Impact of E-Service Quality and Customer Value on Customer Satisfaction in LocalBrand

This research aimed to investigate the impact of e-service quality and customer value on customer satisfaction. This research was descriptive-associative in which researcher described not only the value of research variables but also the influence or relationship between independent variables and the dependent variable. A survey on e-marketplace of online fashion local brand in Indonesia was conducted. Using purposive sampling, the researcher distributed questionnaires to buyers or customers on the LocalBrand (LB) website, and 150 questionnaires were analyzed. Partial Least Square (PLS) was used as the statistic method for analyzing the research model. The results show that customer value has a significant influence on the customer satisfaction. However, e-service quality does not have a significant influence on customer satisfaction.


INTRODUCTION
Nowadays, business and technology become one package in the business strategy.Technology, especially the progress of Internet development, changes the business, e-business or e-commerce, and social media (Octavia & Tamerlane, 2017;Sundjaja, Gaol, Abdinagoro, & Abbas, 2017).Online transactions or commonly called as e-commerce in Indonesia is not new.The dotcom era that gave birth to amazon.com and e-bay has revolutionized conventional shopping.In Indonesia, some well-known online shopping sites are to the public are Tokopedia, Bukalapak, Zalora, Lazada, LocalBrand (LB) and many more.
To compete in today's business competition, LB or other company can not only rely on product quality but also on the quality of service.It can further encourage customers to re-purchase the products the companies offer.Companies are required to focus on the needs of the customers.It can begin with changing the mindset regarding profit-oriented to other potential factors such as customer satisfaction and loyalty.This is because the customer is an important element for the company.Business competition can mean competing for getting the customers.The winner is not the strongest or the greatest, but it can satisfy the customers.In this increasingly fierce business competition, companies need to realize that the customer's decision to buy the goods is based on good product quality, competitive prices, and service quality.From this service, it will lead to customer assessment and result in customer satisfaction.
Service can be defined as deed, effort, or performance.Meanwhile, service quality is an attitude that is formed for the long-term and the overall evaluation of the company's performance.Kotler and Keller (2015) defined service as any action that a party could offer to an essentially intangible party and did not result in the ownership of something.Moreover, Blut, Chowdhry, Mittal, and Brock (2015) and Zemblytė (2015) agreed that e-service quality was as the extent to which the website facilitated shopping, purchasing, delivery of products, and services effectively and efficiently.
According to Candra (2014b), The researches of service quality and e-service quality have been conducted, and different scales have been developed to measure e-service quality.However, the existing research on e-service quality has been described separately.Usually, it is the research related the service quality with customer value impact.
Customer value describes the exchange between quality or benefit received by customers and costs such as money, energy, time.Those encourage the customers to evaluate, obtain, and use the product (Kasiri, Cheng, Sambasivan, & Sidin, 2017).According to Carlos Fandos Roig, Sanchez Garcia, Angel Moliner Tena, and Llorens Monzonis (2006) and Sánchez, Callarisa, Rodríguez, and Moliner (2006), customer value dimension consists of the functional value, emotional value, and social value.Functional value is defined as the perceived utility of the product and service attributes.In this research, there are four functional values.First, it is the functional value of establishment (installation).It addresses the available registration page on the website.Second, there is the functional value of personnel (professionalism).The utility is provided by the personnel (staff).Third, it is the functional value of the service purchased (quality).The perceived utility is derived from the service quality provided.Fourth, the functional value of price.It is the perceived utility of the cost level incurred.Meanwhile, emotional value consists of feelings or affective states created through the experience of consumption.Then, social value is acceptance related to the level of individual relationships with their social environment.
Furthermore, customer satisfaction is a level that the needs, wishes, and expectations of the customer can be met and can result in repeat purchase or continued loyalty.The most important factor for obtaining consumer satisfaction is the performance of the agent.It is usually interpreted by the quality of the agent (Nagengast, Evanschitzky, Blut, & Rudolph, 2014).Bressolles and Durrieu (2011) examined the link between service quality, customer value, and customer satisfaction through the dimensions of the three variables in the online context wine website.The results of this research highlighted the impact of e-service quality (information, ease of use, design, security/ privacy, and reliability) dimensions on customer satisfaction through customer value (transactional, utilitarian, and social value) dimensions.These results ultimately led the researchers to a conclusion that from a managerial point of view and to improve customer value and satisfaction, online wine sellers should focus on different dimensions of e-service quality especially on user convenience; propose user-friendly navigation interface; facilitate internet users by offering several search options based on price, region, type of wine, color, consumption, and others; and provide internal search engine.Then, the reliability in this research indicated the level of inventory in real-time by offering various delivery methods and time ranges.In the design, the company created video regarding the bottle.Finally, in the security/privacy, the company offered various payment methods and provided information relating to policies in the data protection.
LB is one of the e-commerce sites that was launched at the 66th Indonesia Independence Day.LB has a different concept from other established e-commerce in Indonesia.LB is an online fashion mall that sells local product from Indonesian fashion designer.There are several products from each brand, and it will be a different approach offered by LB.Moreover, the quality is the main key point that LB also offers (LocalBrand, 2011).
Based on the explanation, this research aims to investigate the impact of e-service quality and customer value on customer satisfaction from LB.The finding can give suggestions for LB and other e-commerce that have the similar services like LB.
In the future, it can be used as a guide to develop strategies to improve customer's relationships and create customer satisfaction.

METHODS
This type of research is descriptive-associative.The researcher does not only describe the value of research variables but also defines the influence or relationship between independent variables and the dependent variable.A survey is used that the researcher distributes the questionnaires to buyers or customers on the LB website.The used sampling technique is probability sampling with purposive sampling.Purposive sampling is a sampling technique that is restricted to certain types of people who can provide the desired information.It is because they are the only ones who have it or according to some criteria set by the researcher (Sekaran & Bougie, 2016).The analysis technique in this research is Partial Least Square (PLS) method (Candra, 2014a).
The website looks neat and well organized (Y01).

Registration is clear and modern (Y01).
It is easy to find and access (Y01).

Functional Value of Personnel (staff) -Professionalism
Personnel knows their work well (Y01).Personnel's knowledge is up to date (Y01).
The information provided by personnel is invaluable (Y01).
Personnel knows all services owned by the company (Y01).

Functional Value of the Service Purchased -Quality
The overall service is appropriate (Y01).Quality is always maintained (Y01).
The level of quality is acceptable compared to other companies (Y01).The result of received service is as expected (Y01).

Functional Value of Price
The price of the product is in accordance with the offered quality (Y01).The provided service is in accordance with the cost incurred (Y01).The total fees charged are very relevant (Y01).
Emotional Value I am happy with the service (Y01).I feel good (Y01).
The staff gives me positive feeling (Y01).The staff does not bother me (Y01).

Social Value
It is considered as very good at the social level (Y01).
The fact that I come here looks good to the people I know (Y01).

RESULTS AND DISCUSSIONS
The researcher begins to spread the questionnaires through social media (twitter) in @ LocalBrandID.It is a twitter of LB.The questionnaire is sent through newsletter with e-mail by LB.The target amount is 344 respondents.The distribution of this questionnaire is conducted for approximately a month.After a month, there are about 195 respondents who fill out the questionnaires.However, 150 respondents answer that they have shopped on LB before.Meanwhile, 45 respondents say no.It means that only data from 150 respondents can be used or 43,60%.The response rate is acceptable.In the previous researches, the response rate for the online questionnaire is always lower than the result of using a physical questionnaire.The response rate using the online questionnaire in both researches is only 23% and 43 (Saleh & Bista, 2017).
Based on survey results collected from all 150 respondents, the researcher obtains data that the majority of the respondent is female (100 respondents or 67%).Meanwhile, the rest is male (50 respondents or 33%).It can be said that most buyers or customer in LB and the tendency to shop online are dominated by women.
Based on age, the result shows that most respondents are 21-30 years about 104 respondents or 69%.It is followed by <20-year respondents with 40 respondents or 27%.There are 5 respondents with age of 31-40 years or 3%.Meanwhile, the minority of respondents with age of >51 years is only 1 or 1%.It can be concluded that the tendency to shop is more done at the age of 21-30 years that they have a need for fashion.
Moreover, from the education, it shows a various result.About 57 respondents (57%) have high school education.Similarly, with the same number as the previous result, 57 respondents (57%) have a bachelor degree (S1).Then, it is followed by respondents with junior school education about 26 respondents (17%).Meanwhile, in diploma level and others, the number of respondents are 9 (6%) and 1 (1%) respectively.
Next, in status, most of the respondents are students or 99 respondents with the highest percentage of 65%.It is followed by 24 respondents who work as private employees (16%).Then, the respondents who are civil servants are 14 with a percentage of 9%.Meanwhile, the respondents who are self-employed and others are 9 and 6 respondents with the percentage of 6% and 4% respectively.
The experience with the Internet is an important consideration in making purchases online.Inexperienced internet users rarely buy online (Candra, 2015).Based on survey results, most respondents have the experience in online shopping for 1-3 years.It has 75 respondents and 50% percentage.Then it is followed by <1 year of experience with 51 respondents (34%), and 20 respondents (13%) has 3-5 years of experience.Then, 3 respondents have >7 years of experience or 2%.The respondents with 5-7 years of experience are only 1 or 1%.
To answer the hypothesis, the researcher will evaluate the measurement model.It determines the relationship between the latent variable and its indicator or manifest variable.This measurement is done through three steps: convergent validity, discriminant validity, and composite reliability (Candra & Gunawan, 2017).
To measure convergent validity, it can be done in combined loadings and cross-loadings table at SEM output.There are two recommended criteria as a basis to determine whether the measurement model has met the convergent validity.P-value should be less than 0,05, and the loading factor should be equal to or greater than 0,5 (Candra, 2016b).Based on Table 4 to Table 6, the results of p-value and loading factor for e-service quality, customer value, and customer satisfaction are good.All data can be used for the next analysis.However, X05 and X14 are dropped in the next analysis.It is because the loading factor is under 0,5.Average Variance Extracted (AVE) describes the variant or diversity of manifest variables that can be contained by latent variables.The criterion is AVE value > 0,5.It means that the variable has met the convergent validity.Based on Table 7, it is known that AVE value for e-service quality variable is 0,368, customer value with 0,390 and customer satisfaction with 0,739.These mean that AVE value generated by customer satisfaction is greater than 0,5.In accordance with the AVE criteria, customer satisfaction variables show a good convergent validity measurement.Meanwhile, the value of AVE on e-service quality and customer value shows that the results are smaller than 0,5.It implies that these two variables do not have a good convergent validity.Discriminant validity of the measurement model is assessed based on measurement of cross measurement with variables.If the variable correlation with the principal measurement (each indicator) is greater than the size of the other variable, the latent variable predicts the indicator better than the other variables.From Table 8 to Table 10, it shows the finding of discriminant validity.Overall, all variables meet the discriminant validity criteria.It shows that all latent variables predict indicators on their blocks better than the other blocked indicators.Based on the results of PLS output, the e-service quality variables, especially on the loading values X09, X10, X11, and X14, are smaller than the correlation of the e-service quality indicator with other variables.This shows that the loading values of X9, X10, X11, and X14 do not meet the discriminant validity criteria.Then, the correlation of the customer value variable on the value of loading Y01 is smaller than the correlation of the customer value indicator with other variables.It indicates that the loading value of Y01 does not meet the criteria of discriminant validity.Meanwhile, customer satisfaction variable shows bigger correlation variable indicator than correlation indicator of customer satisfaction with another variable.
After testing the validity of variables, the researcher conducts internal consistency reliability test.It is measured by two criteria, namely composite reliability and Cronbach's alpha.Those are from the indicator block that measures the variables.A variable is considered reliable if the values of composite reliability and Cronbach's alpha are above 0,7 (Candra, 2016a).
Table 11 shows the satisfactory composite reliability coefficient of each variable.The result of each variable is e-service quality (0,889), customer value (0,933), and customer satisfaction (0,895).In addition to these composite reliability results, the coefficients of Cronbach's alpha also show good results for e-service quality (0,864), customer value (0,864), and customer satisfaction (0,823).Therefore, it can be concluded that each variable has high reliability.It can be seen from the values of composite reliability and Cronbach's alpha.All variables are greater than 0,7.Therefore, it can be concluded that each variable has high reliability.It can be seen from the values of composite reliability and Cronbach's alpha.All variables are greater than 0,7.
After evaluating the measurement model, the indicators of e-service quality, customer value, and customer satisfaction have fulfilled the requirements of convergent validity, discriminant validity, and composite reliability.Thus, it can be continued to the next stage such as evaluating the structural model including model fit indices, path coefficient, and R2.Before testing the significance of path coefficient and R2, model fit indices have to be tested.The test of model fit indices is done to know whether the research model meets the criteria so it can be used for further analysis.
In this model fit test as seen in Table 12, there are three indices: Average Path Coefficient (APC), Average R-Squared (ARS), and Average Variance Inflation Factor (AVIF).APC and ARS criteria are accepted when p-value is <0,05.Meanwhile, AVIF is accepted if it is smaller than 5 (Candra, 2016a).It can be seen in Table 12 that the APC and ARS indices have p-value of <0,001.It means that it has met the criteria of APC and ARS.Then, the AVIF also shows p-value of <5 such as 3,454.Thus, it can be concluded that the model meets the criteria and it can be used for further analysis.Hypothesis testing of path coefficients between the variables is used to evaluate the structural relationship between latent variables.It compares the p-value with alpha 0,05.This test is intended to measure the correctness of the alleged research or to answer questions about the effect or relationship as described in research background.Testing all hypotheses will be analyzed based on the results obtained from data processing in Table 13 and Table 14.
The e-service quality variable has a significant and positive influence on the customer value variable at LB.It can be observed through the value of the positive value coefficient of 0,844.This result indicates that if there is an increase in the assessment of e-service quality of one unit, the assessment of customer value will increase by 0,844 and vice versa.Similarly, if there is any decrease in the assessment of e-service quality of one unit, the assessment of the customer value will decrease by 0,844.Then, the effect size of R2 value of 0,713 can be seen.It means that e-service quality variable affects customer value variable in LB about 71,3%.The other (28,7%) is influenced by other variables that are not used in this research.The customer value has a significant and positive influence on the customer satisfaction variable at LB.It can be seen from the value of the positive value coefficient of 0,743.This indicates that if there is an increase in the assessment of customer value of one unit, the assessment of customer satisfaction will increase by 0,743 and vice versa.It is similar if there is a decrease in the assessment of customer value of one unit, the assessment of customer satisfaction will decrease by 0,743.Then, the effect size of R2 value of 0,563 can be seen.It implies that customer value variable affects customer satisfaction variable on LB about 56,3%.The rest (43,7%) is influenced by other variables which are beyond this research.
Based on Figure 1, it can be seen that e-service quality variable has a significant influence on customer value variable in LB.Bressolles and Durrieu (2011) stated that service quality was an important form of customer value.In addition, many researchers have validated the strong, positive, and significant effect of the product or service quality on customer value perceptions (Bressolles & Durrieu, 2011;El-Adly & Eid, 2016;Kasiri et al., 2017).Based on 71,3% of customer value obtained from e-service quality, it is necessary for LB to continue to pay attention and improve e-service quality.This finding is in accordance with previous research by Oh (1999).Oh (1999) stated that customer value was an important variable to be considered in service quality and customer satisfaction.
Customer value also has a significant on customer satisfaction at LB as described in data processing.This finding is similar to previous research by Bressolles and Durrieu (2011).They agreed that customer value and customer satisfaction showed a close relationship.It shows that customer value is the beginning of the formation of customer satisfaction.
Moreover, e-service quality has no significant influence on customer satisfaction in LB.This is different from the findings of previous research.The previous researchers state that e-service quality is related to customer satisfaction (Chang, Cheng, & Wang, 2007;Nicolaou & Bhattacharya, 2006).In this research, it is found that e-service quality has no effect on customer value directly.It must go through customer value first.E-service quality and customer value influence the customer satisfaction simultaneously in LB.According to Bressolles and Durrieu (2011), e-service quality can lead to an increase in customer value and customer satisfaction.E-service quality affects the dimension of customer value which ultimately affects customer satisfaction in return.This can be seen in Figure 1.The research model describes that customer satisfaction has R2 of 0,57.It means that the relationship between e-service quality and customer value affects customer satisfaction by 57%.

CONCLUSIONS
Based on the results of the analysis, the researcher recommends completing the service order attribute on the website.Thus, the customer can be easier to order.It is related to the e-service quality.The accuracy of time owned by LB in interacting with its customers is also considered very less.It is stated by most respondents in the results of the questionnaire.In accordance with this indicator, it is recommended that the service can be done in real-time so that customers feel comfortable shopping in LB.Then, it is also advisable to add staff so that the service time can be done up to 24 hours.Then, for customer value, the performance of personnel or staff in LB is considered as less helpful when customers seek important information.It will be better if the personnel or staff in LB are given knowledge about the job desk and the product to improve the service quality and can be more informative to give positive feelings to LB customers.This research still has limited sample and location of study.For future research, this model can be adapted and expanded by applying more than one e-marketplace or e-commerce in different countries or region.Hopefully, with that combination, the models can be generalized to answer the problem that happens among e-service quality.

Figure 1
Figure 1 Result of Research Model

Table 1
Operational Variable for E-Service Quality

Table 2
Operational Variable for Customer Value

Table 3
Operational Variable for Customer Satisfaction

Table 12 Model
Fit Indices