Relationship between Customer Loyalty, Customer Satisfaction, Customer Trust, and Service Quality in E-Commerce Setting: Case Study of Lazada in Indonesia

Copyright © 2018Authors. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Yulita Fairina Susanti , Albert Hasudungan , Arif Wiratama Prasetyo 1* 2 3 Sekolah Tinggi Manajemen Ipmi 1,2,3 International Journal of Business Studies Vol. 2 No. 2 ( June 2018) 94 which in 2016 there are 8.6 Million Indonesia shoppers (Bahar, 2017). Besides, the number of e-commerce shoppers in 2016 in Indonesia has the estimated value of about 4.5 Billion USD (Bahar, 2017), which still 0.83% from the total retail sales estimated in Indonesia (543.1 Billion USD). Even there is a very high gap between the volume sales of retail sales, slowly the share of total online sales as a percentage of retail sales has grown from 0.50% in 2013 to 0.83% in 2016. In fact, according to Harsono (2016), Indonesia has a lot of opportunity of e-Commerce among other emerging Asian economies, which projected become the third largest E-Market countries behind China and India with value 130 Billion USD. This study aims to examine the relationship between Customer loyalty, Customer satisfaction, Customer trust, and Service quality in e-commerce setting from the online sales of LAZADA Indonesia. The contribution to management study is to provide the shifting avenue of customer satisfaction from conventional shopping mall to web-based sales online transaction in the rapid growth of the e-commerce development in Indonesia. The target respondent will focus on Jakarta, the capital city of Indonesia. The core ingredient to support that online sales transaction will be availability and access to internet connection. In 2016, the most contributors for Internet Penetration in Indonesia is Java region with 65% of total Indonesia Internet User (APJII, 2016) or 170.3 Million people. Java Region dominates the Internet Penetration since there are several cities in Java region that include in the 10 most populous cities in Indonesia, such as Jakarta (9.608 Million people), Surabaya (2.765 Million people), Bandung (2.575 Million people), Bekasi (2.51 Million people), Semarang (2.06 Million people), Tangerang (2 Million people), Depok (1.869 Million people). Basically, Jakarta has the largest demographic population throughout big cities in Java island. In this study, the scope of the sampling unit is randomly selected with the scope of Millennial User, which male or female who 15 to 29 years old, ever purchase an item from LAZADA minimum two times in 2017, and the minimum transaction is Rp.100,000. From our literature development, the customer trust and satisfaction are mediating variables that can have the correlation with service quality indicators and customer satisfaction. Service quality dimensions here consist of Ease of Use, Website Design, Responsiveness, Customization, and Assurance. With such complex relationships, Covariance-Based Structural Equation Modelling will be selected as statistical techniques for developing, expanding, and con irming a theory, to make a better understanding of relationships between the variables. This article consists of several sections from literature review, research methodology, discussion & analysis, into conclusion. The literature review will cover the theoretical framework to derive the models for this research. The section for methodology and data will illuminate the strategy to collect the data, hypothesis and statistical conducts. The data analysis and result will be presented in the section of the results and discussion. Finally, the conclusion will synthesize the indings, state the study's limitation, and to suggest a recommendation for future research. LITERATURE REVIEW In the service industry, like e-commerce, the evaluation and assessment given by consumers based on either the poor quality of a service is part of determined indicators to assess customer loyalty. In that e-commerce industry, Ribbink, et al. (2004) develops ive dimensions of e-Service Quality in w-Commerce settings, which representing some humanistic interventions and technical factors in that setting. Those are Ease of Use, Website Design, Customization, Responsiveness, and Assurance. Those indicators have each own unique property that differs from one another. Firstly, the ease of Use refers to the level of ease of site to be used by consumers. Determinants are the level of functionality, accessibility to information on the site, ease of ordering, and ease of navigation


INTRODUCTION
Indonesia has recently been prospected by some global investors as the country that can maintain their economic stability and one of the largest population number in the world. The penetration of e-commerce online business platform has grown signi icantly in Indonesia during the ive years. From their statistics, the growth of e-commerce in Indonesia from the year 2013 to 2016 is 17.27%, Vol. 2 | No. 2 (June 2018) The main objective of this study is to examine the relationship of each dimension of Service Quality to Customer Satisfaction and Customer Trust, Customer Satisfaction to Customer Trust and Customer Loyalty, and Customer Trust to Customer Loyalty. The sampling is selected on the basis of random sampling from Jakarta Millennial generation, below 31 years old, that use online transaction. The questionnaire was distributed in 2017 to that Jakarta Millennial respondents on the basis of random sampling selection. This study uses Covariance-Based Structural Equation Modelling, and the variable is adopted by the previous study to see that relationship. The research inds that Website Design, Responsiveness, and Assurance has signi icant effect to Customer Satisfaction, Customization and Assurance has signi icant effect to Customer Trust, Customer Satisfaction has signi icant effect to Customer Trust and Customer Loyalty, and Customer Trust has signi icant effect to Customer Loyalty. This study recommends equal proportion of business efforts between the service delivery and customize technology to retain their customer loyalty, since the results underscore the importance of both technical and service delivery aspects.
This study aims to examine the relationship between Customer loyalty, Customer satisfaction, Customer trust, and Service quality in e-commerce setting from the online sales of LAZADA Indonesia. The contribution to management study is to provide the shifting avenue of customer satisfaction from conventional shopping mall to web-based sales online transaction in the rapid growth of the e-commerce development in Indonesia.
The target respondent will focus on Jakarta, the capital city of Indonesia. The core ingredient to support that online sales transaction will be availability and access to internet connection. In 2016, the most contributors for Internet Penetration in Indonesia is Java region with 65% of total Indonesia Internet User (APJII, 2016) or 170.3 Million people. Java Region dominates the Internet Penetration since there are several cities in Java region that include in the 10 most populous cities in Indonesia, such as Jakarta (9.608 Million people), Surabaya (2.765 Million people), Bandung (2.575 Million people), Bekasi (2.51 Million people), Semarang (2.06 Million people), Tangerang (2 Million people), Depok (1.869 Million people). Basically, Jakarta has the largest demographic population throughout big cities in Java island. In this study, the scope of the sampling unit is randomly selected with the scope of Millennial User, which male or female who 15 to 29 years old, ever purchase an item from LAZADA minimum two times in 2017, and the minimum transaction is Rp.100,000.
From our literature development, the customer trust and satisfaction are mediating variables that can have the correlation with service quality indicators and customer satisfaction. Service quality dimensions here consist of Ease of Use, Website Design, Responsiveness, Customization, and Assurance. With such complex relationships, Covariance-Based Structural Equation Modelling will be selected as statistical techniques for developing, expanding, and con irming a theory, to make a better understanding of relationships between the variables. This article consists of several sections from literature review, research methodology, discussion & analysis, into conclusion. The literature review will cover the theoretical framework to derive the models for this research. The section for methodology and data will illuminate the strategy to collect the data, hypothesis and statistical conducts. The data analysis and result will be presented in the section of the results and discussion. Finally, the conclusion will synthesize the indings, state the study's limitation, and to suggest a recommendation for future research.

LITERATURE REVIEW
In the service industry, like e-commerce, the evaluation and assessment given by consumers based on either the poor quality of a service is part of determined indicators to assess customer loyalty.
In that e-commerce industry, Ribbink, et al. (2004) develops ive dimensions of e-Service Quality in w-Commerce settings, which representing some humanistic interventions and technical factors in that setting. Those are Ease of Use, Website Design, Customization, Responsiveness, and Assurance.
Those indicators have each own unique property that differs from one another. Firstly, the ease of Use refers to the level of ease of site to be used by consumers. Determinants are the level of functionality, accessibility to information on the site, ease of ordering, and ease of navigation (Reibsten, 2002). Secondly, Website Design is the Display a site ranging from colors that used by the site for users to feel comfortable and website design, with the purposes to pleasing the eye of the consumer (Wol inbarger & Gilly, 2003). Thirdly, Customization is the adjustment of the level of service provided by a site to the desires and needs of each consumer. The example is providing feedback from the previous purchases made by consumers and other information that consumers have provided. In addition, Responsiveness is the speed of company in replying to questions or request that submitted by the consumers. Finally, Assurance is the security of a site in maintaining the data that has been given by consumers and also the entire information owned by the consumers.
To drive more customer loyalty, those ive dimensions are usually mediated through customer satisfaction and customer trust. Dabholkar (1996) states that customer satisfaction is the combination between the cognitive loyalty and affective loyalty elements, which shows Satisfaction can increase the customer loyalty, since satisfaction create a big in luence on future purchase intentions (Kassim & Abdullah, 2008). In customer trust's point of view, (Wang et al., 2016) argues that it is very important to penetrate the consumer to continuously use E-Commerce with trust. Trust is an outcome that was created by customer satisfaction (Kassim & Abdullah, 2008), also directly affected by services quality that gives an effect to customer loyalty (Ribbink et al., 2004).
In previous empirical studies, there are diverse relationships of relationship between variables of Service Quality (SQ) to Customer Satisfaction (CS), Service Quality (SQ) to Customer Trust (CT), Customer Satisfaction (CS) to Customer Trust (CT), Customer Satisfaction (CS) to Customer Loyalty (CL), and Customer Trust (CT) to Customer Loyalty (CT) from several sectors that will help this study stronger to discuss.
From those studies, the researchers consolidate the dependent, intermediaries and independent variables into the structural equation model (SEM) for this study as follows.
From the igure 1 above, there are four variables, which are Service Quality (with dimensions: Ease of Use, Website Design, Responsiveness, Customization, and Assurance), Customer Satisfaction (named as Satisfaction), Customer Trust (named as Trust), and Customer Loyalty (named as Loyalty), which adopted by Ribbink et al. (2004) since their study focuses on the empirical study of E-Commerce Loyalty. Based on the theory above, the hypothesis being tested will be as follows: On the hypothesis's development, elaborating from the research objective, there are six points that the researchers intend to examine. The study examines 5 hypotheses, which 'The Effect of Service Quality to Customer Satisfaction', 'The Effect of Service Quality to Customer Trust', 'The Effect of Customer Satisfaction to Customer Trust', 'The Effect of Customer Trust to Customer Loyalty', and 'The Effect of Customer Satisfaction to Customer Loyalty'. Those hypotheses are as follows: 1. The Effect of Service Quality to Customer Satisfaction H1a: Service Quality (Ease of Use) has a positive impact to Customer Satisfaction H1b: Service Quality (Website Design) has a positive impact to Customer Satisfaction H1c: Service Quality (Responsiveness) has a positive impact to Customer Satisfaction The The

RESET METHOD
The information for this study is collected from the primary data. The data source is gathered with the survey to the respondents included in the target population that has been determined by the researchers. The study adopts judgmental sampling. Malhortra (2010) de ines judgmental sampling as a convenience sampling form with speci ic population elements that have been chosen by the researchers based on the considerations they have. Since the sampling is selected upon the criteria, this judgmental sampling its with.
For the procedures of data collection, the main data source used to determine the results of the study is primary data collected through the survey to the respondents included in the target population that has been determined by the researchers. The population will be the millennial generation to ship with the sales online of LAZADA that has been a resident from Jakarta, the capital city of Indonesia. Speci ically, this research selects 365 respondents with the criteria as follows: The respondent age must in between 15 to 29 years old since this study objective is to scrutinize the perspective from the millennial generation. Live in Jakarta. The minimum transaction history is two times transaction in 2017. The minimum transaction is Rp. 100.000. In this study, the measurement of the variables is listed in the following table.
In the survey, those variables are translated into 33 of questions, which will be the indicator for will be delivered using 6 Likert scales as the technical scaling on the questionnaire survey. The reason for using 6 Likert scale or points is because it tends to give the discrimination and reliability higher than Likert's scale 5 points (Chomeya, 2010). The purpose of Indicator is to avoid misunderstanding between the respondent and the survey on de ining every variable that been used for this research. All of the questionnaires for each measurement item is adopted from Kassim & Abdullah (2008) who adapted from Ribbink et al. (2004).
After data collection, there are several statistical techniques and estimation that are used to support the data analysis.  It is easy to get access to the online organization's website in which I obtained online services. several it indices that can be chosen one of them for the test, which NFI, TLI, CFI, and RNI. c. Parsimony Fit Indices: A test for testing the model savings, which a model that has thehigh degree of it for each degree of freedom. There are several it indices that can be chosen one of them for the test, which AGFI, and PNFI.

Convergent
Validity, Reliability, and Discriminant Validity: on Convergent Validity is to see the outer loading with the rule of thumb not less than 0.500 (Hair et al. 2010). On Convergent Reliability, there are Average Variance Extracted and Composite Reliability with the rule of thumb not less than 0.500 (Ibid.). The last thing is Discriminant Validity with Fornell-Larcker with the square roots must bigger than all of its correlation (Ibid.) 3. Hypothesis Test: Testing the structural model with the hypotheses that already been developed before. There are several elements that can a rule of thumb to judge whether the hypotheses can be accepted or rejected, which: a. P Value: Accept if the P-Value is less than 0.05, Reject if the P -Value is more than 0.05 b. T Value: Accept if the T Value is more than 1.9, Reject if the T Value is less than 1.9

RESULTS AND DISCUSSIONS
The respondents' pro iles are shown in table 3. The total of respondents are 365 people, all from Jakarta. The highest proportion of respondent is male. The majority monthly expenditure of respondent is between Rp. 1,500,001 to Rp. 2,500,000. From that table 3, 60% of respondents spend averagely between Rp. 100,000 up to Rp. 500,000, means the majority of the respondent has a buying power to buy an item from LAZADA, even though it is low, but they did at least do twice transaction on LAZADA. Besides, they are actively do shopping online on LAZADA, since based on table 5 there 81% or the majority of respondents did 2 up to 3 times transaction on LAZADA.
Before testing the hypotheses, we assess the goodness of it, convergent validity, reliability, and discriminant validity of our proposed SEM model. The goodness of it model was summarized in the con irmatory factor analysis as in Table 4.

Average Spending (IDR):
100.000 -500,000 501,000 -1,000,000 1,000,001 -1,500,000 More than 1,500,000 1% 10% 56% 23.5% 9.5% 60% 5% 31% 4% International Journal of Business Studies Vol. 2 No. 2 ( June 2018) From the Con irmatory Factor Analysis above, the result shows that all of the indices are mark as acceptable since they are above the rule of thumb. In the absolute it indices, the Chi-Square value is 2,170 which has a signi icant P-Values, the GFI value is 0.91which higher than the rule of thumb. The RMSEA that less than 0.08, which the value is 0.055. In the incremental it indices, the NFI value is 0.760, and the CFI is higher than 0.92, which 0.95. For the parsimony it indices, the PNFI value is 0.809, which is good since it is closer to 1.
After that goodness of it, the assessment will be convergent validity, reliability, and discriminant validity tests, as presented in With this situation, LAZADA has low point on Customer Loyalty that represents from the four items (LOY1, LOY2, LOY3, and LOY40) that ranged between 2.539 to 2.832, which 40% in luenced by the Customer Satisfaction that build by Website Design, Responsiveness, and Assurance, also 39% in luenced by Customer Trust that was formed by Customer Satisfaction, Customization, and Assurance.
Furthermore, LAZADA also potentially lose their customer that also do online shopping in Customer to Customer E-Commerce, since there are 171 or 46% people who do shopping in Tokopedia, 136 or 37% people in Bukalapak, and 115 or 31% people in S hopee.

CONCLUSION
The paper examines the determinants and intermediaries' variables that have signi icant in luence on customer loyalty of LAZADA Jakarta, Indonesia. In this research, it is found Website Design, Responsiveness, and Assurance has signi icant effect to Customer Satisfaction, Customization and Assurance has signi icant effect to Customer Trust, Customer Satisfaction has signi icant effect to Customer Trust and Customer Loyalty, and Customer Trust has signi icant effect to Customer Loyalty.
There are some limitations to this paper. Firstly, this research is only a perspective from Jakarta Millennial perspective, which makes it not general and get the perspective from people outside Jakarta and people who not in Millennial generation. Secondly, the study has lack of several other variables, since the focus more on 4 variables, which are Service Quality, Satisfaction, Trust, and Loyalty. Despite the limitation, the study recommends for e-commerce industry to carefully maintain both technical and service deliveries as important components to assure the customer loyalty.
Furthermore, considering those limitations, the future recommendation is to retain other sample groups and to extend other variables that can in luence customer loyalty. Besides, for the future research, it is also possible if the case is changed into Consumer to Consumer E-Commerce model to get more insight and information, such as Tokopedia, Bukapalak, and Shopee that become the rival for LAZADA.