Customer satisfaction and trust interaction model

Article history: Received: August 14, 2020 Received in revised format: November 14 2020 Accepted: November 14, 2020 Available online: November 23, 2020 The purpose of this study is to identify and analyze the interaction between customer satisfaction and trust in a structural model. Sample consists of 210 Suzuki motor bikes’ users, aged 17-64, domiciled in Medan. Structural equation model approach is used to process raw data. Respondents' opinions were taken proportionally throughout the Medan city sub-districts. Questionnaires were submitted accidentally. The results show that customer satisfaction and trust partially had positive and significant effect on brand equity and repurchase intention. Brand equity did not show any indirect causal relationship mediator of the three exogenous variables to repurchase intention. Programs related to customer satisfaction and trust better increase brand equity and repurchase intention directly. Interaction variable existence in research model increases Chi-Square probability, decreases standard deviation, and increases t-value. Interaction variable creates synergy on the influence of customer satisfaction and trust to increase brand trust. Any programs to increase brand equity through customer satisfaction and trust should be implemented, simultaneously. The two exogenous variables will bring out the synergy that comes from this interaction. © 2021 by the authors; licensee Growing Science, Canada

The purpose of this study is to identify and analyze the interaction between customer satisfaction and trust in a structural model. Sample consists of 210 Suzuki motor bikes ' users, aged 17-64, domiciled in Medan. Structural equation model approach is used to process raw data. Respondents' opinions were taken proportionally throughout the Medan city sub-districts. Questionnaires were submitted accidentally. The results show that customer satisfaction and trust partially had positive and significant effect on brand equity and repurchase intention. Brand equity did not show any indirect causal relationship mediator of the three exogenous variables to repurchase intention. Programs related to customer satisfaction and trust better increase brand equity and repurchase intention directly. Interaction variable existence in research model increases Chi-Square probability, decreases standard deviation, and increases t-value. Interaction variable creates synergy on the influence of customer satisfaction and trust to increase brand trust. Any programs to increase brand equity through customer satisfaction and trust should be implemented, simultaneously. The two exogenous variables will bring out the synergy that comes from this interaction.

Introduction
The motorcycle industry is making innovations aggressively in an effort to meet market demands. Competition encourages them to keep developing their product attributes or features. The 10 best motorcycle brands in the world in 2017 are Honda, Yamaha, Ducati, Harley Davidson, Kawasaki, Suzuki, BMW, Aprilia, Triumph, and KTM (4muda, 2017). The motorcycle market share in Indonesia in February 2019 was controlled by Honda with a market share of 77.1%, second behind Yamaha (19.4%), third place was Kawasaki (2%), then Suzuki (1.6%), and finally TVS (Nugroho, 2019). Based on researchers' observations during May 2020 at 10 largest universities in Medan, there were approximately 60% Honda motorbikes' users, 25% Yamaha's users, 10% Suzuki, and 5% Kawasaki. Many marketing researchers discuss consumer satisfaction as an antecedent of certain endogenous latent variables (Sari & Giantari, 2020;Waari, 2018;Ibzan et al., 2016;Chang et al., 2014;He et al., 2008;Anić & Radas, 2006;Hellier et al., 2003). Likewise, customer trust is an antecedent (Widodo & Murwatiningsih, 2019;Bricci et al., 2016;Upamannyu et al., 2015;Utami, 2015;Nguyen et al., 2013;Hayati, 2011;Hong & Cho, 2011). The authors also discuss the one-way relationship between satisfaction and customer trust (Setiawan et al., 2020;Rahmatiyah et al., 2017;Leninkumar, 2017;Prameka et al., 2016;Suki, 2011;Franco, 2009). or vice versa, a one-way causal relationship between customer trust and consumer satisfaction (Ruswanti et al., 2020;Dung, 2019;Leonnard, 2019;Soo Ho, 2018;Mahaputra, 2017;Kim, 2017;Hidayat et al., 2015;Ardhiansyah et al., 2014). Authors rarely discuss the interaction between satisfaction and customer trust in a research model. The main substances that will be discussed in this research is the interaction between satisfaction and customer trust and its effect on the brand equity of Suzuki motorbikes in Medan. Besides, it will also discuss the causal relationship between the exogenous latent variables of customer satisfaction and customer trust toward endogenous latent variables of brand equity and repurchase intention. The results are useful for marketing managers in establishing programs to increase customer satisfaction and trust. The interaction between the two exogenous variables has been shown in several studies. There is an interaction between customer satisfaction and customer trust that significantly influences other latent variables in a model (Bilson, n.d.). The interaction between customer satisfaction and antecedent e-loyalty trust in research (Lu & Hakim, 2017). The study by Kim (2012) provides a framework explaining the subsequent relationships of trust, expectation, confirmation, satisfaction, post-expectation, and repurchase intention. Consumer trust has an effect on satisfaction, and satisfaction has an effect on repurchase intention. However, this model does not influence each other (interaction) between trust and satisfaction. Study results by Tirtayani and Sukaatmadja (2018) show that e-satisfaction has no effect on online repurchase intention. In his research Dehghan et al. (2015) found Customer satisfaction has no significant effect on repurchase intention. However, Customer satisfaction may have some effects on repurchase intention (Sari & Giantari, 2020;Ruswanti et al., 2020;Ashfaq et al., 2019). The results of hypothesis testing concluded that not all dimensions of consumerbased brand equity have a positive and significant effect on consumers satisfaction (Susanty & Kenny, 2015). Another author found all the dimensions of brand equity (brand awareness, brand loyalty, brand association and perceived quality) have a significant association with customer satisfaction (Kumar, 2014). Brand equity has no effect on brand satisfaction (Erciş et al., 2012). Other researchers (Basheer et al., 2017;Bilal & Malik, 2014;Kim et al., 2008) say the consumer satisfaction effect on brand equity.

Repurchase Intention
Buyback intention refers to the possibility of using the brand again in the future (Fornell, 1992). Another author says intention to rebuy is a customer self-reported possibility of being involved in a future repurchase behavior (Seiders et al., 2005). So, the intention to repurchase is the consumers' will referring to his/her experience and information received to buy a product that has been purchased on the future. Repurchase intention indicators in its study (Wijaya et al., 2018) are: wants to repurchase, wants to recommend to friends, wants to continue buying, and wants to shop for other goods from the same supplier. The indicators used by Hellier et al. (2003) to measure repurchase intention are: Intend to continue to purchase and intend to continue to contribute. The repurchase intention indicators used by Bulut (2015) are: will make purchase again, will visit again in the future, and intend to recommend. Repurchase intention indicators in this study are: Want to buy another Suzuki motorcycle; Intend to recommend to a friend; Want to buy another Suzuki brand; Intend to ignore competitors' promotions.

Brand Equity
One of the company's equity is brand, name, sign, symbol, term, design or combination of signs to identify goods or services from one seller or a group of sellers is called a brand (Kotler & Amstrong, 2012). Another author said, The notion that brands add value to products has been called brand equity (Pappu & Quester, 2006). Brand equity is a collection of brand assets and liabilities associated with a brand, its name and symbol that increase or decrease the value provided by a product or service (Aaker, 2009). Brand equity is an intangible asset derived from the brand owned by a company, differentiating products from competitors. Another authors established five dimensions of brand equity -consist of brand awareness, brand association, perceived quality, brand loyalty and other proprietary brand assets (Aaker, 2009;Kotler, 2008;Washburn & Plank, 2002;Yoo & Donthu, 2001). The results of a study by Khan (2012) state that brand equity includes: Brand awareness, brand familiarity, brand image, brand association, brand loyalty, brand preference, and brand availability. Formative indicator of brand equity (Gómez et al., 2019) are Brand Associations, Perceived Quality, Loyalty, and Awareness. The indicators of brand equity according to the author are: Brand Associations, brand image. brand availability, and brand preference.

Customer Trust
Transaction occurs when the buyer trusts the seller. Trust is the willingness of consumers to rely on companies based on their confidence (Moorman et al., 1993). Trust is an assessment of one's relationship with others who will conduct certain transactions in line with expectations in an environment full of uncertainty (Ba & Pavlou, 2002). Customer trust is a perception of customer trust reliability is based on experience or more in a sequence of transactions or interactions that are indicated by the fulfillment of product performance and satisfaction expectations (Costabile, 2000). Customer trust is the result of customer evaluation that a certain product is superior after using, consuming, or receiving information from others regarding the product. In their research (Gefen & Straub, 2004) trust includes integrity, predictability, ability, and benevolence. Trust can be measured by Availability, competence (ability), discreteness, fairness, integrity, loyalty, openness, promise fulfillment, receptivity (Butler, 1991). In this study, customer trust is reflected in ability, credibility, benevolence, and openness.

Consumer Satisfaction
Consumer satisfaction plays a very important role in the company sustainability. Satisfaction is a person's feeling of pleasure or disappointment after comparing the perceived performance of a product in relation to expectations ( Kotler & Keller, 2016). Customer satisfaction is defined as customers' feelings of pleasure, fulfillment and desire towards a service rendered (Thaichon & Quach, 2015). The author's definition is related to consumer satisfaction, namely the user or consumer's assessment of the use or after consuming a product compared to his/her standards. If the ratings are the same or above the standard, then they are satisfied. Conversely, if it is below the standard they are not satisfied or even complain. The measurement of consumer satisfaction is of utmost importance to the organization. (Roche & Allen, 2007). Possible dimensions in measuring customer satisfaction can be quality, price, complaints, trust relationships, problems and others (Khadka & Maharjan, 2017). Factors that are often used to evaluate customer satisfaction (Tjiptono et al., 2008) are satisfaction with: Performance, features, reliability, conformance to specifications, durability, serviceability, aesthetics, and perceived quality. The author uses satisfaction on these eight indicators to evaluate the satisfaction of Suzuki motorcycle users in Medan.

Research Framework
Customer satisfaction and brand trust were positively related to each other, besides that the two variables are correlated with brand equity (Basheer et al., 2017). There is a positive relationship between satisfaction and trust (Erciş et al., 2012). This study results indicate that there is an interaction between customer satisfaction and customer trust. Increased customer satisfaction carried out by the company will have an impact on the stronger brand equity of the product concerned. The product impression will stick in their minds. Many previous studies have concluded that the consumer satisfaction effect on brand equity such as (Basheer et al., 2017;Torres & Tribó, 2011;Ossipenko et al., 2009;K. H. Kim et al., 2008). Increasingly satisfied consumers not only increase the brand equity of a product or company, but can increase repurchase intention. Conversely, if not satisfied, it will certainly detrimental to the company. Customer satisfaction effect on repurchase intention (Sari & Giantari, 2020;Ruswanti et al., 2020;Ashfaq et al., 2019;Ayo & Jones, 2018;Li, 2016;Ibzan et al., 2016;Chang et al., 2014;Wen et al., 2011;Santos & Fernandes, 2008). Transactions that occur between sellers and buyers are based on trust. Consumers will not buy a product or service if they do not believe that producers can fulfill their wants. The increase in brand equity is driven by increasing customer trust. Customer trust effect on brand equity (Poerwadi et al., 2019;Basheer et al., 2017;Kim et al., 2008). Besides that, the trust that is maintained or increased encourages a stronger repurchase intention. Customer trust effect repurchase intention (Kim, 2017;Santos & Fernandes, 2008;Wijayajaya & Astuti, 2018). Every company wants to grow or operate continuously. Therefore, their repurchase intention must be strengthened. One strategy is to increase brand equity.

Hypothesis
The problems described above, the theory, and the conceptual framework proposed by the researcher are the basis for formulating the following temporary answers: H1: There is an effect of consumer satisfaction on brand equity; H2: There is an effect of consumer satisfaction on repurchase intention; H3: There is an effect of customer trust on brand equity; H4: There is an effect of customer trust on repurchase intention; H5: There is an effect of brand equity on repurchase intention; H6: There is an effect of the interaction variable between consumer satisfaction and customer trust on brand equity.

Research Method
This research type is quantitative with a survey, located in Medan, which is the third largest city in Indonesia. The main data source was from primary sources. Initially, the research model (main model) did not contain an interaction between the customer satisfaction latent variables and customer trust. Therefore, it consists of 4 (four) latent variables, namely customer satisfaction (COSA), customer trust (CUST), brand equity (BRE), and repurchase intention (REIN). Operationalization of the four latent variables resulted 20 observed variables. The measurement uses ordinal data with a Likert scale of 1-7. The number of respondents was determined based on structural equation model with robust maximum likelihood method. Input data was in asymptotic covariance matrix. The number of samples is based on n ≥ 1 / 2K (K + 1), where K is the number of manifest variables (Jöreskog, 2016). The sample in this study included 210 people. They are currently using a Suzuki brand motorbike age 17 -64, domiciled in Medan. The origin of respondents is determined to be proportional to the population in 21 sub-districts in Medan city (Nasution, 2020). This technique is intended to obtain more representative information from this area. The questionnaire was submitted accidentally. The first step in the Structural Equation Model (SEM) analysis technique is the development of the theory that underlies the model; Model specification development; Examination of measurement models; Structural model development; Model inspection; Model modification. Structural models with interaction are then developed by adding an interaction variable (Ping,1996). The interaction model has 5 (five) latent variables. Modeling the interaction variable with a single indicator using a latent variable score (LVS) (Jöreskog et al., 2006). The interaction variable (SOCU) is made by multiplying consumer satisfaction's LVS to customer trust.

Measurement Model Fit
The measurement model for customer satisfaction is reflected by 8 indicators, but 4 of them have a standardized loading factor (SLF) not ideal, then eliminated from the model.

Structural Model Fit
Structural model examination results include initial model feasibility without interaction, followed by an interaction model. The model without interaction must be fit first (Ping, 1996), then the interaction between latent variables of customer satisfaction and customer trust is developed. The structural model path diagram without interaction is presented in Fig. 2. The path diagram of the structural model below in structural equations form is presented as follows: BRE = 0.29 × COSA + 0.49 × CUST (1) REIN = 0.043 × BRE + 0.47 × COSA + 0.35 × CUST (2) Rsquare1 = 0.53; Rsquare2 = 0.62. The two determinant coefficients (R-Square) are moderate (Chin, 1998).

Fig. 3. Path diagram of structural model with interaction (standardized solution)
Correlation between variables ˂ 0.90. Based on the path diagram, it can be seen that there is no cross-loading indicator. These parameters indicate structural model with interactions is fit (Waluyo, 2016). The direct effect coefficient, standard deviation, and significance of the exogenous latent variable on the endogenous latent variable of the two models above related to the hypothesis can be seen in Table 1. Direct effect Regression weight of COSA → BRE = 0.29, standard deviation = 0.13, t-value = 2.26 with 95% degree of confidence means significant, H1 is accepted. Direct effect regression weight of CUST → BRE = 0.49, standard deviation = 0.13, t-value = 3.82 with 95% degree of confidence means significant, H2 is accepted. Direct effect regression weight COSA → REIN = 0.47, standard deviation = 0.12, t-value = 3.85 with 95% degree of confidence means significant, H3 is accepted. Direct effect regression weight of CUST → REIN = 0.35, standard deviation = 0.11, t-value = 3.13 with 95% degree of confidence means significant, H4 is accepted. direct effect regression weight of BRE → REIN = 0.04, standard deviation = 0.10, t-value = 0.45 with 95% degree of confidence meaning not significant, H5 is rejected. Direct effect regression weight of SOCU → BRE = 0.17, standard deviation = 0.06, t-value = 3.00 with 95% degree of confidence means significant, H6 is accepted.

Direct effect in the no-interaction model
The direct effect of customer satisfaction on brand equity is positive and significant. Increased customer satisfaction will increase brand equity. Conversely, a decrease in customer satisfaction will result in a decrease in brand equity. This fact is in line with the research results (Basheer et al., 2017;Torres & Tribó, 2011;Ossipenko et al., 2009;Kim et al., 2008). The differences in the authors' studies include analysis techniques, between SEM and multiple regression analysis techniques. Differences in the size of the study sites covering respondents from 11 countries (Torres & Tribó, 2011). The location of this research is a major city in Indonesia, differences in research object. The research object (Kim et al., 2008) is the Korean hospital, while (Ossipenko et al., 2009) has a business-to-business setting research object. The object of this research is the Suzuki motorcycle brand. Differences also occur in the customer satisfaction measurement. This study uses consumer opinion about satisfaction with the quality dimensions of Suzuki motorbikes to measure the customer satisfaction latent variables. Other researchers use different measurement techniques. Customer satisfaction has a direct, positive, and significant effect on repurchase intention. Customers' intention to repurchase the same brand will increase, if they are satisfied with the product being used or added to the information they receive and finally it is hoped that a repeat purchase will occur. If customers are not satisfied, their intention to repurchase will ebb. So, marketers cannot expect repeat purchases. This study results are supported by many studies (Sari & Giantari, 2020;Ruswanti et al., 2020;Ashfaq et al., 2019;Ayo & Jones, 2018;Li, 2016;Ibzan et al., 2016;Chang et al., 2014;Wen et al., 2011;Santos & Fernandes, 2008). Consumer satisfaction is the strongest latent variable that directly affects repurchase intention. Customers who feel satisfied with a brand, in addition to increasing brand equity itself, will also directly increase repurchase intention. Customer satisfaction is a very important component in business.
Customer trust affects brand equity directly, positively and significantly. Increased customer trust is an added value to brand equity. Conversely, a decrease in customer confidence will reduce brand equity. In other words, the company's assets are decreasing. This study results are in line with (Poerwadi et al., 2019;Basheer et al., 2017;Kim et al., 2008). In this study, the impact of customer trust on brand equity is stronger than customer satisfaction. Kopi Kapal Api brand trust is measured with quality indicators that match expectations, product reliability, confidence, and recommendations (Poerwadi et al., 2019). In this study, product quality expectations are used to measure customer satisfaction. Recommendations are not used in measuring Suzuki motorcycle brand trust. The analysis technique used (Basheer et al., 2017) is multiple linear regression so that it does not further describe the customer trust latent variables. The research object (Kim et al., 2008) is hospitals in Korea. Customer trust has a direct, positive and significant effect on repurchase intention. Customers who have increased trust in the brand will encourage stronger repurchase intentions. This situation at some point will result in a buyback. Repurchasing will increase efficiency. Recruiting new customers requires a relatively higher cost than retaining old customers. This efficiency will increase the company's profit. However, if customer trust decreases in a product it will result in a decrease in the intention. Furthermore, there will be no repeat purchases. This fact is supported by research results (Kim, 2017;Santos & Fernandes, 2008;Wijayajaya & Astuti, 2018). The research objects of each author are traditional markets, aviation services, and online shopping (Case at Berrybenka online customer in Indonesia). In this study, the three exogenous variables have a direct, positive and significant effect on brand equity. The interaction variable of customer satisfaction with customer trust has a direct, positive, and unidirectional effect on brand equity. Simultaneous application of customer satisfaction and customer trust in increasing equity will create a synergy. New strength arises from the interaction of the two exogenous variables to increase brand equity.

Impact of interaction variables.
Interaction variable presence (SOCU) in this research model directly affects brand equity. Therefore, it causes changes in the parameters of structural model. Chi-Square probability (χ2) in the model without interaction is 0.26. Main criterion for this GOF is 1.00 in the structural model with interaction.

Table 2
Comparison of regression weight, standard deviation, and t-value between models without interaction and models with interactions in direct effects There was an increase of 0.74. Compatibility of the sample covariance matrix with the estimator model covariance matrix in the interaction model is better. Comparison of regression weight, standard deviation, and t-value on model without interaction and model with interaction can be seen in Table 2. Direct effect coefficient of customer satisfaction on brand equity increases in the interaction model from 0.29 to 0.36. This influence is getting stronger, more positive and significant. The standard deviation decreased from 0.13 to 0.09. Variance in the measurement model is decreasing. Model in this aspect is getting better. T-value increased from 2.26 to 4.21. The higher significance of the causal relationship between the two variables COSA → BRE in the interaction model. Direct effect regression weight of customer trust on brand equity changes in the interaction model from 0.49 to 0.47. The influence is decreasing, positive, and significant. Standard deviation decreased from 0.13 to 0.09. Variance in the measurement model is decreasing. The model in this aspect is getting better. T-value increased from 3.82 to 5.49. Significance of the causal relationship between the two variables CUST → BRE is higher in the interaction model. Direct effect regression weight of consumer satisfaction on repurchase intention changes in the interaction model from 0.47 to 0.46. The influence is decreasing, positive, and significant. The standard deviation decreased from 0.12 to 0.08. The variance in the measurement model is decreasing. Model in this aspect is getting better. T-value increased from 3.85 to 5.47.
Significance of the causal relationship between the two variables COSA → REIN is getting better in the interaction model. Direct effect coefficient of customer trust on repurchase intention did not change in the interaction model, namely 0.35. This influence is positive and significant. Standard deviation decreased from 0.11 to 0.09. Diversity in the measurement model is decreasing. Model in this aspect is getting better. T-value increased from 3.13 to 3.86. Significance of the causal relationship between the two variables CUST → REIN is getting better in the interaction model. Direct effect coefficient of brand equity on repurchase intention increases in the interaction model from 0.04 to 0.06. The influence is positive, increasing, but not significant. Standard deviation decreased from 0.10 to 0.08. Variance in the measurement model is decreasing. Model in this aspect is getting better. T-value increased from 0.45 to 0.69.

Table 3
Comparison of regression weight, standard deviation, and t-value between models without interaction and models with interactions in the indirect effect. Indirect effect of consumer satisfaction on repurchase intention through brand equity is not significant (See Table 3). Brand equity is not a mediator of consumer satisfaction on repurchase intention. The indirect effect of customer trust on repurchase intention through brand equity indicates that Brand equity is not a mediator of customer confidence toward repurchase intentions. The indirect effect of the interaction variable on repurchase intention through brand equity is not significant. Brand equity is not a mediator of the interaction variable on repurchase intention. Even though in the three relationships, brand equity cannot mediate the causal relationship between variables in the model. Interaction variable existence has an impact on parameters such as regression weight, standard deviation, and t-value. Indirect effect regression weight of COSA → BRE → REIN increased from 0.01 to 0.02; Unchanged standard deviation of 0.03; T-value increased from 0.44 to 0.68. Indirect effect coefficient CUST → BRE → REIN increased from 0.02 to 0.03; Standard deviation decreased from 0.05 to 0.04. T-value increased from 0.45 to 0.69.

Conclusion and Recommendation
The effect of customer satisfaction and customer trust partially on brand equity and repurchase intention was positive and significant. The increase in the two exogenous variables will have a direct impact on brand equity and repurchase intention. Brand equity is not a second mediator of exogenous variables on repurchase intention. Therefore, the program to increase customer satisfaction and trust is better directed at increasing brand equity and repurchase intention. The emphasis of the program is on increasing customer satisfaction in satisfaction at conformance. Emphasis is also on the program to increase customer trust in abilities. The interaction variables existence increases the goodness of models such as the probability of chi square, standard deviation, and t-value in both direct and indirect relationships. This interaction variable creates synergy encouraging Suzuki motorcycle brand equity in Medan. Programs related to customer satisfaction and customer trust should be implemented, simultaneously. Two variables will interact with each other to create synergy in strengthening brand equity. The research location is still limited of Medan city people. The highest determinant coefficient is 0.62. The change in variance of endogenous variables can be explained by predictors of 62%. The next researcher can expand the research location to provincial or national levels, or add exogenous variables to the research model.