INFLUENCE E-SERVICE QUALITY TOWARDS CUSTOMER RETENTION AND WORD OF MOUTH (WOM) MEDIATED BY CUSTOMER SATISFACTION

ABSTRACT


Introduction
One of the times of the digital society in Indonesia is shown by the rapid development of information and communication technology, especially the telecommunications industry over the past few years, especially the use of the internet through cellular phones which is growing rapidly.Telecommunications have eliminated distance boundaries and reduced the differences between people in urban and rural areas.Time differences, location distance, as well as heterogeneity of population characteristics are no longer obstacles in the speed of information dissemination.The internet is part of telecommunications, playing an important role in daily life activities in Indonesian society.People use the internet to support activities such as working, studying, accessing various digital platforms, shopping online and socializing with each other through social media.
In January 2023, Indonesia recorded the number of internet users reaching 212.9 million, reaching 77% of the total population and showing a significant increase from the previous year.Most users, namely 98.3%, access the internet through mobile devices with an average daily usage of 7 hours 42 minutes.This situation opens up great opportunities for companies, including conventional retailers such as Alfamart, to capitalize on this trend by launching online shopping services.
Alfamart responded to the trend with the launch of an online shopping application called Alfagift in 2015.Although this application has improved and added features, challenges in socialization and consumer education remain.In addition, the COVID-19 pandemic has changed consumers' shopping behavior, forcing them to switch to online services, creating new challenges and opportunities for companies like Alfamart in running their businesses.The increase in the use of Alfagift for online shopping which now reaches 45 million transaction receipts indicates a positive boost, although it also brings negative complaint issues that need to be addressed further to maintain and improve service quality.
Complaints that come through Alfacare related to Alfagift cause disappointment in consumers, because consumers do not get what they want.The impact of consumer disappointment causes consumers not to want to shop anymore or reduce their transactions at Alfagift specifically and can have an indirect impact on Alfamart.The greater impact is on consumers who are disappointed but do not submit negative complaints to Alfacare, but spread their bad experiences to others through direct storytelling or consumer social media which can cause a bad image of Alfagift & Alfamart.
Based on negative complaints, it is necessary to improve the business process at Alfagift through a structured system.In this study using e-service quality as an improvement system at Alfagift.Broadly speaking, service is a positive impact service provided by the company to consumers.Service quality or service quality comes from consumer evaluation of service quality is a function of the magnitude and direction of the gap between consumer expectations of service and consumer assessment of the actual service delivered (Schiffman, Kanuk, & Wisenblit, 2010).Service quality starts from consumer needs and ends with customer satisfaction and a positive perception of service quality.Continuous improvement of service quality is not a cost, but an investment to increase greater profits.
The study outlines a number of hypotheses based on findings from a number of related studies.It was found that service quality, especially in the dimensions of design, fulfillment, security & service, is positively related to repurchase intention (Garcia, Freire, Santos, & Andrade, 2020) and has a direct influence on consumer satisfaction and retention (Arshad Khan & Alhumoudi, 2022).However, other studies have shown that e-service quality does not significantly affect repurchase intention (Ginting, Chandra, Miran, & Yusriadi, 2023).Furthermore, it was found that e-service quality has a positive and significant influence on customer retention (H1), customer satisfaction (H2), and word of mouth (H3), with customer satisfaction mediating the relationship between e-service quality and customer retention (H6) and word of mouth (H7).These findings support an understanding of the importance of service quality in influencing consumer satisfaction, retention, and word of mouth in the online retail industry.
This study aims to investigate various significant relationships in the context of the online retail industry.The focus includes analyzing the positive influence of E-Service Quality on Customer Retention, Customer Satisfaction, and Word of Mouth, as well as identifying the positive influence of E-Service Quality on Customer Satisfaction.Furthermore, this study looked for the positive influence of Customer Satisfaction on Customer Retention and Word of Mouth.In addition, the study will also consider the effect of E-Service Quality on Customer Retention and Word of Mouth by mediating through Customer Satisfaction.With a deep understanding of these dynamics, this research will provide valuable insights into the management and development of strategies in the online retail industry.This research has three aspects of benefit.First, theoretical contributions by enriching marketing literature on the relationship of electronic service quality to customer retention and word of mouth in the online retail industry.Second, it provides practical guidance for managers to manage those factors in business planning.Finally, it has the potential to influence retail industry regulation to support sustainable growth.

Materials and Methods
This study aims to measure the relationship between electronic service quality variables (E-service quality) as independent variables, especially efficiency, reliability, responsiveness, fulfillment, and contact, with consumer performance variables, such as consumer satisfaction, consumer repurchases, and word of mouth (recommendations from consumers).This study used quantitative methods with survey techniques using questionnaires.The research sample was 190 respondents domiciled in Jabodetabek.The research instrument is prepared based on the operationalization of variables measured using the Likert scale.To test the validity, reliability, and quality of instruments, statistical tests such as the Outer Model and the Inner Model are used.The results of this statistical analysis will be used to determine the extent to which electronic service quality variables contribute to consumer performance, as well as whether there is a mediating influence of consumer satisfaction in the relationship.The data were analyzed using the Partial Least Square Structural Equation Modeling (PLS-SEM) method.
In addition, statistical analysis in this study also included hypothesis testing performed to determine the significance of the relationship between the independent variable (quality of electronic services) and the dependent variable (consumer performance).In addition, mediation testing will be conducted to check whether consumer satisfaction plays a role as a mediator in the relationship between electronic service quality and consumer performance.This statistical analysis technique aims to provide a deeper understanding of the extent to which these factors influence consumer behavior, especially in the context of the Alfagift application in the Jabodetabek area.

Characteristics of Respondents
The data used in this study were 202 respondents within 1 month.The demographics of 202 respondents are 202 or 100% have shopped with Alfagift using the delivery method and 128 or 63.37% using the pickup method.As many as 89 or 44% of respondents routinely shop every 1 month and with a nominal value of less than Rp 100,000 in 1 month 90 or 44.5% of respondents.
Of the 202 respondents, the majority are domiciled in Tangerang, 67 or 33% and male as many as 109 or 54%.The average age of the respondents was 26-30 years old as much as 54 or 26.7%, the last S1 education was 108 or 53.5% and 164 or 81% worked as private employees.

Instrument Test: Validity & Reliability Validity Test (Validity)
Based on the PLS method that the loading factor of each indicator used to measure the variables of e-service quality, customer satisfaction, customer retention and word of mouth is greater than 0.70.This proves that the indicators used to measure the variables of e-service quality,  The first stage of convergent validity test and the second discriminant validity test have been carried out for indicators in the variables e-service quality, customer satisfaction, customer retention and word of mouth and meet the validity testing criteria .

a. Reliability Test
Reliability testing using internal consistency reliability evaluated using Cronbach's alpha and composite reliability values with a minimum reliability coefficient value greater than 0.70 or 0.60 is still accepted (Hair, Black, Babin, & Anderson, 2019).Cronbach' s Alpha testing on each variable of e-service quality, customer satisfaction, customer retention and word of mouth met the criteria of Cronbach's Alpha value of >0.70.This shows that the variables used are reliable or meet Cronbach's Alpha test.Based on the table below composite reliability testing on each variable e-service quality, customer satisfaction, customer retention and word of mouth meets the criteria of composite reliability value >0.70.This shows that the variables used are reliable or meet the composite reliability test.Cronbach's Alpha tests and composite reliability tests have been performed for indicators in the variables e-service quality, customer satisfaction, customer retention and word of mouth and meet reliability testing criteria.

Structural Model Testing (Inner Model)
Structural model testing (inner model) using bootstrapping calculations.The evaluation is carried out by looking at the criteria for the coefficient of determination test value (R-Square), effect size test (f^2)and significance value (hypothesis test).Test Coefficient of Determination (R Square) R-square testing is used to assess the magnitude of the influence of the independent variable on the dependent variable.The interpretation of R-square values is 0.75 (strong), 0.50 (moderate), and 0.25 (weak).Effect size testing to show whether exogenous and endogenous variables have a weak, medium or large influence based on testing.A f 2 value of 0.02 means weak influence, 0.15 means medium influence and 0.35 means large influence.(Ghozali, 2016) Table 7  weak influence.Customer satisfaction variables on customer retention and word of mouth have a medium influence.

Significance Test (Hypothesis Test)/Direct Effect
Significance test testing or hypothesis testing is obtained through procedures bootstrapping and can be seen in the table path coefficients value p-values dengan level of significant with a 95% confidence level or margin error as big as 5% (α = 0,05) and test t-statistics with the following hypothesis: H0: The independent variable has no positive and significant effect on the dependent variable.H1: The independent variable has a positive and significant effect on the dependent variable.The criteria for testing the hypothesis in this study are as follows: If the p-value >0.05 or t-statistics <1.96 then H1 is rejected and Ho ≤0.05 and ≥1.96 then H1 is accepted and Ho is rejected.≤0.05 dan ≥1.96 then H1 accepted and Ho rejected.The first hypothesis (H1) states that "e-service quality has a positive and significant effect on customer retention", based on testing the first hypothesis (H1) is accepted because it meets the criteria of T statistics value is 2.741 greater than T table 1.96 and P-value is 0.006 smaller than 0.05 (significant), while A positive coefficient value of 0.231 means that if the E-service quality variable is increased by 1 unit, it results in an increase in the customer retention variable by 0.231 units.This is also supported by research conducted by (Arshad Khan & Alhumoudi, 2022;Martínez., 2019) that service quality has a significant direct influence on consumer satisfaction and consumer retention.So based on this hypothesis, Alfagift can increase customer retention both shopping with delivery and pick up methods by adding e-service quality variables using 5 dimensions, namely efficiency, reliability, responsiveness, fulfilment & contact in the process of improving its business.

Influence E-Service Quality towards Customer Satisfaction
The second hypothesis (H2) states that "e-service quality has a positive and significant effect on customer satisfaction", based on testing the second hypothesis (H2) accepted shown by the value of T statistics is 27.718 greater than T table 1.96 and the value of P-value is 0.000 less than 0, 05 (significant), while the positive coefficient value of 0.859 means that if the e-service quality variable is increased by 1 unit, it results in an increase in the customer satisfaction variable by 0.859 units.This is also supported by research conducted by (Venkatakrishnan, Alagiriswamy, & Parayitam, 2023) that e-service quality is positively related to customer satisfaction.Based on this hypothesis, it can be seen that the contribution of the e-service quality variable as an independent variable is very high, which is 85.9% in the dependent variable, namely customer satisfaction.Customer satisfaction has a positive impact on companies such as switching barriers, switching costs, customer loyalty, e-ISSN: 2723-6692  p-ISSN: 2723-6595 increasing market share, and reducing costs so that in the long run the company will survive, be able to compete and generate profits.

The Effect of E-Service Quality on Word of Mouth
The third hypothesis (H3) states that "e-service quality has a positive and significant effect on word of mouth" based on testing the third hypothesis (H3) is accepted as shown by the value of T statistics is 3.753 greater than T table 1.96 and the value of P-value is 0.000 less than 0.05 (significant), while the value of A positive coefficient of 0.313 means that if the e-service quality variable is increased by 1 unit, it results in an increase in the word of mouth variable by 0.313 units.This is also supported by research conducted by (Zia, Rafique, Rehman, & Chudhery, 2022) and (Vukethwele & Chinomona, 2023), namely e-service quality has a significant effect on word of mouth and customer perceived value.Based on this hypothesis, it can be interpreted that adding eservice quality variables can make consumers tell the shopping experience at Alfagift to other consumers so that the added value for the WOM variable.
Influence Customer Satisfaction towards Customer Retention.
The fourth hypothesis (H4) states that "customer satisfaction has a positive and significant effect on customer retention" based on testing the fourth hypothesis (H4) is accepted shown by the value of T statistics is 7.886 greater than T table 1.96 and the value of P-value is 0.000 less than 0.05 (significant), while the value of A positive coefficient of 0.673 means that if the customer satisfaction variable is increased by 1 unit, it results in an increase in the customer retention variable by 0.673 units.
This is also supported by research conducted by (Ananda, Hanny, Hernández-García, & Prasetya, 2023) that consumer satisfaction (customer satisfaction) is positively related and affects consumer repurchase (repurchase intention) in multichannel retail.The relationship between the influence of customer satisfaction on customer retention is quite high at 67.3%, which means that increasing customer satisfaction can affect customer retention at Alfagift.Satisfied consumers will spend more and buy more regularly and do not easily move to other applications or competitors.

Influence Customer Satisfaction towards Word of Mouth
The fifth hypothesis (H5) states that "customer satisfaction has a positive and significant effect on word of mouth" based on testing the fifth hypothesis (H5) is accepted as shown by the value of T statistics is 7.151 greater than T table 1.96 and the value of P-value is 0.000 less than 0.05 (significant), while the value of A positive coefficient of 0.555 means that if the customer satisfaction variable is increased by 1 unit, it results in an increase in the word of mouth variable by 0(Rizal, Yussof, Amin, & Chen-Jung, 2018).Consumer satisfaction (customer satisfaction) is positively related and affects (word of mouth).The relationship between the influence of customer satisfaction on word of mouth is quite high at 55.5% which means that consumers who are very satisfied (satisfaction) will buy the same products and brands, provide positive promotions and encourage word of mouth (WOM) to others and will become consumers for life.
Influence E-Service Quality towards Customer Retention Mediated Customer Satisfaction.
The sixth hypothesis (H6) states that "E-service quality to customer retention mediated customer satisfaction" based on testing the sixth hypothesis (H6) is accepted shown by the value of T statistics is 7.204 greater than T table 1.96 and the value of P-value is 0.000 less than 0.05 (significant), while the positive coefficient value of 0.477 means that if the E-service quality variable is increased by 1 unit, it results in an increase in the customer retention variable mediated by the customer satisfaction variable of 0.477 units.This is also supported by research conducted by (Khoo, 2020) that service quality is an important element that can cause revisit intention mediated by customer satisfaction.To increase customer retention, you can use e-service quality variables and customer satisfaction variables in the Alfagift application.The e-service quality dimension used will improve the consumer experience of shopping at Alfagift and is expected to increase customer satisfaction and not cause consumer complaints which will ultimately increase customer retention and have an impact on increasing sales, profits and resistance to competitors that are beneficial to the company in the future.

Influence E-Service Quality towards Word of Mouth Mediated Customer Satisfaction
The seventh hypothesis (H7) states that "e-service quality has a positive and significant effect on word of mouth mediated customer satisfaction" based on testing the seventh hypothesis (H7) accepted indicated that the value of T statistics is 7.956 greater than T table 1.96 and the value of P-value is 0.000 smaller from 0.05 (significant), while the positive coefficient value of 0.578 means that if the E-Service Quality variable is increased by 1 unit, it results in an increase in the Word of Mouth variable mediated by the Customer Satisfaction variable of 0.578 units.This is also supported by research conducted by (Stribbell & Duangekanong, 2022) that service quality is an important element that can cause positive word of mouth mediated by customer satisfaction.Alfagift can increase word of mouth by using the dimensions of e-service quality and customer satisfaction because it has a fairly high influence of 57.8%.The e-service quality dimension used will improve the consumer experience of shopping at Alfagift, hopefully it can increase customer satisfaction and there will be no business failures that cause consumer disappointment which can have an impact on consumer complaints.When consumers are satisfied with Alfagift, they can spread positive WOM which can be beneficial for the company such as bringing new consumers to Alfagift.

Conclusion
This research concludes that E-Service Quality plays an important role in influencing Customer Retention, Customer Satisfaction, and Word of Mouth in the context of the online retail industry.The quality of electronic services has a positive and significant impact on customer retention, retaining customers, and influencing their satisfaction.In addition, the quality of electronic services also plays a role in the development of brand reputation through positive recommendations and word-of-mouth influence.The results also show that Customer Satisfaction has a central role in linking the quality of electronic services with Customer Retention and Word of Mouth.Theoretical suggestions include expanding the study by considering the broader dimensions of E-Service Quality variables and additional mediating variables, while practical suggestions emphasize the importance of taking E-Service Quality into account in a company's strategic planning, effectively managing customer reviews, and strengthening consumer protection and regulations that protect companies.Thus, this research provides valuable insights for policymakers and companies in understanding the important role of e-service quality in the online retail industry.
e-ISSN: 2723-6692  p-ISSN: 2723-6595 customer satisfaction, customer retention and word of mouth are valid or have met the criteria of the convergent validity test and meet the first stage of the validity test.

Figure 1
Figure 1 Test ModelThe results of the hypothesis based on the figure above in detail can be seen in the following table:Table8Results of T Statistics and P Values with smartPLS (direct effect) e-ISSN: 2723-6692  p-ISSN: 2723-6595

Table 1 Uji Outer Loading variabel E-Service Quality, Customer Satisfaction, Customer Retention, and Word of Mouth
-ISSN: 2723-6692  p-ISSN: 2723-6595 retention and word of mouth were valid or had met the criteria of the convergent validity test and met the first stage of the validity test.Tabel The second test using Average Variance Extracted (AVE) greater than 0.50, proved that the indicators used to measure the variables of e-service quality, customer satisfaction, customer e

2 Uji Average variance extracted (AVE) variabel E-Service Quality, Customer Satisfaction, Customer Retention, and Word of Mouth
The next test, test discriminant validity using cross loading testing.The indicators used in eservice quality, customer satisfaction, customer retention and word of mouth have higher loading compared to other indicators in different variables and meet the criteria that cross loading >0.70.

Table 6
When viewed the results of processing data in table 6, the relationship between constructs based on the R-square value can be explained as follows: 1.The R-Square value in the Customer Satisfaction variable is 0.738 greater than 0.50 which can mean that the model is moderate.The R-Square value means that 73.8% of the Customer Satisfaction variable can be influenced by the E-Service Quality variable.While the remaining 26.2% was influenced by other variables outside the variables studied.2. The R-Square value of the Customer Retention variable is 0.774 greater than 0.70 which means that the model is strong.The R-Square value means that 77.4% of Customer Retention variables can be influenced by E-Service Quality variables.While the remaining 22.6% was influenced by other variables outside the variables studied.3. The R-Square value of the Word of Mouth variable is 0.704 greater than 0.50 which means that the model is moderate.The R-Square value means that 70.4% of Word of Mouth variables can be affected by E-Service Quality variables.While the remaining 29.6% was influenced by other variables outside the variables studied.

Effect Size f test results
Based on table 7, it is shown that the e-service quality variable and the customer satisfaction variable have a large influence, while the customer retention and word of mouth variables have a e-ISSN: 2723-6692  p-ISSN: 2723-6595

Table 8 Results of T Statistics and P Values with smartPLS (direct effect)
.96, then the significant value (Solling Hamid & M Anwar, 2019) ≤0.05 and bootstrapping and can be seen in the table total indirect effects Value column p-values dan t-statistics.≤0.05 and p-value t-statistics ≤0.05 dan ≥1.96, hence the significant value (Solling Hamid & M Anwar, 2019)