. In: Venkatesan, M. (Ed.), Proceedings of the Third Annual Conference, Association for Consumer Research, Champaign, IL, pp. 382–393. Janda, S., Trocchia, P., Gwinner, K., 2002. Consumer perceptions of Internet retail service quality. International Journal of Service Industry Management 13 (5), 412–431. ¨ Joreskog, K., 1971. Statistical analysis of sets of congeneric tests. Psychometrika 36 (2), 109–133. ¨ ¨ Joreskog, K., Sorbom, D., 1993. LISREL 8: Structural Equation Modeling with the SIMPLIS Command Language. Scientific Software International, Chicago, IL. Kim, J., Jin, B., Swinney, J.L., 2009. The role of etail quality, e-satisfaction and e-trust in online loyalty development process. Journal of Retailing and Consumer Services 16 (4), 239–247. Kim, E., Eom, S., 2002. Designing effective cyber store user interface. Industrial Management and Data Systems 102 (5), 241–251. Kulviwat, S., Bruner, G., Al-Shuridah, O., 2009. The role of social influence on adoption of high tech innovations: the moderating effect of public/private consumption. Journal of Business Research 62 (7), 706–712. Ladhari, R., 2010. Developing e-service quality scales: a literature review. Journal of Retailing and Consumer Services 17 (6), 464–477. Lee, M., Turban, E., 2001. A trust model for consumer Internet shopping. International Journal of Electronic Commerce 6 (1), 75–91. Liao, C., Chen, J.L., Yen, D.C., 2007. Theory of Planned Behavior (TPB) and customer satisfaction in the continued use of e-service: an integrative model. Computers in Human Behavior 23 (6), 2804–2822. Lichtenstein, D.R., Netemeyer, R.G., Burton, S., 1990. Distinguishing coupon proneness from value consciousness: an acquisition—transaction utility theory perspective. Journal of Marketing 54, 54–67. Lin, C., Wu, S., Tsai, R., 2005. Integrating perceived playfulness into expectationconfirmation model for web portal context. Information & Management 42 (5), 683–693. Liu, I., Chen, M.C., Sun, Y.S., Wible, D., Kuo, C.H., 2010. Extending the TAM model to explore the factors that affect intention to use an online learning community. Computers & Education 54 (2), 600–610. Lowry, P., Spaulding, T., Wells, T., Moody, G., Moffit, K., Madariaga, S., 2006. A theoretical model and empirical results linking website interactivity and usability satisfaction. In: Proceedings of the 39th Hawaii International Conference on System Sciences, Hawaii. Maditinos, D.I., Theodoridis, K., 2010. Satisfaction determinants in the Greek online shopping context. Information Technology and People 23 (4), 312–329. Neter, J., Wasserman, W., Kutner, M.H., 1990. Applied Linear Statistical Models. Irwin, Boston, MA. Oliver, R., 1980. A cognitive model of the antecedents and consequences of satisfaction decisions. Journal of Marketing Research 17 (4), 460–469. Oliver, R., 1999. Whence consumer loyalty? Journal of Marketing 63, 33–44. Oliver, R., DeSarbo, W., 1988. Response determinants in satisfaction judgments. Journal of Consumer Research 14 (4), 495. Oneupweb, 2010. Revolutionizing Website Design. Available: /http://www.oneup web.comS [Accessed 19 October 2010]. RED.ES, 2009. Perfil sociodemogra´fico de los internautas: Ana´lisis de datos INE 2008. Available: /http://observatorio.red.es/hogares-ciudadanos/articles/id/ 3027/perfil-sociodemografico-los-internautas-analisis-datos-ine-2008.htmlS (Accessed 19 October 2010). Roy, M., Dewit, O., Aubert, B., 2001. The impact of interface usability on trust in web retailers. Internet Research: Electronic Networking Applications and Policy 11 (5), 388–398. Schlosser, A.E., White, T.B., Lloyd, S.M., 2006. Converting web site visitors into buyers: how web site investment increases consumer trusting beliefs and online purchase intentions. Journal of Marketing 70, 133–148. Severt, D., 2002. The customer’s path to loyalty: a partial test of the relationships of prior experience, justice, and customer satisfaction. Doctoral Thesis. Faculty of the Virginia Polytechnic Institute and State University. Available: /http:// en.scientificcommons.org/1527127S (Accessed 19 October 2010). Shankar, V.S., Smith, A.K., Rangaswamy, A., 2003. Customer satisfaction and loyalty in online and offline environments. International Journal of Research in Marketing 20 (2), 153–175. Smith, J., Barclay, D., 1997. The effects of organizational differences and trust on the effectiveness of selling partner relationships. Journal of Marketing 61 (1), 3–21. Smith, K., 1974. On estimating the reliability of composite indexes through factor analysis. Sociological Methods & Research 2 (4), 485–510. Steenkamp, J., Geyskens, I., 2006. How country characteristics affect the perceived value of Web sites. Journal of Marketing 70 (3), 136–150. Stone, R., Grønhaug, K., 1993. Perceived risk: Further considerations for the marketing decision. European Journal of Marketing 27 (3), 39–50.
132
D. Belanche et al. / Journal of Retailing and Consumer Services 19 (2012) 124–132
Szymanski, D.M., Hise, R.T., 2000. E-satisfaction: an initial examination. Journal of Retailing 76 (3), 309–322. Wu, J., Chang, Y., 2005. Towards understanding members’ interactivity, trust, and flow in online travel community. Industrial Management and Data Systems 105 (7), 937–954.
Wu, I.L., Chen, J.L., 2005. An extension of trust and TAM model with TPB in the initial adoption of on-line tax: an empirical study. International Journal of Human–Computer Studies 62, 784–808. Zaichkowsky, J., 1985. Measuring the involvement construct. Journal of Consumer Research 12 (3), 341–352.
Contents lists available at SciVerse ScienceDirect
Journal of Retailing and Consumer Services journal homepage: www.elsevier.com/locate/jretconser
Website usability, consumer satisfaction and the intention to use a website: The moderating effect of perceived risk Daniel Belanche a,1, Luis V. Casalo´ b, 2, Miguel Guinalı´u a,n a b
´n de Marketing e Investigacio ´n de Mercados, Facultad de Economı´a y Empresa, University of Zaragoza, Gran Vı´a 2, 50005 Zaragoza, Spain Departamento de Direccio ´n de Marketing e Investigacio ´n de Mercados, Facultad de Empresa y Gestio ´n Pu ´ blica, University of Zaragoza, Ronda Misericordia 1, 22001 Huesca, Spain Departamento de Direccio
a r t i c l e i n f o
a b s t r a c t
Available online 12 November 2011
This paper analyzes the influence of website usability on both consumers’ satisfaction and intention to use a website, as well as the impact of satisfaction on usage intentions. Additionally, we study the moderating effect that consumer risk perceptions may have on the influence of website usability. Results show that website usability affects satisfaction which in turn affects intention to use. Contrary to expected, usability does not directly affect intention to use but has an indirect effect through consumer satisfaction. Finally, the usability effect on consumer satisfaction is moderated by perceived risk. & 2011 Elsevier Ltd. All rights reserved.
Keywords: Usability Satisfaction Intention to use Perceived risk Online consumer behavior
1. Introduction According to Oneupweb (2010), in the context of online transactions, online users’ mainly expect websites that facilitate their purchases. In this line, 95.5% of users expect pricing and shipping information to be clearly presented, so usability must be an essential component of e-retail strategy. Website usability can be defined considering the following aspects (Flavia´n et al., 2006): (a) the ease of understanding the structure of a website, its functions, interface and the contents that can be observed by the user; (b) simplicity of use of the website in its initial stages; (c) the speed with which the users can find what they are looking for; (d) the perceived ease of site navigation in terms of time required and action necessary in order to obtain the desired results; and (e) the ability of the user to control what they are doing, and where they are, at any given moment. Analogously to merchandising in offline stores, when a customer accesses to an online store, usability issues may affect customer’s perceptions and behaviors. Thus, more usable websites tend to create more positive attitudes toward online stores and increase conversion rates, whereas less usable websites have the opposite effect (Becker and Mottay, 2001). Considering the influence of usability on website performance, online retailers require a full understanding of this variable. Previous
n
Corresponding author. Tel.: þ34 976761000x4695; fax: þ 34 976761667. E-mail addresses: [email protected] (D. Belanche), [email protected] (L.V. Casalo´), [email protected] (M. Guinalı´u). 1 Tel.: þ34 976761000x4636; fax: þ 34 976761667. 2 Tel.: þ34 976761000x4695; fax: þ 34 976761667. 0969-6989/$ - see front matter & 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.jretconser.2011.11.001
literature in the field of retailing has studied the concept of usability, especially as a component of quality of service (Ladhari, 2010). However, the integration of usability management in the e-business strategy and the role of perceived risk are outstanding issues. This research allows us to understand the relevance of usability on strategic indicators of companies’ success analyzing the influence of perceived usability on the levels of consumer satisfaction and intention to use a website. Moreover, although the Internet shopping has been proposed to involve more risk than traditional shopping (Lee and Turban, 2001), different consumers may perceive different levels of risk in the same online environment. Thus, this research contributes to the literature analyzing whether risk moderates the influence of perceived usability on consumer satisfaction and intention to use a website. Bearing these considerations in mind, we structure the remainder of this article as follows. In the following section we formalize the working hypotheses. Next, data collection and measure validation processes are explained, followed by the results of the analyses. Finally, we discuss the main conclusions, managerial implications, and limitations of the study, as well as some possibilities for future research.
2. Hypotheses formulation Traditionally, several authors have stated that satisfaction – defined as an affective condition that results from a global evaluation of all the aspects that make up a relationship – is a crucial antecedent of re-purchase and re-usage intentions of a product or service (e.g. Oliver, 1980; Anderson and Sullivan, 1993). Focusing on the new technologies context, satisfaction
D. Belanche et al. / Journal of Retailing and Consumer Services 19 (2012) 124–132
with prior use is also found to be the strongest predictor of users’ continuance intentions (Bhattacherjee, 2001). In the same line, Devaraj et al. (2002) measured customer satisfaction in the e-commerce context and supported empirically that satisfaction is a key determinant of customer channel preference. More recently, researchers have found that satisfaction positively affects loyalty intentions toward online shopping (Chiu et al., 2009) and boosts the use of Internet portals (Lin et al., 2005), e-services (Liao et al., 2007) and online communities as well (Liu et al., 2010). Broadly speaking, literature supports that satisfied consumers exhibit a greater intention to use firm’s products, have a greater re-purchase intention, favor positive word-of-mouth and have a lower to look for alternative providers (Oliver, 1999; Kim et al., 2009). Consequently, it is expected that once users achieve certain levels of satisfaction with prior use of a website, the perceptions of satisfaction will influence their re-usage intentions. Thus, we propose our first hypothesis: H1. Consumer satisfaction has a positive effect on consumer intention to use a website. In one of the first works focusing on the determinants of satisfaction in online exchanges, Szymanski and Hise (2000) noted the ability of website design to promote satisfactory purchase experiences. This result has been supported by several authors afterwards. For example, Kim and Eom (2002) concluded that usability is of critical importance in achieving user satisfaction. More recently, Maditinos and Theodoridis (2010) pointed out that both quality of the interface and of the information provided to consumers (two key aspects of website usability) have a significant effect on the levels of users’ satisfaction. We must also note that making the purchase easier is one of the main motivations of consumers’ online purchasing (Bridges and Florsheim, 2008). Therefore, when visiting a website, consumers expect to find a channel whose features facilitate search, selection, payment and post-purchase actions. Consequently, it is reasonable to say that satisfaction with the purchase experience depend on website ease of use (Shankar et al., 2003). Complex and not intuitive interfaces, long purchasing processes, non-updated information or non-relevant information create an interaction atmosphere that negatively affects consumer satisfaction. According to all these ideas, we propose the following hypothesis: H2. Perceived website usability has a positive effect on consumer satisfaction. However, literature on online consumer behavior found that usability impact is even more relevant since it not only affects satisfaction but also favor future purchase intentions. In this respect, Flavia´n et al. (2006) found that usability positively affects consumer loyalty to a website. As well, Abdeldayem (2010) noted that attitudes toward online shopping and the intention to shop online are affected by ease of use. Therefore, we propose in our third hypothesis that: H3. Perceived website usability has a positive effect on consumer intention to use the website. 2.1. Moderating effect of perceived risk Focusing on the online context, the importance of risk perceptions is evident when conducting an e-purchase. Compared to traditional shopping, online commercial establishments are less known to consumers (Lee and Turban, 2001), and the absence of face-to-face interaction has introduced more uncertainty and risk (Wu and Chen, 2005). Apart from risk associated with intangibility and time lag, customers have to release personal and
125
financial data to not always well known providers (De Ruyter et al., 2001). Moreover, the outcome of this process depends on the behavior of the e-marketer and this is not within the consumer’s control (Lee and Turban, 2001). Therefore, perceived risk is a relevant parameter, especially in the early stages of a process of customer adoption of online purchase of products and services (De Ruyter et al., 2001; Hsu and Chiu, 2004). As a result of numerous online frauds and piracy crimes in the network, users need more and more signs to make sure that financial and personal information is under their control and that privacy and security are guaranteed. In doing so, usability could be understood as a sign of competence, ability and honesty by the organization hosting the website. Firms often use signals to communicate the level of some unobservable quality, especially to reduce the information asymmetry between buyers and sellers (Schlosser et al., 2006). Since an online transaction is usually unobservable by consumers before purchase (Schlosser et al., 2006), website usability might serve to demonstrate that the company is able to provide high-quality services and understands and is sensitive to online consumers’ fears. Besides, consumer perceptions about their own control of the online transaction process could be increased by perceived usability (Casalo´ et al., 2007), since greater usability is associated to low levels of difficulty to manage a system (Davis, 1989). That is, users perceptions of their own skills depend not only on their experience but also on the usability of the website (Flavia´n et al., 2006). In those cases in which consumers perceive high risk in an online transaction, website usability may help reduce this uncertainty and thus its effect in forming consumer satisfaction and intentions to use a website may be reinforced. To be precise, consumers perceiving high risk would more likely try to eliminate any suspicion when interacting with the website. Usability communicates information about performance, thus consumers might infer that a company that has invested in usability and website design can successfully handle online transactions (Schlosser et al., 2006). Thus, in a high-risk situation consumers will be more prone to carefully analyze all the information about the website and, as a result, website usability may help overcome these fears and form a more favorable opinion of using a website. In turn, when perceived risk is low, consumers will be less influenced by usability perceptions since they do not need these reinforcements to overcome worry about the possibility of nondesired outcomes. Taking into account the previous considerations we propose in the following hypothesis a moderating role of risk in the effect of usability on both satisfaction and intention to use. H4. If perceived risk increases, the relationship between perceived website usability and: (a) consumer satisfaction and (b) consumer intention to use the website will be strengthened. To sum up, the research model can be seen in Fig. 1.
3. Data collection Data were obtained though a web survey targeted to the users of a Spanish online retailer, which helped us to recruit participants. This practice is consistent with common online market research (e.g. Steenkamp and Geyskens, 2006). To be precise, data was collected among the users of a bus ticket e-selling service offered by one of the most important transport companies in Spain. It is important to note that only few companies operate online in this sector in Spain. In order to measure the variables, a structured questionnaire containing closed questions was developed. This questionnaire was designed to gather information about the studied constructs
126
D. Belanche et al. / Journal of Retailing and Consumer Services 19 (2012) 124–132
Consumer Satisfaction
H2: + Perceived Usability
H1: +
H3: + H4a
Intention to Use H4b
Perceived Risk Note: “- - >” reflects hypotheses affected by moderation Fig. 1. Research model.
Table 1 Representative nature of the data collected.
Sample size Age (years) o24 25–34 35–44 444 Gender Males Females
Current research project
AIMC (2009)
RED.ES (2009)
214
30,705
19,131
28.5% 27.6% 18.2% 25.7%
24.4% 28.5% 21.8% 25.2%
20.3% 30.5% 24.7% 24.5%
52.3% 47.7%
54.6% 45.4%
52.7% 47.3%
and other data in order to provide a more detailed insight of the socio-demographic characteristics of the user. Specifically, the consumer was asked to indicate his/her level of agreement or disagreement with a series of statements about the perceived usability and risk3 in the retailer’s website as well as their levels of satisfaction and intention to use the website (see Appendix). All questions related to construct used 7-point Likert scale. Finally, this method of collecting data, which relies on volunteer sampling, generated 214 valid questionnaires (atypical cases, repeated responses and incomplete questionnaires were removed). However, because we cannot statistically assess the reliability or possible bias associated with this non-random sample, we compare our sample characteristics with available information about a wider population, that is, the socio-demographic characteristics of the respondents to the largest studies of the online Spanish-speaking population (AIMC, 2009; RED.ES, 2009). The results appear very similar in terms of the age and sex of the respondents (see Table 1), which supports the representative nature of our sample.
4. Measures validation 4.1. Content and face validity An initial set of items was proposed once the relevant literature was revised. This review guaranteed the content validity of the 3 In order to measure risk perceptions, Jacoby and Kaplan (1972) distinguished between five risk dimensions in the overall construct – financial, performance, psychological, physical and social risk – although not every transaction involve all kind of risks. Consequently, in this work we decide to adapt a general measure of overall risk (Stone and Grønhaug, 1993).
measurements instruments, or the degree to which items correctly represent the theoretical content of the construct. To be precise, our measures were inspired by previous scales regarding perceived website usability (Flavia´n et al., 2006; Roy et al., 2001), customer satisfaction (Brockman, 1998; Severt, 2002; Janda et al., 2002; Smith and Barclay, 1997; Guinalı´u, 2005), intention to use (Wu and Chang, 2005; Cronin et al., 2000) and overall perceived risk (Stone and Grønhaug, 1993). These previous scales had been already proposed on either a service or online context, so that adaption to our specific context becomes natural. More efforts were needed to adapt the overall perceived risk scale to the online context since the original scale was tested in offline shopping. Due to this adaptation, face validity (the degree that respondents judge that the items are appropriate to the targeted construct) was also tested through a variation of the Zaichkowsky method (1985), whereby each item is qualified by a panel of experts (a total of 8 people who are professionals in different disciplines such as marketing, sociology, new technologies and Internet) as ‘‘clearly representative’’, ‘‘somewhat representative’’ or ‘‘not representative’’ of the construct of interest. In line with Lichtenstein et al. (1990) an item was retained if a high level of consensus was observed among the experts.
4.2. Exploratory analysis of reliability and dimensionality The validation process started with an initial exploratory analysis of reliability and dimensionality (Anderson and Gerbing, 1988). Cronbach’s alpha indicator was used to assess the initial reliability of the scales, considering a minimum value of .7 (e.g. Cronbach, 1970). The item-total correlation was used to improve the levels of Cronbach’s alpha, considering a minimum value of .3 (De Vaus, 2001). For these initial tasks we used statistical software SPSS v.14.0. All items were adjusted to the required levels. Second, we proceeded to evaluate the unidimensionality of the scales proposed by carrying out a principal components analysis. Factor extraction was based on the existence of eigenvalues higher than 1. Moreover, it was required that factorial loadings were higher than .5 points and a significant total explained variance (Hair et al., 1998). Construct validity was supported by principal component analysis with varimax rotation (see Table 2). After conducting the exploratory factor analysis using all items together, only one factor
Table 2 Principal components analysis with varimax rotation. Item
USAB1 USAB2 USAB3 USAB4 USAB5 USAB6 USAB7 SAT1 SAT2 SAT3 SAT4 IUSE1 IUSE2 IUSE3 RISK1 RISK2 RISK3
Construct loadings Factor 1 Usability
Factor 2 Int. to use
Factor 3 Satisfaction
Factor 4 Risk
.768 .830 .824 .864 .839 .761 .733 .411 .547 .513 .506 .076 .157 .240 .066 .125 .121
.115 .172 .160 .085 .120 .066 .086 .333 .269 .230 .252 .896 .914 .752 .071 .158 .147
.337 .132 .199 .186 .274 .226 .304 .699 .664 .667 .755 .124 .165 .336 .203 .078 .050
.070 .060 .035 .130 .117 .181 .203 .133 .196 .222 .138 .092 .140 .219 .788 .889 .918
D. Belanche et al. / Journal of Retailing and Consumer Services 19 (2012) 124–132
was extracted from each scale corresponding to: usability, satisfaction, perceived risk and intention to use.
127
Table 4 Convergent validity. Item
Loading
t-value
Item
Loading
t-value
USAB1 USAB2 USAB3 USAB4 USAB5 USAB6 USAB7
.826a .790a .822a .872a .890a .782a .786a
14.051 15.308 14.993 14.778 15.800 12.054 13.363
SAT1 SAT2 SAT3 SAT4 IUSE1 IUSE2 IUSE3 RISK2 RISK3
.817a .912a .882a .941a .819a .956a .823a .933a .884a
12.361 15.696 14.563 16.334 13.086 12.964 8.147 12.352 10.752
4.3. Confirmatory analysis of dimensionality With the aim of confirming the dimensional structure of the scales, we used the Confirmatory Factor Analysis. That is, we included all individual-level constructs in a single confirmatory factor model. In addition, this process also allows for a stringent test of convergent and discriminatory validity (Steenkamp and Geyskens, 2006). We employed the statistical software EQS v.6.1. As an estimation method we chose Robust Maximum Likelihood, since it affords more security in samples which might not present ¨ multivariate normality. The criteria proposed by Joreskog and ¨ Sorbom (1993) were followed:
The weak convergence criterion, which supposes eliminating
indicators that do not show significant factor regression coefficients. The strong convergence criterion, which involves eliminating non-substantial indicators (those indicators whose standardized coefficients are lower than .5). ¨ ¨ According to the suggestion of Joreskog and Sorbom (1993), we also eliminated the indicators that least contributes to the explanation of the model, taking R2 o.3 as a cut-off point.
a
Table 5 Discriminant validity. Pair of constructs
Correlation
Standard deviation
95% confidence interval
USAB–SAT USAB–IUSE USAB–RISK SAT–IUSO SAT–RISK IUSE–RISK
.824a .378a .286a .566a .381a .349a
.031 .067 .077 .065 .084 .080
.76324 .24668 .43692 .4386 .54564 .5058
a
Following these recommendations, we eliminated the item RISK1 (see Appendix) and we finally obtained acceptable levels of convergence, R2 (see Table 4) and model fit (Chi-square¼245.789, 98 d.f., po.001; Satorra–Bentler scaled Chi-square¼160.565, 98 d.f., p ¼.00007; Bentler–Bonett Normed Fit Index¼.912; Bentler– Bonett Nonnormed Fit Index¼.955; Comparative Fit Index (CFI)¼ .963; Bollen (IFI) Fit Index¼.964; Root Mean Sq. Error of App. (RMSEA) ¼.055; 90% Confidence Interval of RMSEA (.039; .069); normed Chi-square¼2.5081). 4.4. Composite reliability Additionally, we used the composite reliability indicator to asses construct reliability. Although Cronbach’s alpha indicator is the most frequent test to assess reliability, some authors consider that it underestimates reliability (e.g. Smith, 1974), so that the use of ¨ composite reliability (rc) has been suggested (Joreskog, 1971). Table 3 Descriptive statistics, construct reliability and convergent validity. Construct
Mean
S.D.
rc
AVE
Usability USAB1 USAB2 USAB3 USAB4 USAB5 USAB6 USAB7 Satisfaction SAT1 SAT2 SAT3 SAT4 Intention to use IUSE1 IUSE2 IUSE3 Risk RISK2 RISK3
5.01 5.13 4.82 4.98 5.16 5.14 4.99 4.88 5.40 5.58 5.35 5.24 5.45 6.08 5.82 6.14 6.29 1.99 2.05 1.93
1.16 1.19 1.43 1.42 1.33 1.38 1.34 1.41 1.17 1.23 1.32 1.32 1.27 1.10 1.32 1.18 1.14 1.40 1.48 1.47
.937
.681
.938
.791
.901
.754
Expresses that coefficients are significant at the level of .01.
.88476 .50932 .13508 .6934 .21636 .1922
Expresses that correlations are significant at the level of .01.
Table 3 shows that all values exceeded the .65 benchmark that literature suggests as acceptable (Steenkamp and Geyskens, 2006).
4.5. Construct validity Construct validity was assessed by considering two types of criteria: convergent and discriminant validity: a. Convergent validity. This shows if the items that compose a determined scale converge on only one construct. We used the Average Variance Extracted (AVE) to contrast the convergent validity and obtained acceptable values greater than .5, which implies that items that compose a determined scale contain less than 50% error variance and converge on only one construct (Fornell and Larcker, 1981), as Table 3 shows. We also checked that the factor loadings of the confirmatory model were statistically significant (level of .01) and higher than .5 points (Steenkamp and Geyskens, 2006). Results showed that all the indicators loaded significantly (po.001) and substantively (all factor loadings went beyond .5) on their proposed constructs, providing evidence of convergent validity of the measures (see Table 4). b. Discriminant validity. This verifies if a determined construct is significantly distinct from other constructs that are not theoretically related to it. To guarantee discriminant validity, correlations between the constructs must differ significantly at the .05 level from 1 (Bagozzi et al., 1991). In this sense, we checked that the value 1 did not appear in the 95% confidence interval of the correlations between the different constructs in any pair of constructs. Results showed an acceptable level of discrimination since all pairs satisfied this criterion (see Table 5). 5. Results
.904
.826
To test Hypotheses 1–3 we developed a structural equation model. Fig. 2 shows the results corresponding to these hypotheses. Results reveal the acceptance of hypotheses 1 and 2 to a level of .01,
128
D. Belanche et al. / Journal of Retailing and Consumer Services 19 (2012) 124–132
intention to use a website is positively affected by customer satisfaction (b ¼.785; po.01; t¼4.516); however the influence of website usability is slightly negative but non-significant (l ¼ .271; p4.05; t¼ 1.719). As a result, hypothesis 1 is confirmed whereas hypothesis 3 is not supported. The non-significant effect of website usability on intention to use a website is especially interesting since in previous works, the role of usability in forming consumer behavioral intentions has been confirmed (e.g. Flavia´n et al., 2006). A first explanation for this non-significant direct effect might be found in multicollinearity problems. To check this we calculated the VIF (variance inflation factor) and contrasted that its value (2.544) was lower than the acceptable limit of 10 points (Neter et al., 1990), so that we have to think in additional explanations rather than multicollinearity. In this respect, this non-significant effect might be due to the fact that usability may not guarantee future intentions to use the website if there are better alternatives available or, following attribution and equity theories (Oliver and DeSarbo, 1988), if the consumer perceive that the website is obtaining unfair benefits from the relationship (i.e. the website is easy to use but the e-retailer is obtaining a great market share because of this). As well, if there are no alternatives available (as it might be in the sector we used to collect data, which is very concentrated), using a given e-service might be the only possibility for consumers even if it exhibits low levels of usability. However, usability could positively affect intention to use a website indirectly through customer satisfaction. In other words, intention to use a website might depend on the development of consumer satisfaction, which emerges as a result of the proper management of usability on the Web site. Thus, although we posit that usability relates to future intention to use the website, its influence might be mediated by consumer satisfaction. Therefore, to examine the possible mediating role of customer satisfaction, we develop four models using structural equation modeling. We first analyze the direct effects of website usability (Fig. 3, Panel A) and satisfaction (Fig. 3, Panel B) on intention to use a website. According to Bloemer and de Ruyter (1998), an initial condition to support a possible mediating effect is the fact
while Hypothesis 3 was not supported. Besides, the model fit showed acceptable values (Chi-square¼ 199.095, 74 d.f., po.001; Satorra–Bentler scaled Chi-square¼129.3661, 74 d.f., p¼.00007; Bentler–Bonett Normed Fit Index¼ .920; Bentler–Bonett Nonnormed Fit Index¼ .955; Comparative Fit Index (CFI)¼.964; Bollen (IFI) Fit Index¼.964; Root Mean Sq. Error of App. (RMSEA)¼.059; 90% Confidence Interval of RMSEA (.042, .076); normed Chi-square¼2.6905). It was also notable that this simple model allow us to partially explain both intention to use a website (R2 ¼.340) and satisfaction (R2 ¼.678), which are two key factors to guarantee website success. To be precise, according to the standardized estimations, we may say first that customer satisfaction with a website is positively influenced by the perceived website usability (l ¼.823; po.01; t¼10.206), confirming hypotheses 2. At the same time, consumer
R2 = .678 .823* t = 10.206
Consumer Satisfaction
Perceived Usability
.785* t = 4.516
-.271(n.s.) t = -1.719
Intention to Use R2 = .340
Note: (*) expresses that coefficients are significant at the level of .01. (n.s.) expresses that coefficients are non-significant.
Fig. 2. Structural equation model: standardized solution.
A
B
(Direct Effect)
(Direct Effect)
.359* t = 4.595 USAB
.558* t = 6.417 I.USE
SAT
I.USE
R2 = .129
R2 = .311
χ2 = 99.799, 34 d.f., p < .05; NFI = .932; NNFI = .951; CFI = .963; IFI = .964; RMSEA = .071
C
χ2 = 58.662, 13 d.f., p < .05; NFI = .944; NNFI = .945; CFI = .966; IFI = .967; RMSEA = .082
(Full Mediating Effect)
.819* t = 10.148 USAB
.551* t = 6.937 SAT
I.USE
R2 = .671 R2 = .304 χ2 = 204.017, 75 d.f., p < .05; NFI = .918; NNFI = .954; CFI = .962; IFI = .962; RMSEA = .060
Note: (*) expresses that coefficients are significant at the level of .01. Fig. 3. Structural equation models.
D. Belanche et al. / Journal of Retailing and Consumer Services 19 (2012) 124–132
129
Table 6 Multisample analysis: perceived risk. Constraints
Estimated coefficients (lower perceived risk)
Estimated coefficients (higher perceived risk)
D.F.
w2 differences
Probability
USAB–SAT
.581a t ¼6.934
.868a t ¼11.053
1
4.047
.044
USAB–I.USE
.199 t ¼ 1.155
.272 t ¼ 1.082
1
.008
.928
FIT INDICES
NFI ¼.871, NNFI¼ .947, CFI¼ .957; IFI ¼.958; RMSEA ¼ .063
NFI ¼ .879, NNFI ¼.952, CFI¼ .961; IFI ¼ .962; RMSEA ¼ .068
a
Expresses that coefficients are significant at the level of .01.
that these two direct effects must be significant. Specifically, results show a positive effect of both usability (l ¼.359; p o.01; t ¼4.595) and satisfaction (l ¼.558; po.01; t¼ 6.417) on intention to use a website. Thus, in the absence of any other determinant factor, the direct effect of usability on intention to use a website is positive. The next step would consist on including satisfaction as a partial mediator; that is, allowing for both direct and indirect effects (i.e., mediated through consumer satisfaction) of usability on intention to use a website (note that this is the model already shown in Fig. 2). In this situation, the direct effect of usability on intention to use a website becomes non-significant (l ¼ .271; p 4.05; t ¼ 1.719) suggesting that, at least, a partial mediation of consumer satisfaction exists. Finally, the model shown in Panel C (Fig. 3) depicts the indirect effect when satisfaction fully mediates the relationship between usability and intention to use a website. Then, this last model was nested within the previous one, and a chi-square (w2) difference test was performed to determine whether consumer satisfaction fully or only partially mediates the effect of usability on intention to use a website (Kulviwat et al., 2009). The test indicates that the partial mediation model provides the best fit to the data (Dw2(1) ¼4.922, p o.05). Thus, we may state that usability exerts an indirect effect on intention to use a website through consumer satisfaction.
Chi-squared when the constraint of equalizing one of the coefficients is eliminated. Thus LM-Test assesses if the elimination of this constraint supposes a significant change in the Chi-squared, and as a consequence a significant improvement in the model fit. Table 6 suggests that there is a significant difference between the groups to a level of .05 for the relationship between perceived usability and consumer satisfaction. To be precise, the effect of perceived usability on satisfaction is strengthened when perceived risk is high, which supports hypothesis 4a. On the contrary, there is no significant difference on the other constraint considered, so that hypothesis 4b is not supported. Indeed, the direct effect of usability on intention to use is found to be nonsignificant in both cases. However, taking into account that: (1) usability also exerts an indirect effect on intention to use through consumer satisfaction and (2) the direct effect of usability on satisfaction is strengthen when perceived risk is high, we may conclude that the indirect effect of usability on intention to use a website may be reinforced for high-risk perceivers. Thus, when perceived risk is high, results suggest that the role of usability in forming behavioral intentions becomes more relevant.
5.1. Multisample analysis
Online stores usually expend significant sums of money to set up and maintain their websites. It is not a trivial question since important performance ratios may depend on users’ judgment about the online establishment and their intention to use this more times. However, nowadays consumers have different skills and abilities regarding the Internet use, so that companies must focus not only on complex and attractive website design but also on developing easy-to-use websites. Although previous literature has noted the importance of website usability, it is necessary to understand the real influence of usability on key variables that may benefit the host-firm and to find under what conditions usability affects in a greater extent to customers’ perceptions. The results of this work suggest that usability has a significant impact on consumer behavior. More specifically, website usability has been first found to directly influence customer satisfaction. Although some authors argue that usability will not necessarily favor a positive impression among users (Chen and Yen, 2004, Lowry et al., 2006), this finding is consistent with most of previous studies (e.g. Kim and Eom, 2002, Casalo´ et al., 2008). Second, this work suggests that website usability also influences the consumer intention to use a website indirectly through consumer satisfaction. This mediating role of consumer satisfaction represents a contribution to the literature since previous authors have proposed either a direct influence of usability on consumer loyalty to a website (e.g. Flavia´n et al., 2006) or a
In order to assess the moderating role of perceived risk (Hypothesis 4) a multisample analysis was performed. To do that, we employed the statistical software EQS version 6.1. More specifically, we divided the total sample into two groups according to their perceived risk in the selected website. Following Garcı´a et al. (2008), total sample was divided into two groups according to the arithmetic mean of the moderating variable. Around this mean we eliminated some cases (71/2 standard deviation). The first group was formed by 107 cases representing consumers that showed a low perceived risk in the selected website. The second group was formed by 90 cases representing consumers with a high perceived risk. The risk mean for the first group (M¼1.07) is significantly lower (t¼ 17.103; po.01) than the mean for the second group (M¼3.26), which supports the creation of these groups. First, multisample analysis generates an individual structural solution for each group. Fit indices for the whole multisample analysis shows acceptable fit levels (NFI¼.859, NNFI¼.943, CFI¼ .950); also fit indices for each subsample (reported in Table 5) are close to the recommended values in spite of the sample size reduction. Second, multisample analysis offers information about the significance of the differences between the coefficients of the two models. To assess these differences we use the LM-Test. This contrast analyzes the variation of the
6. Discussion and managerial implications
130
D. Belanche et al. / Journal of Retailing and Consumer Services 19 (2012) 124–132
non-significant relationship between usability and the attraction of viewers to the website (Chen and Yen, 2004, Lowry et al., 2006). In other words, a logical sequence among the study variables has appeared. This logical sequence implies that intention to use a website might depend on the development of consumer satisfaction, which emerges in part as a result of the proper management of usability on the website. As a consequence, although usability may not be one of the most relevant direct antecedents of intention to use a website, managers should not forget usability when designing their websites since it still influences consumer intentions by means of other variables such as satisfaction. In this situation, the smaller direct effect of usability may be compensated by the indirect effect through satisfaction. In addition, in this work we have analyzed under which conditions the role of usability is especially relevant. In this respect, due to the fact that perceived risk plays an important role in affecting users’ decision of e-services adoption (Hsu and Chiu, 2004), we have concentrated on how risk perceptions might moderate the influence of usability. Indeed, because consumers have different skills in managing and controlling a website, they may perceive different levels of risk when interacting with a website. Our results suggest that the influence of usability is strengthened in those consumers that perceive a great risk, probably because usability helps overcome these consumer fears and form a more favorable opinion of using a website. All these findings suggest that website usability still plays a relevant role in the e-commerce context. 6.1. Managerial implications Thus, these findings allow us to offer some alternatives to improve the levels of customer intention to use a website and, as a result, the retention-rate and profits of the e-business, which are key objectives of most organizations:
First, due to the influence of usability on consumer behavior,
management should not give priority to the design of complex websites full of multi-media effects but concentrate instead on designs and structures that are simple and easy for the user to understand. That is, the most effective website may not be the most sophisticated one, but the most easy to use (Casalo´ et al., 2007). Especially interesting is the role of usability when perceived risk is high. Indeed, companies must ensure great levels of website usability in those cases that consumers perceived as especially uncertain. Among others, when a company is not wellknown or when credit card is required to pay an online purchase, managers should prioritize usability in website design. For this goal, companies should take into account aspects such as: anticipation (i.e. websites must be designed according to the potential visitors’ needs and requirements), consistency (websites must be consistent with consumers’ knowledge and skills), reversibility (i.e. websites must allow consumers to undo all the tasks), legibility, efficiency (i.e. simplifying processes and minimizing errors), visibility (i.e. hidden menus are not recommended), adequacy of information (i.e. too much information might be counterproductive) or provision of feedback (this will reduce consumers’ uncertainty regarding the security of the purchasing processes) among others. Second, due to the relevance of consumer satisfaction in forming behavioral intentions, companies should try to maximize the satisfaction of their customers during their interactions through the companies’ websites. To be precise, customer satisfaction will be generated if the customer’s expectations about the relationship are met (e.g. Oliver, 1980). Therefore, companies should try to identify the needs of their online customers (not only in terms of website usability and design but also in terms of services offered, etc.) in order to offer them what they require in an efficient way.
6.2. Limitations and future research Despite the contributions of this work, we acknowledge several limitations of this research that also offer some possibilities for further research. First, in this work we have only considered consumer perceptions regarding a specific website (the one of the retailer that collaborated in the research), which means that extrapolating the findings to other websites requires great care. As we have mentioned in the data collection section, data was collected among the users of a bus ticket e-selling service offered by a wellknown transport company in Spain, which might reduce risk perceptions (the mean of perceived risk is quite low as can be seen in Table 3). The election of a well-known retailer was based on data collection convenience; however, this might represent a limitation of the study that opens new research avenues. For instance, it would be interesting to investigate if the role of perceived risk is different when retailers are unknown, or analyze possible differences in the role of risk when considering multichannel retailers versus those operating exclusively online. As well, the sample consists of only Spanish-speaking consumers, so that an interesting route to extend this research would be to repeat the study using a wider sample of customers (for example in terms of different cultures). This effort might help firms develop websites more efficiently depending on where their potential consumers’ come from. In addition, it would be useful to extend this work by analyzing the influence of usability on other key variables that determine consumer behavior depending on the product/service category provided. For instance, it is possible that greater levels of usability may be also associated to greater levels of perceived security in the website, increasing consumer trust as well. This analysis would help to clarify in more detail the relevance of website usability for managers in different sectors. Finally, it would be also a good idea to analyze the effects of website usability on consumer behavior when the consumer accesses the Internet by new methods (e.g. mobile phone, PDA, etc.). This would serve to understand the role of usability when individuals and companies interact by means of these new technologies.
Acknowledgments The authors are grateful for the financial support received from Spanish Ministry of Education (ECO2009-10157).
Appendix
USABILITY (USAB) USAB1 In this website everything is easy to understand USAB2 This website is simple to use, even when using it for the first time USAB3 It is easy to find the information I need from this website USAB4 The structure and contents of this website are easy to understand USAB5 It is easy to move within this website USAB6 The organization of the contents of this site makes it easy for me to know where I am when navigating it USAB7 When I am navigating this site, I feel that I am in control of what I can do SATISFACTION (SAT) SAT1 I think I made the correct decision to use this website SAT2 The experience that I have had with this website has been satisfactory
D. Belanche et al. / Journal of Retailing and Consumer Services 19 (2012) 124–132
SAT3
In general, I am satisfied with the way that this website has managed interactions SAT4 In general, I am satisfied with the service I have received from the website INTENTION TO USE (I_USE) I_USE1 I have the intention to use this website again in the near future I_USE2 The likelihood of using this website again is high I_USE3 If I had to use this website again, I will use it without any doubt PERCEIVED RISK (RISK) RISK1 Using this website causes me to be concerned with experiencing some kind of loss in the future RISK2 I am vulnerable to the actions conducted by this website RISK3 The actions conducted by this website may cause problems and uncertain consequences for me. Notes: Item in italics were rejected during the scale validation process. The scales were presented in Spanish due to respondent’s nationality.
References Abdeldayem, M.M., 2010. A study of customer satisfaction with online shopping: evidence from the UAE. International Journal of Advanced Media and Communication 4 (3), 235–257. AIMC—Asociacio´n para la investigacio´n de los medios de comunicacio´n, 2009. Audiencia en Internet segu´n el EGM febrero/marzo de 2009. Available: /http://www.aimc.esS (Accessed 19 October 2010). Anderson, E., Sullivan, M., 1993. The antecedents and consequences of customer satisfaction for firms. Marketing Science 12 (2), 125–143. Anderson, J., Gerbing, D., 1988. Structural equation modeling in practice: a review and recommended two-step approach. Psychological Bulletin 103 (3), 411–423. Bagozzi, R., Yi, Y., Phillips, L., 1991. Assessing construct validity in organizational research. Administrative Science Quarterly 36 (3), 421–458. Becker, S., Mottay, F., 2001. A global perspective on web site usability. IEEE Software 18 (1), 54–61. Bhattacherjee, A., 2001. Understanding information systems continuance: an expectation-confirmation model. MIS Quarterly 25 (3), 351–370. Bloemer, J., de Ruyter, K., 1998. On the relationship between store image, store satisfaction and store loyalty. European Journal of Marketing 32 (5/6), 499–513. Bridges, E., Florsheim, R., 2008. Hedonic and utilitarian shopping goals: the online experience. Journal of Business Research 61 (4), 309–314. Brockman, B., 1998. The influence of affective state on satisfaction ratings. Journal of Consumer Satisfaction, Dissatisfaction and Complaining Behavior 11, 40–50. Casalo´, L., Flavia´n, C., Guinalı´u, M., 2007. The role of usability and satisfaction in the consumer’s commitment to a financial services website. International Journal of Electronic Finance 2 (1), 31–49. Casalo´, L., Flavia´n, C., Guinalı´u, M., 2008. The role of satisfaction and website usability in developing customer loyalty and positive word-of-mouth in the e-banking services. International Journal of Bank Marketing 26 (6), 399–417. Cronbach, L., 1970. Essentials of Psychological Testing. Harper and Row, New York, NY. Cronin, J., Brady, M., Hult, G., 2000. Assessing the effects of quality, value, and customer satisfaction on consumer behavioral intentions in service environments. Journal of Retailing 76 (2), 193–218. Chen, K., Yen, D., 2004. Improving the quality of online presence through interactivity. Information & Management 42 (1), 217–226. Chiu, C.M., Lin, H.Y., Sun, S.Y., Hsu, M.Y., 2009. Understanding customers’ loyalty intentions toward online shopping: An integration of Technology Acceptance Model and Fairness Theory. Behavior and Information Technology 28 (4), 347–360. Davis, F., 1989. Perceived usefulness, perceived ease of use, and user acceptance of information technology. MIS Quarterly 13 (3), 319–340. De Ruyter, D., Wetzels, M., Kleijnen, M., 2001. Customer adoption of e-service: an experimental study. International Journal of Service Industry Management 12 (2), 184–207. De Vaus, D., 2001. Surveys in Social Research, 5th edition Social Research Today. Devaraj, S., Fan, M., Kohli, R., 2002. Antecedents of B2C channel satisfaction and preference: validating e-commerce metrics. Informaton Systems Research 13 (3), 316–333. Flavia´n, C., Guinalı´u, M., Gurrea, R., 2006. The role played by perceived usability, satisfaction and consumer trust on website loyalty. Information & Management 43 (1), 1–14. Fornell, C., Larcker, D., 1981. Evaluating structural equation models with unobservable variables and measurement error. Journal of Marketing Research 18, 39–50.
131
Garcı´a, N., Sanzo, M., Trespalacios, J., 2008. Can a good organizational climate compensate for a lack of top management commitment to new product development? Journal of Business Research 61 (2), 118–131. Guinalı´u, M., 2005. La gestio´n de la confianza en Internet. Un factor clave para el desarrollo de la economı´a digital. Doctoral Thesis. University of Zaragoza, Spain. Hair, J., Anderson, R., Tatham, R., Black, W., 1998. Multivariate analysis. Prentice Hall International, Englewood. Hsu, M., Chiu, C., 2004. Predicting electronic service continuance with a decomposed theory of planned behaviour. Behaviour and information Technology 23 (5), 359–373. Jacoby, J., Kaplan, L., 1972. The components of perceived risk. In: Venkatesan, M. (Ed.), Proceedings of the Third Annual Conference, Association for Consumer Research, Champaign, IL, pp. 382–393. Janda, S., Trocchia, P., Gwinner, K., 2002. Consumer perceptions of Internet retail service quality. International Journal of Service Industry Management 13 (5), 412–431. ¨ Joreskog, K., 1971. Statistical analysis of sets of congeneric tests. Psychometrika 36 (2), 109–133. ¨ ¨ Joreskog, K., Sorbom, D., 1993. LISREL 8: Structural Equation Modeling with the SIMPLIS Command Language. Scientific Software International, Chicago, IL. Kim, J., Jin, B., Swinney, J.L., 2009. The role of etail quality, e-satisfaction and e-trust in online loyalty development process. Journal of Retailing and Consumer Services 16 (4), 239–247. Kim, E., Eom, S., 2002. Designing effective cyber store user interface. Industrial Management and Data Systems 102 (5), 241–251. Kulviwat, S., Bruner, G., Al-Shuridah, O., 2009. The role of social influence on adoption of high tech innovations: the moderating effect of public/private consumption. Journal of Business Research 62 (7), 706–712. Ladhari, R., 2010. Developing e-service quality scales: a literature review. Journal of Retailing and Consumer Services 17 (6), 464–477. Lee, M., Turban, E., 2001. A trust model for consumer Internet shopping. International Journal of Electronic Commerce 6 (1), 75–91. Liao, C., Chen, J.L., Yen, D.C., 2007. Theory of Planned Behavior (TPB) and customer satisfaction in the continued use of e-service: an integrative model. Computers in Human Behavior 23 (6), 2804–2822. Lichtenstein, D.R., Netemeyer, R.G., Burton, S., 1990. Distinguishing coupon proneness from value consciousness: an acquisition—transaction utility theory perspective. Journal of Marketing 54, 54–67. Lin, C., Wu, S., Tsai, R., 2005. Integrating perceived playfulness into expectationconfirmation model for web portal context. Information & Management 42 (5), 683–693. Liu, I., Chen, M.C., Sun, Y.S., Wible, D., Kuo, C.H., 2010. Extending the TAM model to explore the factors that affect intention to use an online learning community. Computers & Education 54 (2), 600–610. Lowry, P., Spaulding, T., Wells, T., Moody, G., Moffit, K., Madariaga, S., 2006. A theoretical model and empirical results linking website interactivity and usability satisfaction. In: Proceedings of the 39th Hawaii International Conference on System Sciences, Hawaii. Maditinos, D.I., Theodoridis, K., 2010. Satisfaction determinants in the Greek online shopping context. Information Technology and People 23 (4), 312–329. Neter, J., Wasserman, W., Kutner, M.H., 1990. Applied Linear Statistical Models. Irwin, Boston, MA. Oliver, R., 1980. A cognitive model of the antecedents and consequences of satisfaction decisions. Journal of Marketing Research 17 (4), 460–469. Oliver, R., 1999. Whence consumer loyalty? Journal of Marketing 63, 33–44. Oliver, R., DeSarbo, W., 1988. Response determinants in satisfaction judgments. Journal of Consumer Research 14 (4), 495. Oneupweb, 2010. Revolutionizing Website Design. Available: /http://www.oneup web.comS [Accessed 19 October 2010]. RED.ES, 2009. Perfil sociodemogra´fico de los internautas: Ana´lisis de datos INE 2008. Available: /http://observatorio.red.es/hogares-ciudadanos/articles/id/ 3027/perfil-sociodemografico-los-internautas-analisis-datos-ine-2008.htmlS (Accessed 19 October 2010). Roy, M., Dewit, O., Aubert, B., 2001. The impact of interface usability on trust in web retailers. Internet Research: Electronic Networking Applications and Policy 11 (5), 388–398. Schlosser, A.E., White, T.B., Lloyd, S.M., 2006. Converting web site visitors into buyers: how web site investment increases consumer trusting beliefs and online purchase intentions. Journal of Marketing 70, 133–148. Severt, D., 2002. The customer’s path to loyalty: a partial test of the relationships of prior experience, justice, and customer satisfaction. Doctoral Thesis. Faculty of the Virginia Polytechnic Institute and State University. Available: /http:// en.scientificcommons.org/1527127S (Accessed 19 October 2010). Shankar, V.S., Smith, A.K., Rangaswamy, A., 2003. Customer satisfaction and loyalty in online and offline environments. International Journal of Research in Marketing 20 (2), 153–175. Smith, J., Barclay, D., 1997. The effects of organizational differences and trust on the effectiveness of selling partner relationships. Journal of Marketing 61 (1), 3–21. Smith, K., 1974. On estimating the reliability of composite indexes through factor analysis. Sociological Methods & Research 2 (4), 485–510. Steenkamp, J., Geyskens, I., 2006. How country characteristics affect the perceived value of Web sites. Journal of Marketing 70 (3), 136–150. Stone, R., Grønhaug, K., 1993. Perceived risk: Further considerations for the marketing decision. European Journal of Marketing 27 (3), 39–50.
132
D. Belanche et al. / Journal of Retailing and Consumer Services 19 (2012) 124–132
Szymanski, D.M., Hise, R.T., 2000. E-satisfaction: an initial examination. Journal of Retailing 76 (3), 309–322. Wu, J., Chang, Y., 2005. Towards understanding members’ interactivity, trust, and flow in online travel community. Industrial Management and Data Systems 105 (7), 937–954.
Wu, I.L., Chen, J.L., 2005. An extension of trust and TAM model with TPB in the initial adoption of on-line tax: an empirical study. International Journal of Human–Computer Studies 62, 784–808. Zaichkowsky, J., 1985. Measuring the involvement construct. Journal of Consumer Research 12 (3), 341–352.