“Evolution of customer satisfaction in the e-banking service industry”

Since several commercial activities such as banking, shopping, transfers, and payments had been conducted online, many banks in Cambodia provided e-banking services to their customers to support these activities. Meanwhile, if the banks could provide such e-banking services to satisfy their customers’ needs, they could maintain their customers and profits. Thus, finding the main factors influencing customer satisfaction in the e-banking service industry is significant. Therefore, the objective of this paper is to investigate how customer satisfaction develops through examining the impacts of perceived risk (financial risk and performance risk) and perceived value on customer satisfaction in the e-banking. This study applied convenience sampling to get data from the respondents at convenient locations (near banks, markets, supermarkets, universities, and workplaces). 700 respondents who were currently using mobile bank- ing or internet banking services at either commercial or retail banks in Cambodia were invited to fill in the questionnaires. In addition, the results of this study were gener- ated through structural equation model (SEM) analysis based on 546 valid responses. The results revealed that perceived value was mainly influenced by performance risk, whereas financial risk did not significantly affect perceived value. Finally, perceived value and performance risk significantly influenced customer satisfaction, except fi- nancial risk. In addition, despite both perceived value and performance risk significantly influencing customer satisfaction, promoting customer satisfaction through increasing perceived value was far more effective than minimizing performance risk.


INTRODUCTION
Since technological innovation is advancing in electronic channels, many businesses no longer face industrial, geographical, and regulatory barriers. The internet has become the new tool for interactions between customers and businesses (Firdous & Farooqi, 2017). Based on the current changes in market trends, e-banking services have been developed to become the new means for market interaction (Asad et al., 2016).
Meanwhile, Wong (2020) has reported that the percentage of Cambodians actively using the internet is around 90%. While the major commercial activities, such as shopping, banking, and payments, can be conducted online (Firdous & Farooqi, 2017), many e-banking services have been created to serve current market transactions in Cambodia (Wong, 2020;Yang et al., 2021). However, the banks have to maintain high customer satisfaction if they want those customers to continue using their e-banking services because customer satisfaction can positively influence an individual's purchase decisions (Abdul-Muhmin, 2010). Therefore, investigating the factors that can influence satisfaction is very important for the e-banking service industry.
Due to the significant impact of satisfaction on many businesses, previous investigations conducted by Tran (2020) in the online shopping service industry and by Jin et al. (2016) in the restaurant industry have suggested minimizing perceived risk for customers. High perceived risk can cause a high threat to customers and lower customer's confidence and joy in using the service. Therefore, minimizing customer perceived risk can ensure high customer satisfaction. In contrast, the investigations conducted by Tukiran et al. (2021) in the education industry and by Samudro et al. (2020) in the chemical industry have suggested maximizing perceived value. Enhancing service value can allow customers to see the advantages of investing in the firms. Thus, increased perceived value can promote high customer satisfaction.
Even though the current perceived risk and perceived value have been individually tested with satisfaction in different industries, integrating these variables to explain how customer satisfaction develops has been studied only slightly in the e-banking service industry. Moreover, developing a complex satisfaction model, which includes the perceived risk dimensions (performance risk and financial risk) and perceived value as the systematic impacts on customer satisfaction, has also remained narrow in the current marketing literature. Oliver (1980) defines customer satisfaction as an individual's evaluation that results from a comparison between service performance and his or her expectation. Namkung and Jang (2007) explain that satisfaction occurs when service performance meets customer expectations; otherwise, customers may not be happy with the firm. Konuk (2019) has added that customer satisfaction can be evaluated based on two main stages: emotional and cognitive. Omar et al. (2011) reveal that the emotional stage shows a person's sense or feeling to the service performance with their expectations. In contrast, the cognitive stage reveals the rational evaluation of the service utility while using it. Many industries have considered customer satisfaction a significant factor that can increase customer repurchase behavior (Abdul-Muhmin, 2010). Therefore, many studies have been conducted to investigate the evolution of customer satisfaction in different industries.

Customer satisfaction
For example, Tien et al. (2021) used multiple regressions to study personal factors, price, product quality, and service quality with satisfaction in the banking service industry. Geebren et al. (2021) used the structural equation model (SEM) to study the effect of trust, service quality, information, and system quality on satisfaction.
Zaid and Patwayati (2021) used partial least square analysis in the online shopping industry to study the influence of experience and customer engagement on satisfaction. To et al. (2020) used the SEM to study perceived quality, brand awareness, brand image, and customer satisfaction.
In tourist industries, Cho (2020) used ANOVA analysis to study the effect of sustainability, culture, trust, service quality, and price on satisfaction. Tran et al. (2021) used SEM to study the influence of perceived quality, brand awareness, and brand image on customer satisfaction.
In other industries, Samudro et al. (2020) used SEM to study the effect of perceived value and quality on satisfaction in the B2B industry. For example, Kumar et al. (2021) used SEM to study service quality, competency, product price, and efficiency in telecommunications and their relationship with satisfaction. In the e-commerce industry, Nguyen (2020) used SEM to study security, external incentives, customer service, shopping experience, and customer satisfaction.
Based on the current gap in the literature, although many factors have been studied with customer satisfaction in different industries, the impacts of perceived risk and perceived value on customer satisfaction have not been studied extensively, especially in the e-banking service industry. In addition, a combined model structure that consists of the perceived risk dimensions (fi-nancial risk and performance risk), and can create systematic relationships with perceived value and customer satisfaction, has remained studied only slightly. Therefore, the purpose of this study is to examine the impacts of perceived risk (financial risk and performance risk) and perceived value on customer satisfaction in the e-banking service industry of Cambodia.

Perceived risk and perceived value
Piri and Lotfizadeh (2016) have conceptualized perceived risk as an individual's perceived uncertainty about the service or product evaluated before or after he or she buys it. From the psychology of the subject to consumer behavioral science, perceived risk underscores a customer's negative evaluation after having an unpleasant shopping experience with the firms ( Based on a conceptual comparison, risk indicates customer skepticism of service performance, causing unexpected losses to customers (Piri & Lotfizadeh, 2016), while perceived value indicates the benefits obtained from service performance (Tuncer et al., 2021). The current perspectives between the two variables are quite the opposite. From the mobile consumer behavioral perspective, fewer benefits from the product or service utility occur when a poor product or service performance is perceived by customers (Piri & Lotfizadeh, 2016). In online and offline shopping value perception, risk can create discomfort to customers who may expect a low return from their investment in the firms (Broekhuizen & Jager, 2003

Perceived value and customer satisfaction
Yusiana and Widodo (2020) define perceived value as the overall benefit assessment resulting from the trade-off between cost and financial value. According to mean-end theory, the concept of the value indicates the comparison between price and quality (Zeithaml, 1988 (2020) reveal that when products are expected to provide more advantages, customers definitely enjoy using them more.

AIM AND HYPOTHESES
Based on the contemporary literature, this study has attempted to fill the research gap by developing a new research model to investigate the factors affecting customer satisfaction in the e-banking service industry.
As a result, the research objective is to investigate the systematic impacts of the perceived risk dimensions (performance risk and financial risk) and perceived value on customer satisfaction in the e-banking service industry. Based on the above theoretical discussions, the study has created systematic impacts among variables (performance risk, financial risk, perceived value, and customer satisfaction). Thus, the study has generated the following hypotheses ( Figure 1): H1: Financial risk negatively influences perceived value. H3: Financial risk negatively influences customer satisfaction.
H4: Performance risk negatively influences customer satisfaction.
H5: Perceived value positively influences customer satisfaction.

Sample and data collection
700 bank customers were asked to participate in this study. However, bank customers in this study could be either mobile banking or internet banking users at either commercial or retail banks in Cambodia. To obtain the data from these respondents, convenience sampling technique was used to interview the respondents at convenient places such as banks, stores, supermarkets, and other workplaces . Thus, all of the bank customers located at the above-mentioned places were invited to fill in the questionnaires.
For the survey processes, the respondents, firstly, were approached and asked a screening question "Are you using the e-banking service at any bank in Cambodia?" Next, if they said "yes," they were asked for their agreement to fill in the questionnaires voluntarily. After that, those who volunteered to participate in the survey were strictly guided on how to fill in the questionnaires correctly, within approximately 10-15 minutes. Finally, all questionnaires were collected at the end of December, after three months of data collection in Cambodia (October-December, 2021). However, the data of only 546 were good enough for processing in data analysis.

Measurement constructs
The measurement constructs in this study were adopted from previous studies. For example, the items of perceived risk (financial risk = 2 items and performance risk = 3 items) were adopted from Yang et al. (2016). Next, the items of perceived value were adopted from Kim and Jindabot To obtain the answers from the bank customers, the items in each variable were rated using a 5-point Likert scale (1 = strongly disagree to 5 = strongly agree). Babakus and Mangold (1992) supported that using the current scale technique could provide adequate time to respondents in order to complete the questionnaires. In addition, Garland (1991) highly recommended using the current scale because it has a mid-point (3 = neutral), dividing the boundaries between positive and negative scales for the respondents to answer. Thus, it could reduce time as well as frustration to customers.

Structural equation model (SEM) measurement
This study used the structural equation model (SEM) to analyze 546 responses. However, the model built in this study was checked to ensure good model measurement and model fitness before performing SEM analysis.
First, the SEM model measurement was examined to ensure a reliable construct for each variable in this study. First, the loading factors with scores above 0.6 were kept for analysis. Second, each Cronbach's alpha score above 0.7 indicated the content reliability (Mosavi et al., 2018). Then, all composite reliability (CR) scores above 0.7 were also acceptable . In Table 1, all of the average variance extracted (AVE) scores above 0.5 clearly proved the acceptable reliable construct of the model (Kim et al., 2021).
Second, the fitness degree of the model was examined to ensure the acceptable regression results of the SEM analysis. Therefore, confirmatory factor analysis (CFA) was conducted to adjust the leading indicators' index scores (CMIN 2 /df, GFI, CFI, NFI, AGFI, RMSEA, and PCLOSE) in order to pass the minimum requirement of model fitness recommended by . After using CFA to modify the indicators' index, all indicators passed the thresholds in Table 2. Therefore, SEM analysis was conducted in Figure 2.

SEM results
The results obtained from the SEM analysis are summarized in both Figure 2 and Table 3. Based on the current results of this study, financial risk showed an insignificant impact on perceived value (β = -0.09, p < 0.05), whereas performance risk showed a significant impact on perceived value (β = -0.91, p < 0.001). Finally, perceived value and performance risk showed significant impacts to customer satisfaction (β = 0.54, p < 0.001 and β = -0.36, p < 0.001, respectively). In contrast, financial risk showed an insignificant impact to customer satisfaction (β = -0.07, p < 0.05). Furthermore, the explanations of the impacts to the main dependent variables in this study were provided in the discussion section.

Effects on perceived value
Regardless of the effects on perceived value, performance risk negatively affected perceived value, which accepts hypothesis 2. Likewise, Yang et al. (2016) also found that the value of the service could reduce over time if the customers highly suspected performance risk. In fact, poor service performance could result in high frustration and stress to customers. In addition, the risk of service performance could also waste customers' time and effort. Likewise, having many errors and inaccurate financial information probably made customers have deep concerns with the current e-banking services. Therefore, these possible losses created a negative perspective for the service providers. Based on this circumstance, performance risk was taken seriously by many customers because it showed how vulnerable the e-banking system was.
In contrast, financial risk negatively affected perceived value; however, it showed an insignificant effect on perceived value, rejecting hypothesis 1.
Obviously, the current result contrasted with previous studies such Agarwal and Teas (2015) in the drug production industry and Yang et al. (2015) in the mobile devices industry. They claimed that financial risk was the main predictor of perceived value. However, the insignificant current effect of financial risk on perceived value could be ex-plained based on the current empirical results of this study. First, the insignificant effect of financial risk on perceived value outlined no serious customer concerns on future financial losses when customers decided to use the e-banking service with the banks. Second, the e-banking service was important to customers only if the service demonstrated high service performance. High service performance indicated that high quality could possibly gain high system protection, especially on credit card information and personal information. Therefore, financial risk may not have altered a customer's positive outlook on the e-banking service.

Effects on customer satisfaction
Regardless of the effects on customer satisfaction, perceived value positively affected customer satisfaction, which accepts hypothesis 5. The current result was similar to Tukiran et al. (2021), who claimed that increasing high-perceived value could enhance high satisfaction with the product or service of the firms. At the same time, perceived value also acted as the partial mediator between performance risk and customer satisfaction. The current results underscored the significant effect of perceived value that could obtain more customer satisfaction by lowering the performance risk. The value of the e-banking service increased depending on the degree of performance risk. Lower performance risk gained high positive opinions, which resulted in more favorable positions for customers. Thus, it probably created more customer enjoyment of the service utility.
Second, performance risk negatively affected customer satisfaction, which accepted hypothesis 4. Similarly, the current result was similar to Tran (2020), who found that high performance risk led to low customer satisfaction. In fact, performance risk created performance uncertainty which could cause an unfavorable desire to continue using the e-banking service. Based on the current result, its direct impact on customer satisfaction revealed how customers may have been dissatisfied immediately with the e-banking service. Performance risk reflected poor service offerings (e.g., errors and an unsecured e-banking service system); thus, if it remained very high, customers were not happy with the banks, which could result in an unfavora-ble decision to leave at any time. However, customers were happy to continue using it with the banks if it remained very low.
Unlike perceived value and performance risk, financial risk negatively affected customer satisfaction, but it showed an insignificant effect on customer satisfaction, rejecting hypothesis 3. In contrast, this result was the opposite of Nguyen-Phuoc et al. (2021), who identified financial risk as a significant factor in customer satisfaction. According to the current empirical results of this study, customer satisfaction increased only if the e-banking service contained high value, along with the contribution of low performance risk. Suppose these two main components became better and met customer expectations. In that case, customers might not have had to be concerned much about financial losses because when the e-banking service system had high protection and performed its functions well, the chance of facing financial loss in the e-banking service could be almost zero for customers. Thus, financial risk was not a serious factor in customer satisfaction in the e-banking service industry.
Based on the above discussion, the results of hypotheses testing are summarized in Table 3. Three hypotheses were accepted, while two (1 and 3) were rejected.

CONCLUSION
The objective of this study is to examine the impacts of perceived risk (performance risk and financial risk) and perceived value on customer satisfaction in the e-banking service industry of Cambodia. To fulfill the current objective of this study, this study applies the convenience sampling method to investigate 700 bank customers who are currently using mobile banking or internet banking services at both commercial and retail banks in Cambodia. Therefore, 546 valid responses are finally analyzed using SEM analysis technique.
To sum up, the results of this study have revealed that performance risk significantly influenced perceived value, while financial risk did not display any significant influence on perceived value. Finally, performance risk and perceived value, except financial risk, significantly influenced customer satisfaction. Although both performance risk and perceived value have been found to have significant impacts on e-banking satisfaction, the strength of perceived value on e-banking satisfaction was more potent than the strength of performance risk. To sum up, customers can be happier when they see strong e-banking service performance and more significance of using e-banking service with the bank. In addition, the degree of satisfaction with the current e-banking service among Cambodian customers can severely fluctuate if the banks fail to promote service value to their customers since perceived value has a significant impact on e-banking satisfaction.
In spite of achieving the study's objective, some limitations have been found. For example, the results of this study focused mainly on the e-banking service industry. Applying these results to other industries, such as restaurants, green products, and airline industries, may not be appropriate. Future studies are required to investigate these factors in those industries to come up with new conclusions regarding cus- Note: *** reveals significance level p < 0.001.