The Influence of Electronic Service Quality and Electronic Recovery on Online Re-Purchase Intention: Role of E-Loyalty as Intervening Variable

This study intends to examine how online service quality and online recovery affect online loyalty and how they affect consumers' willingness to repurchase online. In this study, a non-probability sampling strategy with a purposeful sampling procedure was applied. A sample of 100 respondents who had purchased products from the marketplace were given questionnaires in order to obtain the data. The data in this study are analyzed using the Partial Least Square (PLS) method. The findings of this study demonstrate that online repurchase intention and electronic service quality have an impact on online loyalty. This study also discovered that electronic service quality has an impact on online repurchase intention via electronic loyalty. Through electronic loyalty, electronic recovery also has an impact on online repurchase intention.


Introduction
The world has entered a new era because to the quick advancement of information technology, known as the internet era. Since the invention of the computer, this data-processing medium has become the first medium used to access the internet. Then proceed with the invention of mobile phones, which are also a medium for accessing the internet. In January 2019, there were 4,388 billion people worldwide using the internet; the number has increased by 367 million since January 2018 [1]. This increase shows that worldwide internet penetration has reached 57%, or you could say there are more people using the internet than those who don't use the internet. In Indonesia, there were 143.26 million internet users in 2017, or around 54.68% of the 262 million people in the country. Java Island has a penetration rate of 58.08%, followed by Sumatra Island (19.09%), Kalimantan Island (7.97%), Sulawesi Island (6.73%), Bali-Nusa Islands (5.63%), and Maluku-Papua Islands (2.49%) in terms of internet users. This fact can benefit entrepreneurs or online store owners in Indonesia, especially those on the island of Java. The most common use of the internet in the economic field in 2017 was by internet users looking for prices, namely 45.14%. use of the internet to make online purchases of 32.19% and online sales of 16.83%. It is clear that the interest of online buyers is quite high, even though more users use the internet to find prices [2]. Businesspeople who want to take their companies online now have a new chance thanks to the rapid expansion of the digital world [3]. As can be seen from the rising number of transactions between 2011 and 2015, Indonesia has already made significant progress in the growth of online retailers through internet media. The ease of shopping and the variety of types of products and services offered make online stores a preferred shopping place for Indonesians [4]. This makes many online shop sellers vying to attract the attention of consumers. One of the online stores in Indonesia that is growing rapidly is Sociolla. Within four years, e-commerce had caught up with its predecessors [5]. This online store is a beauty e-commerce platform that provides a wide selection of beauty merchandise from various brands to cater to the diverse needs of skincare, cosmetics, haircare, bodycare, beauty tools, and fragrances. The range of products available varies from domestic to international brands. There are five reputable beauty e-commerce websites in Indonesia, namely Sephora, Althea Korea, Allyoung, Sociolla, and Bensrub. Sociolla is not the initial beauty e-commerce website in Indonesia. Despite Bensrcub being established prior to this e-commerce platform, the growth rate of the latter surpasses that of Bensrcub. This is due to the fact that Bensrcub primarily offers products from the Western, Australian, and New Zealand regions. Meanwhile, the e-commerce platform offers a diverse range of local, Western, Korean, and Japanese products. The variety and suitability of these products for the Indonesian population make it more preferable, thus facilitating rapid development. Especially for Sephora, this beauty e-commerce company has been using the internet since 1999 [6]. In addition to a range of women's products, it also has products for men. skin care products for men. E-commerce does not only include photos and product prices but also a more in-depth explanation of the products being sold. Thus, consumers can see the benefits of each product and choose according to their needs. Prices for products sold in e-commerce are quite affordable, ranging from tens of thousands to hundreds of thousands of rupiah. But ecommerce also sells certain products whose prices can reach millions of rupiah. Each product in e-commerce has its own column, which contains reviews of the products they have used. In this column, consumers not only provide testimonials but also names, skin types, and ratings [7]. This rating is an average assessment of effectiveness, texture, packaging, and value for money given by consumers. The completeness of information on e-commerce websites is one way to fulfil consumer satisfaction. With the hope that consumers will continue to buy products through e-commerce [8]. E-commerce retail performance is considered good if it can make its customers loyal. Loyalty is an important attitude for e-commerce consumers to have. Customer loyalty is influenced by service recovery and quality in numerous ways [9]. Recovery has a direct correlation with loyalty, whereas quality only indirectly influences loyalty through perceived value and satisfaction. Previous research has demonstrated that website functionality and data security, which affect customer loyalty, have an impact on the quality of e-services. Customer satisfaction may also rise as a result of e-recovery, which is determined by responsiveness, payment, and communication. Online repurchase intents will rise as a result, eventually leading to long-term profitability. According to previous studies, "efficiency" is the aspect of e-service quality that has the biggest impact on eloyalty. Meanwhile, "contact" is the aspect of e-recovery that has the biggest impact on e-loyalty [10]. E-service quality differs from traditional services in that it depends on the exchange of information between customers and service providers. More than ever, offering online services is essential to retaining customers. Customers are persuaded to visit the website again by the sense of loyalty that is fostered by the company's level of service [11]. The effectiveness of the website's ability to support efficient and effective shopping, purchase, and delivery is what is meant by the term "electronic service quality." This definition makes it clear that the idea of electronic service quality encompasses both the pre-purchase and post-purchase phases, including ease of use, product information, order information, and the security of personal information. Because it discusses website services in response to issues or inquiries from users, service recovery is a crucial component of e-service. Service recovery refers to systematic efforts made by businesses to address issues brought on by service failures and keep consumers. Procedures for service recovery are crucial for businesses to ensure customer satisfaction and loyalty. The likelihood of a poor or inadequate service recovery is increased customer dissatisfaction. Customers may leave for competitors or spread unfavorable word of mouth as a result [12]. Even if there may be situational effects and marketing campaigns that have the capacity to impact consumer behaviour, consumer loyalty is a commitment to survive in depth by repeatedly repurchasing or resubscribing to chosen items or services in the future [13]. E-loyalty is a positive attitude towards e-business that leads to repeat purchasing behaviour and is often associated with online services. The consumer's intention to make a purchase from a website or their intention to visit a specific website again are both examples of this e-commerce loyalty. Eloyalty is described as a positive attitude towards customers displayed by an online business that promotes behaviour for recurrent purchases. This is acknowledged as a client commitment to a specific website. Repurchase intention can be defined as a person's assessment of whether to repurchase the specified service from the same business, taking into account both the current situation and potential future situations. The idea of repurchase intention was mostly borrowed and modified from marketing and social psychology [14]. Repurchase intention is defined by social exchange theory and the commitment investment model as the desire to maintain an existing connection, which is known as relationship maintenance [15]. Customer Repurchase Intention is cited as a crucial Defensive Marketing tactic that affects a company's ability to succeed. Instead of using an aggressive marketing strategy to try to win over new consumers and gain market share, businesses concentrate on existing customers to convince them to make repeat purchases [16]. By cultivating relationships with their current clientele, businesses can spend less on marketing campaigns than they would if they were trying to win over new ones. This will result in more repeat business. Repurchasing behaviour is a significant issue that most businesses are concerned about because it is significantly more expensive to get new consumers than to keep existing ones.

Research Methods
All individuals who have made e-commerce purchases, had issues with those purchases, and have made ecommerce purchases in the past make up the population of this study. In this case, the researcher does not categorise respondents by age or gender. The sampling method used in this inquiry was the Non-Probability Sampling Method. Purposive sampling was used to choose the sample for this study. Yamin and Kurniawan's (2011) hypothesis, which indicates that the number of samples employed in the Partial Least Square (PLS) application was very moderate, with a minimum of 30-100 examples, guided the sampling in this work. Answers to surveys given to consumers who made purchases on the website served as the basis for the primary data in this study. A Likert scale is the assessment tool employed in the questionnaire. Researchers gathered secondary data through reading books in libraries on and off campus, as well as on the internet and websites. employing the outer model, the inner model, and partial least squares (PLS) to analyze the data. The link between the indicator blocks and their latent variables is described by the outer model, also known as the measurement model. The construct validity and reliability of the instrument are examined through an analysis of the measurement model. The structural model, often known as the inner model, is a theory that explains how latent variables and their manifest variables interact. The R2 for the dependent construct is used in PLS to assess the structural model.

Results and Discussion
Based on 100 respondent data points obtained from the questionnaire, the characteristics obtained based on gender, age, last education, income, and length of subscription are as follows: There were 92 responders overall, with 92.2% of them being female. The number of male respondents was 8 with an 8% percentage. A total of 52 respondents (52%) were in the 17-24 age group, followed by 22 respondents (22%) in the 25-34 age group, 17 respondents (17%) in the 35-44 age group, and 9 respondents in the 45-plus age group. The percentage of respondents with the last education of high school or equivalent was 31 people (31%), the percentage of respondents with the last education of diploma was 14 people (14%), the percentage of respondents with the last education of S1 was 53 people (53%), and the percentage of respondents with the last education of S2/S3 was 2 people (2%). The percentage of respondents with an income of less than $2,000,000 was 35 people (35%), the percentage of respondents with an income of $2,000,000-5,000,000 was 41 people (41%), and the percentage of respondents with an income above $5,000,000 was 24 people (24%). The percentage of respondents with a subscription of less than 6 months was 30 people (30%), the percentage of respondents with a subscription of 1 year was 52 people (52%), and the percentage of respondents with a subscription of more than 1 year was 18 people (18%).
According to the respondents' answers, the majority of respondents (59.7%) selected "Agree" as their response. And of the 6 statement items, there was one that received the most positive responses, namely the FUL1 statement (e-commerce is able to send the right products to consumers) with a percentage of 94% (total ratings on scores of 4 and 5). While the statement items that have the lowest percentage are PRIV1 (e-commerce is able to protect consumers' personal information) with a percentage of 89%. These percentages' findings lead to the conclusion that one benefit of e-commerce is its capacity to send customers the proper products. And consumers who rate ecommerce services as protecting personal information make up 89% of the total number of respondents. The majority of respondents answered "agree," with a percentage of 54.7%. And of the six statement items, one received the most positive responses, namely CONT1 (e-commerce provides a contact I can contact if a problem occurs), with a percentage of 85%. While the statement item that has the lowest percentage is COMP1 (ecommerce gives me compensation in the form of money or product replacement if a problem occurs) with a percentage of 54%, It can be concluded that providing e-commerce contacts through various media is an advantage, although e-commerce must improve its services in providing consumers with product compensation.
According to the respondents' answers, the majority of respondents (59.8%) selected "Agree" as their choice. And of the 5 statement items, there was one that received the most positive responses, namely WOM2 (I gave positive comments about e-commerce to other users), with a percentage of 93%. While the statement items that have the lowest percentage are FBI1 (I consider e-commerce the first choice for buying products online) with a percentage of 58%, It can be concluded that positive comments from consumers regarding e-commerce to other users are an advantage, even though consumers who consider e-commerce their first choice for buying products are only 58% of the total number of respondents. The majority of respondents answered "agree," with a percentage of 62.7%. And of the 3 statement items, there was one that received the most positive responses, namely ORI1 (I will reuse services in e-commerce), with a percentage of 90%. While the statement item that has the lowest percentage is ORI3 (e-commerce is my first choice from other marketplaces) with a percentage of 70%. It can be concluded that 90% of respondents will reuse services in e-commerce, and only 70% of respondents will make beauty e-commerce their first choice from other marketplaces. The R2 value for electronic loyalty is 0.349, which shows that in moderation, the research model can explain the variability. Electronic service quality and electronic recovery are able to explain the variability of the electronic loyalty construct by 34.9%, and the remaining 65.1% is explained by other constructs that are not hypothesized in the model. And the R2 value for online repurchase intention is 0.572, indicating that the research model can moderately explain the variability. Electronic service quality and electronic recovery are able to explain the variability of the online repurchase intention construct of 57.2%, and the remaining 42.8% is explained by other constructs that are not hypothesized in the model. Because Q2 has a value above 0 and has values of 0.349 for electronic loyalty and 0.572 for online repurchase intention, it can be argued that Q2 predictive relevance is good. According to the Gof calculation, the value hits 0.560 and is regarded as having a large Gof. Therefore, it can be said that this research paradigm is reliable and effective. 4 The effect of electronic service quality on electronic loyalty has a t-statistic of 4.078 and a p-value of 0.000. The two-tailed t-statistic value is greater than 1.96. Therefore, it can be said that electronic loyalty is directly impacted by the quality of electronic service. Ha1 can thus be approved while Ho1 is refused. This is consistent with other studies, which found that the quality of an e-service can influence a customer's loyalty. Electronic recovery has a t-statistic of 2.930 and a p-value of 0.004 on electronic loyalty. The two-tailed t-statistic value is greater than 1.96. Therefore, it may be said that electronic loyalty is directly impacted by electronic recovery. Then Ha2 can be accepted and Ho2 rejected. The results of previous studies also showed the same thing. Both mentioned in their respective research results that e-recovery can affect consumer loyalty. This research is also supported by other research that states that e-recovery can directly affect e-loyalty. Electronic loyalty to online repurchase intelligence has a t-statistic value of 3.731 and a p-value of 0.00. The two-tailed t-statistic value is greater than 1.96. We can therefore draw the conclusion that electronic loyalty directly influences online repurchase intentions. Ha3 can thus be approved while Ho3 is refused. The findings of earlier studies support this finding, which states that e-loyalty can influence online repurchase intention and result in long-term advantages. The author's findings are also supported by other investigations. They conduct studies that demonstrate that loyalty influences repurchase intention in a favourable and meaningful way. The p-value is 0.000, and the t-statistic for the impact of online repurchase intention is 4.577. The two-tailed tstatistic value is greater than 1.96. Therefore, it may be inferred that intentions for online repurchase are directly influenced by the quality of electronic services. Ha4 can thus be approved while Ho4 is refused. The findings of this study are consistent with those of investigations carried out in China, Australia, Asia, Europe, and the United States. These two studies demonstrate how online repurchase intentions can be affected by the quality of an eservice. Online repurchase intention's electronic recovery t-statistic value was 3.391, and the p-value was 0.01. The two-tailed t-statistic value is greater than 1.96. Therefore, it may be said that online repurchase intentions are directly impacted by electronic recovery. Ho5 can then be refused while Ha5 is approved. These findings are consistent with research done in China. They conducted studies that shown recovery has a major impact on repurchase intention. Other studies that claim that recovery can alter repurchase intention lend support to this research. The p-value of the statistics on electronic service quality's effect on online repurchase intent is 0.010, and the tvalue is 2.588. Greater than 1.96 (two-tailed) is the t-statistic value. It is therefore possible to draw the conclusion that electronic service quality indirectly influences online repurchase intention through electronic loyalty. Ha6 can then be approved while Ho6 is disapproved. This study discovered that electronic service quality has an impact on online repurchase intention through electronic loyalty. The findings of this study are consistent with earlier research, which showed that high-quality e-services will directly influence consumer e-loyalty, and if this trend persists, it will lead to consumers making online repurchase intentions. Other studies that concur with this research show that loyalty is more influenced by effectiveness than by system accessibility or privacy. According to the research, service quality is more crucial than recovery since it is preferable to prevent service failure than to deal with it after it occurs. With a p-value of 0.014 and a t value of 2.471, the electronic recovery data on electronic loyalty and its effect on online repurchase intention are statistically significant. The two-tailed t-statistic value is greater than 1.96. We can therefore draw the conclusion that electronic loyalty, through electronic recovery, indirectly influences online repurchase intention. Ho7 can then be refused while Ha7 is approved. The findings of this study are consistent with earlier studies that have shown that a strong e-recovery will have a direct impact on e-loyalty. If this trend persists, it will lead to customers gaining long-term gains through online repurchase intentions. Other research that asserts that responsiveness has a stronger impact than compensation and contact backs up this study. Recovery does not have a stronger impact on loyalty than service quality, according to study. Furthermore, research has shown that there is no solid evidence linking contact with loyalty.

Conclusion
E-loyalty is positively and significantly impacted by the efficiency and privacy-driven e-service quality. Erecovery, which results from response, payment, and communication, considerably and positively affects e-loyalty. Online repurchase intention and e-loyalty, which includes word-of-mouth, future purchase intention, and complaining behavior, are positively and significantly related. Online repurchase intentions are positively and significantly impacted by e-service quality. Online repurchase intentions are positively and significantly impacted by e-recovery. Through e-loyalty, e-service quality has a favorable and considerable impact on online repurchase intention. Through e-loyalty, e-recovery has a favorable and considerable impact on online repurchase intentions. According to the findings of the available data, e-recovery has been shown to have an impact on e-loyalty. According to the survey's findings, an average of 54.7% of respondents said they were happy with the recovery services offered by e-commerce. However, around 30% of respondents said they were dissatisfied with the guarantees provided, and 31% of respondents said they were dissatisfied with the procedure for returning products in e-commerce. In addition, around 46% of respondents said they were dissatisfied with the compensation provided by e-commerce. This is something that must be addressed by improving recovery services. In this case, e-commerce parties must further clarify the guarantees provided and simplify the procedure for returning products to consumers. Especially at the beginning of creating an account or at the beginning of purchasing a product. From the results of the existing data, it also shows that around 25% of respondents were dissatisfied with the response from e-commerce when consumers had problems. This is something that must be addressed by improving recovery services. In this case, e-commerce parties must respond more quickly to consumer problems by improving standard operating procedures (SOP) and providing training to their employees.