The Impact Of Service Quality, Customer Satisfaction, And Trust On Customer Loyalty Among Brilink Agents In North Sumatra

. The shift of business into the digital era demands the banking industry to innovate in serving customers. One of the innovations by PT. Bank Rakyat Indonesia is the e-channel service named Brilink, which is spread across all regions. Local entrepreneurs are appointed as Brilink agents with the aim of providing banking services to individuals who are not yet served by banks, both unbanked and unbankable as stipulated in Article 16 of POJK No. 19/POJK.03/2014. Public distrust towards Brilink agents is attributed to the quality of service provided by these agents, which does not align with BRI's service standards, particularly in utilizing the Brilink application. Therefore, this research aims to examine the influence of the service quality, customer satisfaction, and trust variables on customer loyalty at Brilink agents in North Sumatra, Indonesia. This study employs a quantitative approach with a sample size of 400 respondent customers of Brilink agents. The research offers insights into the role of Brilink agents in marketing products, thus facilitating their understanding and utilization by customers. Furthermore, it holds social benefits in enhancing financial literacy and inclusion among the population in North Sumatra, ultimately contributing to welfare improvement and economic growth.

http://ijstm.inarah.co.id 1382 employment, mobile credit, gaming vouchers, OVO, Shopeepay top-ups, Gopay top-ups, Link Aja top-ups, Dana top-ups, loan applications, microloans, BRI finance, PNM payments, train tickets, and flight tickets.As of December 2022, the number of Brilink agents in North Sumatra reached 36,791, with transactions totaling 6.153 trillion (BRI Performance Report, 2022).While the types and services offered by Brilink agents continue to expand, with the range growing from 10 types of services between 2015 and 2018 to 30 between 2019 and 2022, the number of customers utilizing Brilink services decreased by 16.6% in 2022.Out of the 30 services, 69% of customer transactions were limited to transferring funds within BRI, making deposits, and withdrawing cash.BRI, as the owner of the Brilink product, has established service standards, conducts monthly evaluations and training for Brilink agents to market and serve customers.This indicates that the service quality provided by Brilink agents does not meet the set service standards.Despite the growing number of services, complaints against Brilink agents totaled 1,183 in 2022, [2]Public distrust towards Brilink agents is attributed to the mismatch between the service quality provided by these agents and the established BRI service standards.Consequently, customers still visit physical bank branches to conduct financial transactions, even if these branches are far from their residences, due to the higher satisfaction derived from excellent bank services.This has led to an increase in customer loyalty toward the bank but a decrease in loyalty toward Brilink agents.This issue cannot be overlooked as it impacts the bank's future performance in serving customers.Thus, there is a need for improvement to rectify the service quality of Brilink agents, fostering customer satisfaction and trust, ultimately creating customer loyalty.If a company can better satisfy customer needs compared to its competitors, it can easily cultivate customer loyalty, [3].Service quality centers around efforts to meet customer desires and needs and ensuring accuracy in delivering services to align with customer expectations.Trust is established when customers observe and acknowledge the abilities and responsiveness of employees, subsequently evaluating trust based on other dimensions of service quality separately, [4].

Population and Sample
The population used in this study encompasses all customers of Brilink agents in North Sumatra.This population is chosen as they possess knowledge and experience in transacting through Brilink agents.For this research, the sample size is determined using the Cochran formula, considering that the entire population of Brilink customers in Sumatra is 400 respondents.

Measurement Scale
The measurement scale used to assess perceived service quality, satisfaction, trust, and customer loyalty (both attitudes and behaviors) is derived from prior research.Respondents provide their answers to the questionnaire items using a standard five-point Likert scale.The study of service quality extensively examines five dimensions within the banking sector: reliability, assurance, tangibility, empathy, and responsiveness, as indicated in, [4], [5] Each dimension is evaluated using three questions.Customer satisfaction is gauged through overall satisfaction, as outlined in, [6], [7] and while Trust is assessed based on, [8].Customer loyalty is categorized into attitudes and behaviors.Customer loyalty attitudes encompass three question items, as specified in, [9] whereas customer loyalty behaviors consist of three question items as indicated in, [4].The measurement model illustrates how the observed variables indicate the latent variables to be measured.The structural model demonstrates the strength of estimations between latent variables and constructs, as elaborated in, [10].

Data Analysis Method
SEM PLS is used to analyze the data because the research model to be estimated is quite complex, encompassing variables of mediation and moderation.Additionally, SEM PLS does not require a minimum or maximum sample size, making it suitable even for small sample sizes.However, SEM PLS also performs well with large samples, [11].Source: Processed data (2023) In this research, the relationships between variables in the proposed model will be examined through path analysis, facilitated by the SmartPLS software.The path analysis process using the SmartPLS software involves evaluating the model's overall fitness.During the outer model testing phase, the validity and reliability of constructs are assessed, and during the inner model phase, the research hypotheses are put to the test.

Outer Model Testing
During this assessment, an indicator is deemed valid if its loading factor > 0.7, and every construct is expected to possess an Average Variance Extracted (AVE) > 0.5.The outcomes from the external model evaluation in Table 2 indicate that all indicators within the PLS model are indeed valid for gauging their designated constructs, as their loading factors > 0.7.Furthermore, the analysis establishes that each construct maintains an AVE > 0.5.

Composite Reliability and Cronbach Alpha
The outcomes of the assessment of construct reliability in Table 2 demonstrate that the Cronbach's alpha values for all constructs > 0.7, and the composite reliability values for each construct also > 0.7.This signifies that all constructs within this Structural Equation Modeling (SEM) Partial Least Squares (PLS) model can be considered dependable.The analysis results in Table 3 indicate that the estimated SEM PLS model fits the analyzed data, as it possesses a moderate model strength (moderate category), good predictive relevance (>0), and the SRMR model value meets the fit criteria.Therefore, this model can be considered suitable for testing the research hypotheses.

Inner Model Testing
Fig 1 .Inner model testing In SEM PLS analysis, the direct influence between variables can be observed from the p-value and T statistics.At a significance level of 5%, exogenous variables are considered to have a significant effect on endogenous variables if the p-value is < 0.05 or the T statistic is > 1.65 (one-tailed) and > 1.96 (two-tailed).The direction of the influence (positive effect/negative effect) is determined by the sign accompanying the path coefficient.The analysis outcomes presented in the aforementioned table reveal the following findings:Customer satisfaction exhibits a favorable and noteworthy influence on attitudinal loyalty, Customer satisfaction displays a positive and substantial impact on behavioral loyalty, Customer satisfaction demonstrates a constructive and significant effect on trust., Service quality showcases a beneficial and significant effect on attitudinal loyalty, Service quality presents a positive and meaningful impact on behavioral loyalty, Service quality illustrates a positive and considerable influence on customer satisfaction, Service quality exerts a positive and noteworthy effect on trust, Trust has a positive and significant impact on attitudinal loyalty, Trust yields a constructive and significant impact on behavioral loyalty.In summation, the analysis outcomes underscore that diverse factors including customer satisfaction, service quality, and trust exert positive and substantial repercussions on various dimensions of loyalty, encompassing both attitudinal and behavioral loyalty., [12]- [18].The analysis results from the table above indicate that: In a specific indirect pathway, the impact of service quality on attitudinal loyalty through customer satisfaction is reflected in a p-value of 0.005, a T statistic of 2.798, and a positive path coefficient of 0.104.The p-value being less than 0.05 and the T statistic exceeding 1.96 lead to the conclusion that service quality indirectly influences attitudinal loyalty when customer satisfaction acts as the intermediary.In this Partial Least Squares (PLS) model, it's established that customer satisfaction serves as a mediator for the indirect effect of service quality on attitudinal loyalty.This indicates that when service quality results in increased customer satisfaction, it consequently contributes to elevated attitudinal loyalty.The statistical findings lend support to the notion that customer satisfaction assumes a mediating role in the dynamic between service quality and attitudinal loyalty.In the context of the indirect pathway involving the impact of customer satisfaction on attitudinal loyalty through trust, the results indicate a p-value of 0.004, a T statistic of 2.905, and a positive path coefficient of 0.092.With the p-value falling below 0.05 and the T statistic surpassing 1.96, it can be deduced that customer satisfaction indeed indirectly influences attitudinal loyalty when trust is a mediating factor.In the PLS model, trust is demonstrated to act as a mediator for the indirect effect of customer satisfaction on attitudinal loyalty.The analysis illustrates that when customer satisfaction indirectly impacts attitudinal loyalty through trust, this impact holds statistical significance.Trust's role as a mediator is validated, indicating its function in establishing a connection between the influence of customer satisfaction and attitudinal loyalty.http://ijstm.inarah.co.id 1386 Within the indirect path involving the impact of service quality on attitudinal loyalty through the combined mediation of customer satisfaction and trust, the analysis yields a p-value of 0.006, a T statistic of 2.735, and a positive path coefficient of 0.053.Given that the obtained p-value is below 0.05 and the T statistic exceeds 1.96, it can be determined that service quality does indeed indirectly influence attitudinal loyalty when both customer satisfaction and trust serve as mediators.In the PLS model, the roles of both customer satisfaction and trust as mediators for the indirect effect of service quality on attitudinal loyalty are confirmed.The analysis provides evidence that when service quality indirectly impacts attitudinal loyalty through the joint mediation of both customer satisfaction and trust, this effect holds statistical significance.Both customer satisfaction and trust are substantiated as mediators, affirming their involvement in establishing the link between service quality and attitudinal loyalty.In the context of the indirect route depicting the influence of service quality on attitudinal loyalty via trust, the analysis yields a p-value of 0.002, a T statistic of 3.050, and a positive path coefficient of 0.100.The fact that the obtained p-value is less than 0.05 and the T statistic exceeds 1.96 leads to the conclusion that service quality indeed indirectly impacts attitudinal loyalty when trust is acting as a mediator.In this specific Partial Least Squares (PLS) model, trust's role as a mediator for the indirect effect of service quality on attitudinal loyalty is firmly established.The analysis highlights that when service quality indirectly influences attitudinal loyalty through the medium of trust, this influence carries statistical significance.
Within the indirect pathway examined, where the impact of service quality on behavioral loyalty is routed through customer satisfaction, the analysis results in a p-value of 0.000, a T statistic of 4.945, and a positive path coefficient of 0.176.The fact that the p-value is below 0.05 and the T statistic surpasses 1.96 leads to the deduction that service quality indeed exercises an indirect influence on behavioral loyalty through the intermediary role of customer satisfaction.In the context of the Partial Least Squares (PLS) model, customer satisfaction's role as a mediator in the indirect effect of service quality on behavioral loyalty is affirmed.This suggests that as service quality elevates, it contributes to heightened customer satisfaction, which subsequently fosters an improvement in behavioral loyalty.The analytical findings substantiate the notion that enhancing service quality can indirectly enhance customer loyalty, facilitated by the mediating impact of customer satisfaction.Within the specified indirect pathway, where the impact of customer satisfaction on behavioral loyalty is mediated by trust, the analysis produces a p-value of 0.003, a T statistic of 3.014, and a positive path coefficient of 0.084.Given that the obtained p-value is less than 0.05 and the T statistic surpasses 1.96, it is established that customer satisfaction does indeed possess an indirect influence on behavioral loyalty through the intermediary role of trust.Within the context of the Partial Least Squares (PLS) model, trust is validated as a mediator for the indirect effect of customer satisfaction on behavioral loyalty.This implies that as customer satisfaction increases, it leads to a heightened level of trust in the organization, and this enhanced trust subsequently contributes to an augmentation of behavioral loyalty.The findings underscore the concept that customer satisfaction indirectly shapes behavioral loyalty by way of trust's mediating role.
Within the described indirect pathway, where the impact of service quality on behavioral loyalty is mediated by both customer satisfaction and trust, the analysis reveals a p-value of 0.005, a T statistic of 2.841, and a positive path coefficient of 0.048.Given the p-value is below 0.05 and the T statistic exceeds 1.96, it can be deduced that service quality indeed exerts an indirect effect on behavioral loyalty through the dual mediation of both customer satisfaction and trust.In the Partial Least Squares (PLS) model, the roles of both customer satisfaction and trust as mediators for the indirect influence of service quality on behavioral loyalty are established.This implies that service quality indirectly shapes behavioral loyalty through the combined influence of heightened customer satisfaction and trust.The enhanced satisfaction and trust levels among customers contribute to an elevation of behavioral loyalty.The findings underscore the notion that the intertwined mediation of customer satisfaction and trust plays a pivotal role in the relationship between service quality and behavioral loyalty.Within the outlined indirect pathway, where the impact of service quality on behavioral loyalty is mediated by trust, the analysis unveils a p-value of 0.001, a T statistic of 3.207, and a positive path coefficient of 0.090.With the p-value being less than 0.05 and the T statistic surpassing 1.96, it can be inferred that service quality indeed exerts an indirect influence on behavioral http://ijstm.inarah.co.id 1387 loyalty through the intermediary role of trust.In the framework of the Partial Least Squares (PLS) model, trust is firmly established as a mediator for the indirect impact of service quality on behavioral loyalty.This underscores that service quality indirectly shapes behavioral loyalty through its impact on trust.When customers perceive an elevated level of service quality, it contributes to a heightened sense of trust in the organization.This enhanced trust subsequently fosters improvements in behavioral loyalty.The findings align with the concept that trust plays a mediating role in the relationship between service quality and behavioral loyalty.In the indirect path considered, where the influence of service quality on trust is channeled through customer satisfaction, the analysis reveals a p-value of 0.000, a T statistic of 7.824, and a positive path coefficient of 0.217.With the obtained p-value being below 0.05 and the T statistic exceeding 1.96, it can be inferred that service quality indeed has an indirect impact on trust through the mediation of customer satisfaction.In the context of the Partial Least Squares (PLS) model, the role of customer satisfaction as a mediator for the indirect effect of service quality on trust is corroborated.This suggests that higher levels of service quality lead to an enhancement in customer satisfaction, which, in turn, contributes to fostering greater trust in the organization.The findings underscore the concept that customer satisfaction serves as a mediating factor in establishing the link between service quality and trust.

IV. CONCLUSION
From the above research results, a solution has been obtained to ensure that the quality of Brilink Agent services aligns with the service standards established by BRI.Several aspects need improvement.While the overall service quality variable has been good, there are certain areas that require enhancement.Specifically, agents need to provide end-to-end service processes, enhance debtor security during transactions, maintain up-to-date service equipment, fulfill customer needs, and provide accurate information.Clear explanations for each desired transaction and an improved impression in debtor service at Brilink Agents are necessary.Regarding the customer satisfaction variable, the overall level has been positive.However, there are areas that still need improvement.Satisfaction with service quality, satisfaction with offered prices, and satisfaction with facilities at Brilink Agents need further enhancement.In terms of the trust variable, the overall level has been satisfactory.Nonetheless, there are specific aspects that need more attention.
Ensuring consistent customer security, acknowledging the presence of Brilink Agents in the community, being reliable for delivering added value, and enhancing high integrity in providing information to Brilink Agent customers are areas that need improvement.In the loyalty variable, the overall level has been good.However, there are aspects that need to be heightened.Improving recall of the "BRI Link" name when conducting transactions, delivering excellent service, fostering trust in Brilink Agents, enjoying transactions at Brilink Agents, and recommending Brilink Agents to others are areas that still require improvement.This research has identified certain limitations and avenues for future investigation.Firstly, the results are specific to Brilink Agent customers, and the sample was only taken from North Sumatra.It would be more intriguing to broaden future studies by including samples not only from North Sumatra but from across Indonesia, aiming to gain a broader understanding of Brilink Agents by increasing the sample size.Secondly, this study is based solely on Brilink Agents in the North Sumatra region.Therefore, consumer behavior could be influenced by cultural variables, and satisfaction as well as loyalty might vary between regions.[19] It would also be interesting to determine whether the findings of this research hold strong in other regions. V.

Table 1 .
Respondent Characteristics Description

Table 2 .
Convergent Validity & Composite Reability The assessment of the SEM PLS model's goodness of fit is based on the values of R Square, Q Square, and SRMR.R Square evaluates the model's capability to predict endogenous variables.R Square values range from 0 to 1 and are classified into three categories: strong, moderate, and weak, R Square value > 0.67 suggests a strong PLS model, values between 0.33 -0.67 indicate a moderate PLS model, while values between 0.19 -0.33 represent a weak model.The Q Square value signifies the model's predictive relevance, where a value > 0 indicates predictive relevance.Conversely, SRMR assesses how well the sample explains the population.An SRMR value is considered favorable if it approaches 0.

Table 3 .
Goodness Of Fit Model