E-Trust and Website Quality Can Improve E-Loyalty Through E-Satisfaction

This research aims to know the effect of E-Trust and Website Quality to E-Loyalty through E-Satisfaction. The research population is obtained from Tokopedia users who transact at least 2 times. The population sample taken in the research is about 156 respondents. Within the data collection, this research employs questionnaire. From the path analysis method, it shows that E-Trust and Website Quality directly affect to E-Loyalty or indirectly affect to E-Satisfaction . This research concludes that E-Trust and Website Quality both directly and indirectly affect to build the E-Loyalty through their E-Satisfaction. In last, based on the research result, it recommends to create new innovations to boost E-Satisfaction.

H1: E-Trust has a positive and significant effect on E-Satisfaction H2: Website Quality has a positive and significant effect on E-Satisfaction H3: E-Satisfaction has a positive and significant effect on E-Loyalty H4: E-Trust has a positive and significant direct effect on E-Loyalty H5: Website Quality has a positive and significant direct effect on E-Loyalty H6: Website Quality has a positive and significant indirect effect on E-Loyalty through E-Satisfaction H7: E-Trust has a positive and significant indirect effect on E-Loyalty through E-Satisfaction

Research Method
This research used quantitative research method with descriptive and Explanatory research approaches.. The method of quantitative research refers to the methods to clarify the certain theories by examining the relation between variables (Creswell, 2013). Descriptive approach was used to describe the circumstances of each variable of E-Trust (x1), Website quality (x2), E-loyalty (y), dan E-satisfaction (z). Explanatory Research is a research which aims to explain the ralation between variable and phenomenon of research (Cooper & Schinder, 2011).

Population and Sample
Population is as generalization area consisting of abject or subject which has certain quality and characteristic determined by researcher to be studies, processed and drawn the conclusion (Sugiyono, 2014). The population in this study were unknown numbers of Tokopedia users. The sample used in this study were 130 people who had made product purchase transactions at Tokopedia at least twice. The sampling method used purposive sampling. Data were obtained from distributing questionnaires using Google Form media. The result of trial showed that questionnaire is valid and reliable.

Data Analysis
After the data are collected, the researcher will analyze the data. This research analysis uses method of descriptive and inferential analysis. In the data analysis process, the researcher describes and figures the collected data without having intention to draw general conclusion (Sugiyono, 2014). This descriptive analysis refers to the descriptive variable of research: E-Trust (X1), Website Quality (X2), Customer's Satisfaction (Z), and Customer's Loyalty (Y). Inferential analysis is a research method that analyze the sample data and make generalization (valid generally) in population.  Tokopedia is able to secure customer transaction from the fraud 3,40 ET8

Result of Descriptive Analysis
Tokopedia will responsible over he problems in transaction 3,47 ET9 The delivery is not more than the estimation 3,72 ET10 Received product is accordance with the description of seller 3,79 ET11 There is return of goods when the product received is wrong / damaged 3,37 Total of average 3.   (X2), statistically affect significantly to the E-Satisfaction (Z). This is indicated by the Sig Value. Each value is smaller than alpha 5%, 0.000. Z = ρzx1+ ρzx2+ε2 = 0,513x1 + 0,361x2 + 0,625ε2 Second Comparison E-Trust (X1), Website Quality (X2), through E-Satisfaction (Z) simultaneously to E-Loyalty (Y). Here are the discussion about the line analysis related to the variables.   , T test / partial, it is seen that the variable of E-Trust (X1) has value 0.000 higher than alpha 5% which means Sig Value or Ho is rejected and Ha is accepted. E-Trust (X1), Website Quality (X2) and E-Satisfaction (Z) affect positively and significantly to the E-Loyalty (Y), which is partially shown from each Sig. Value higher than Alpha 5%, 0.000 and 0. 000. Y = ρyx1+ ρyx2+ ρyz + ε1 = 0,233x1 + 0,243x2 + 0,483z + 0,519ε1  Vol.11, No.12, 2019 SEa = standard error of estimation from the effect of independent variable toward mediation variable SEb = Standard error of estimation from direct effect of E-Treust toward E-Satisfaction.   1) The first Hypothesis test. The test of direct effect of E-Trust toward E-Satisfaction was obtained that t-count (8,232) was bigger than t-table (1,976) or Sig. (0,000) was smaller than alpha 5% (0,050). This shows that there was a significant positive effect on E-Trust on E-Satisfaction. 2) The second hypothesis test. The test of direct effect of Website Quality toward E-Satisfaction was obtained t-count (5,793) was bigger than t-table (1,976) or Sig. (0,000) was smaller than alpha 5% (0,050). This shows that there was a significant positive effect of Website Quality on E-Satisfaction.
3) The third hypothesis test. In the test direct effect of E-Satisfaction toward E-Loyalty, it was obtained that tcount (7.172) was bigger than t-table (1.976) or Sig. (0,000) was smaller than alpha 5% (0,050). This shows that there was a significant positive effect of E-Satisfaction on E-Loyalty. 4) The fourth hypothesis test. In the test of direct effect of E-Trust toward E-Loyalty, it was obtained that tcount (3.737) was bigger than t-table (1.976) or Sig. (0,000) was smaller than alpha 5% (0,050). This shows that there was a significant positive effect of E-Trust toward E-Loyalty. 5) The fifth hypothesis test. In the test of direct effect of Website Quality toward E-Loyalty, it was obtained that t-count (4,246) was bigger than t-table (1,976) or Sig. (0,000) was smaller than alpha 5% (0,050). This shows that there was a significant positive effect of Website Quality toward E-Loyalty. 6) The sixth hypothesis test, In the test of direct effect of T-Trust toward E-Loyalty through E-Satisfaction, it was obtained that t-count (6,505) was bigger than t-table (1,976) or Sig. (0,000) was smaller than alpha 5% (0.050), besides, the indirect effect of E-Trust on E-Loyalty through E-Satisfaction was resulted from the multiplication of the value of beta E-Trust (X1) toward E-Satisfaction (Z) with value beta E-Satisfaction (Z) toward E-Loyalty (Y) which was 0.513 x 0.483 = 0.247779. Based on the calculations above, it is known that the indirect effect value is 0.247779 which means that the value of indirect influence is bigger than the value of direct influence. These results showed that there was a significant indirect effect on E-Loyalty through E-Satisfaction as a intervening variable, so the sixth hypothesis was accepted. 7) The seventh hypothesis test. In the test of indirect Website Quality toward E-Loyalty through E-Satisfaction, it was obtained that t-count (5,415) was bigger than t-table (1,976) or Sig. (0,000) was smaller than alpha 5% (0,050). Other than that, Website Quality's indirect effect toward E-Loyalty through E-Satisfaction resulted from the multiplication of the beta value of Website Quality (X2) toward E-Satisfaction (Z) with beta value E-Satisfaction (Z) toward E-Loyalty (Y) was 0.361 x 0.483 = 0.174363. Based on the calculations above, it is known that the indirect effect value was 0.174363 which meant that the value of indirect effect was greater than the direct effect value. These results indicated that there was a significant indirect effect of Website Quality on E-Loyalty through E-Satisfaction as an intervening variable. Thus, the seventh hypothesis was accepted.

Discussion a. The effect of E-Trust toward E-Satisfaction
The result of the research showed that there is positive and significant effect of E-Trust (X1) toward E-Satisfaction. It was proven that t-count (8,232) was bigger than t-table (1,976) or Sig. (0,000) was smaller than alpha 5% (0,050). it indicated that the higher the e-trust or the level of consumer confidence in the site, the higher the e-satisfaction or customer satisfaction with the site was. Conversely, the lower the e-trust, the lower the e-satisfaction was. This study supported previous studies of  and Liao and Zhong (2013) which explained that there was a significant influence between e-trust and e-satisfaction. b. The effect of Website Quality toward E-Satisfaction The result of the research showed that there was positive and significant website quality (X2) toward E-Satisfaction (Z). This statement was proven by the value of t-count (5.793) which was than t-table (1.976) or Sig. (0,000) was smaller than alpha 5% (0,050). Thus, this study supported the previous study  which stated that there was the positive and significant effect of website quality on customer satisfaction. Research was in line with Chen, et al. (2012) that website quality affected positively and significantly toward satisfaction. c. The effect of E-Satisfaction toward E-Loyalty The result of the research showed that there was positive and significant effect of E-Satisfaction (Z) toward E-Loyalty (Y). This statement was proven by the value of t-count (7.172) which was bigger than t-table (1.976) or Sig. (0,000) was smaller than alpha 5% (0,050). Thus, this study supported the previous research examined that the E-Satisfaction variable had a significant and positive effect toward E-Loyalty. The results of this study were in line with the opinion of Miguens et al. (2017) which stated that E-Satisfaction had a positive and significant effect toward E-Loyalty. The results of this study were also consistent with the research of  which explained that E-loyalty was the most significant variable in online retailing that received the most effect from other variables, especially E-Satisfaction. Research from Anderson and Srinivasan (2011) also stated that e-satisfaction had a positive effect toward e-loyalty. d. The direct effect of E-Trust toward E-Loyalty In the figure of analysis, it was shown that there was direct effect of E-Trust toward E-Loyalty. It was obtained that tcount (3.737) was bigger than t-table (1.976) or Sig. (0,000) was smaller than alpha 5% (0,050). Therefore, this study supported previous research of Miguens et al. (2017) stated that E-Trust had a direct effect toward E-Loyalty. The results of the study were also in line with the research of  and Siagian et al. (2014) which showed that E-Trust had a significant effect toward E-Loyalty. e. The direct effect of Website Quality toward E-Loyalty In the figure of analysis, it showed that there was direct effect of Website Quality toward E-loyalty. It was obtained that t-count value (4,246) was bigger than t-table (1,976) or Sig. (0,000) was smaller than alpha 5% (0,050). Therefore, this study supported previous researchof Siagian et al. (2014) andO'cass &Carlson (2012) stating that Website quality had a direct effect toward E-Loyalty. The results of the study were also in line with the results of a study from Puspitasari et al. (2013) and Bai et al. (2008) which stated that website quality proved to have a direct effect toward customer loyalty. f. The indirect effect of Website Quality toward E-Loyalty through E-Satisfaction In the figure of analysis, it showed that there was indirect effect of Website Quality through E-Satisfaction. It was obtained that t-count (6,505) was bigger than t-table (1,976) or Sig. (0,000) as smaller than alpha 5% (0.050) and based on the multiplication calculation between beta values, it was known that the indirect effect value was 0.247779. it meant that the value of indirect influence was bigger than the value of the direct effect. It is concluded that there were significant indirect effects of E-Trust toward E-Loyalty through E-Satisfaction. These results indicated that Website Quality had an indirect influence on E-Loyalty through E-Satisfaction, or it can be concluded that E-satisfaction was a variable that mediated Website Quality toward E-Loyalty. This was in line with the research conducted by Bai et al. (2008) which revealed that the quality of a website affected loyalty through satisfaction.

g. The Indirect effect of E-Trust toward E-Loyalty through E-Satisfaction
In the analysis figure, it showed that there was indirect effect of E-Trust toward E-Loyalty through E-Satisfaction. It was obtained that t-count (5,415) was bigger than t-table (1,976) or Sig. (0,000) as smaller than alpha 5% (0.050). Based on the multiplication calculation between beta values, it is known that the indirect effect value was 0.174363 which meant that the value of indirect influence was bigger than the value of direct influence. Therefore, it is concluded that there were significant indirect effects on Website Quality towards E-Loyalty through E-Satisfaction. This showed that E-Satisfaction could be a variable that mediated E-Trust and E-Loyalty. The results of this study supported the study of Fitriani (2018) that there was an indirect influence between E-Trust toward E-Loyalty through E-Satisfaction.

Conclusion
Based on the results and discussion of the research about the effect of E-Trust, Website Quality toward E-Loyalty through E-Satisfaction as intervening variable of users of E-commerce C2C Tokopedia, it can be concluded as follow: a. E-trust and Website Quality affect positively and significantly toward E-Satisfaction. b. E-Trust, Website Quality and E-Satisfaction affect positively and significantly toward E-Loyalty. c. E-Satisfaction can mediate directly and indirectly between E-Trust and Website Quality toward E-Loyalty.

Recommendation
a. Tokopedia is expected to be able to maintain and improve Website Quality by mentioning product information in detail and easy to understand in website, improve security regarding customer personal information, design menus that are easier to understand, improve website responsiveness, display more attractive and comfortable visual designs to see, complete transaction guidelines, and improve customer service. This is because the website quality variable has a significant effect toward customer E-Satisfaction and E-Loyalty. b. Furthermore, it is to make E-trust and E-Satisfaction is going to increase and it gives impact on the enhancement of E-Loyalty. Tokopedia is expected to always improve the expertise and skills in serving customers, immediately be responsible if there is an error in the transaction and make every effort to listen and consider the suggestions submitted by the customer. c. The further research is suggested to use other factors which can affect or add other variables such as word of mouth, repurchase intention or customer perception.