On the Factors Influencing Banking Satisfaction and Loyalty: Evidence from Denmark

Abstract

This article proposes and tests a conceptual model examining the antecedents of customer satisfaction and loyalty in the Danish banking sector. Given the changing landscape, the importance of personnel contact, the bank’s website, and the bank’s ethics and reputation are of particular interest. Data were collected through convenience sampling. A total of 1132 usable questionnaires were received. This study reveals that competent personnel, as well as bank ethics and reputation, are major factors influencing customer satisfaction and loyalty. Personal contact with staff is also crucial in shaping customers’ perceptions of the bank’s ethics and reputation. The results also demonstrate that switching barriers have a significant influence on customer satisfaction and loyalty. Finally, the multi-group analysis revealed no significant differences in terms of gender or generation. This study offers valuable insights for retail banks to enhance customer satisfaction and loyalty, demonstrating that competent personnel remains an important driver, even during the shift to digital banking.

1. Introduction

The retail banking market in Denmark has undergone significant changes over the past few decades. Several key factors can characterize these changes: (1) a significant shift towards digital banking facilities; (2) a substantial decrease in the number of banks; (3) declining customer loyalty towards large banks; and (4) the occurrence of several influential external events, including the financial crisis in 2008, the money laundering incidents involving Danske Bank in 2018, and the COVID-19 pandemic in 2020 (KFST, 2022). Due to the high level of digitalization in retail banking, there has been a significant increase in the number of customers using mobile and online banking services. In 2024, 98% of banking customers used digital banking in various forms, such as Netbank, mobile banking, or Mobilepay (Danmarks Statistik, 2024). The competitive landscape of the retail banking sector is characterized by a decline in the number of banks and an increase in market concentration. The number of banks also decreased by about 70% between 1990 and 2019, from 220 in 1990 to 63 in 2019 (KFST, 2022). Meanwhile, customer satisfaction has reached its lowest point since 2006 (EPSI Rating, 2021). These changes have led to a decline in customer satisfaction and loyalty in the banking sector, which is impacting the profitability of Danish banks. However, little is known about which consumer segment had switched their primary banks and why they departed. This phenomenon aligns with the service management literature, suggesting that customer satisfaction plays a crucial role in influencing customer loyalty and, consequently, profitability (Taoana et al., 2022).

Customer loyalty has thus been one of the essential aspects of competition for service organizations. Given the nature of service, the increasing role of technology, and higher customer involvement in the service delivery process, customer loyalty has become even more important (Alok & Srivastava, 2013). Furthermore, customer loyalty is strongly related to the bank’s profitability (Keisidou et al., 2013). Therefore, many businesses often design tailored marketing programs to target consumers and encourage loyalty.

Several researchers have analyzed drivers for customer loyalty in the retail banking sector. Kant et al. (2019) revealed that service quality, perceived value, customer satisfaction, and corporate image are critical antecedents of customer loyalty. Bhat et al. (2018) suggested that customer knowledge and satisfaction have a positive impact on customer trust, and that consumer trust significantly influences loyalty; furthermore, consumer trust mediates the effect of knowledge and satisfaction on loyalty. According to Ebert (2009) and Howard (2023), a retail bank’s likability, competencies, and security significantly influence customer trust, which in turn impacts customer loyalty. Research has also shown that customer satisfaction influences perceived service quality, while customer satisfaction and switching costs directly influence customer loyalty (Moosa & Kashiramka, 2023; Taoana et al., 2022). Furthermore, only a few studies examine the influence of a retail bank’s image and value recognition on customer loyalty. Vyas and Raitani (2014) suggested that bank reputation/image is a significant factor in customers switching to other banks in India. Others suggested that the bank’s image and customer satisfaction have a significant influence on customer loyalty (Amin, 2016; Kant et al., 2017; Liang & Nguyen, 2018). However, Omoregie et al. (2019) found that a bank’s image influences customer satisfaction and trust, not customer loyalty. Lastly, customer perceptions of a company’s ethical reputation positively impact customer satisfaction and loyalty to financial institutions (Mulki & Jaramillo, 2011).

While the extensive literature has investigated the relationship between customer satisfaction and customer loyalty in the banking sector, there is a limited understanding of how emerging factors, such as digitalization and banking ethics, influence these outcomes. Addressing these gaps is important for retail banks to reformulate their marketing strategies and effectively meet customers’ needs and preferences.

Given the changing landscape of retail banking and the aforementioned four contextual issues, this paper aims to identify and assess the antecedents of loyalty through an empirical investigation in Danish retail banking. The results of this study provide a nuanced understanding of customer loyalty. It is expected to provide bank managers with insights into building and maintaining customer loyalty. This knowledge can subsequently help formulate effective marketing strategies to enhance customer retention and loyalty.

The remainder of this paper is structured as follows. Section 2 presents the theoretical framework and develops the hypotheses. Section 3 describes the research method. Section 4 presents an empirical analysis. Section 5 concludes with a summary of the key findings and managerial implications for retail banks.

2. Conceptual Framework and Development of Hypotheses

In this section, we present the theoretical foundation of the proposed model and its associated hypotheses.

2.1. Customer Satisfaction and Its Antecedents

In the service marketing literature, customer satisfaction is defined as “an evaluative judgment from experiencing a specific service encounter” (Ekström, 2010). Customer satisfaction influences consumer behavior through word-of-mouth, repurchases, complaints, etc. (Oliver, 2010). Oliver (2010) examined the theoretical and conceptual foundations of customer satisfaction. Fornell et al. (1996) developed the American Customer Satisfaction Index (ACSI), which serves as an indicator of the customer evaluations of the quality of goods and services. The ACSI model comprises three core components, customer expectations, perceived quality, and perceived value, along with three key outcomes: customer satisfaction, customer complaints, and customer loyalty (Fornell et al., 1996). Yet, there is no consensus about measuring customer satisfaction. The literature has demonstrated that perceived quality significantly influences customer satisfaction (Cronin et al., 2000; Cronin & Taylor, 1992). Perceived quality is “a consumer’s evaluative judgment of a product or service’s overall excellence or superiority” (Zeithaml, 1988). The formation of perceived quality is based on the consumer’s assessment of the service provider’s importance and the performance of service attributes (Parasuraman et al., 1994). The well-known scale for measuring the consumer perception of service quality in the retail banking industry is the SERVQUAL scale (Parasuraman et al., 1988). Although studies have reported that the SERVQUAL scale exhibited good reliability and validity, research conducted in other different contexts (e.g., China, South Korea, and Cyprus) has demonstrated a discrepancy between the empirical results and the original form (Ladhari, 2009). Two important criticisms concerning the SERVQUAL framework are (1) that service quality is measured by using the “gap scores” and (2) the reliability, dimensionality, and validity of the scale (Ladhari, 2009). Therefore, there is a need to adjust the measurement items to fit the research context. Garg et al. (2014) developed a 14-factor scale to measure customer experience in banks, with convenience being the most significant factor influencing customer satisfaction. Both the SERVQUAL scale and the customer experience scale have some similarities. Furthermore, Cronin and Taylor (1992) developed the SERVPERF, a measurement model for service quality that solely focuses on the performance aspect of service. Although the SERVPERF model is based on the same five dimensions of service quality as SERVQUAL, it is often considered more efficient and reliable due to its simplicity (Cronin & Taylor, 1994).

Employees are the primary source of customer service delivery (Garg et al., 2014). The employees’ service skills, such as providing financial services and relevant information to assist customers with their decisions and handling problems and complaints, determine customer trust and, eventually, customer loyalty (Sekhon et al., 2014; van Esterik-Plasmeijer & van Raaij, 2017). Research has also shown that the service skills of employees determine service quality and eventually influence customer satisfaction (Omoregie et al., 2019). Furthermore, van Esterik-Plasmeijer and van Raaij (2017) claimed that technical and marketing competencies are the most important.

Corporate social responsibility activities help form a favorable impression in customers’ minds and build customer trust, positively influencing companies’ corporate reputation and brand equity (Fatma & Rahman, 2016; Fatma et al., 2015). Omoregie et al. (2019) claimed that the corporate image perceived by a customer toward a particular retail bank influences customer satisfaction and loyalty.

The marketing literature suggests that a company’s ethical behavior can be a differentiator in gaining a competitive advantage (Omoregie et al., 2019). A company’s commitment to ethical behavior can enhance its image and reputation, ultimately fostering a favorable impression in the minds of consumers. The existing literature indicates that ethics is positively related to customer satisfaction in a bank (Omoregie et al., 2019). Furthermore, research has shown that a corporation’s reputation and image are crucial in building consumers’ trust in banks (Kant et al., 2017; Omoregie et al., 2019; Osakwe & Yusuf, 2021). Finally, research has shown that ethics and reputation, salesperson and organization, and customer satisfaction are intercorrelated (Ozkan et al., 2020). Hence, we hypothesize the following:

H1. 

Competent personnel are positively associated with customer satisfaction.

H2. 

Banking ethics and reputation have a positive impact on customer satisfaction.

H3. 

Competent personnel andbanking ethics and reputation are intercorrelated.

Due to the high level of digitalization penetration, banks’ websites thus become an essential touchpoint between customers and banks in Denmark. For this reason, we replace the tangible dimension in the original SERVQUAL and the SERVPERF model with the website. A bank’s website can give customers a unique experience every time they log on. The functionality of the bank’s website, including its usability and user interface, has a significant impact on customer experience, as indicated by previous studies (Garg et al., 2014). Research suggests that customer satisfaction is associated with banks’ electronic services (Khatoon et al., 2020; Shaikh et al., 2023; Shankar & Jebarajakirthy, 2019). Thus, the bank website’s functionality can influence customer satisfaction.

Security and privacy protection are key concerns for customers interacting with banks online or offline. Research suggests that security and privacy are important factors in consumers’ choice of online service (Amin, 2016; Ayo et al., 2016; Khatoon et al., 2020; Liang & Nguyen, 2018; Ramesh et al., 2021; Shaikh et al., 2023; Shankar & Jebarajakirthy, 2019; Ul Haq & Awan, 2020). Therefore, like many other electronic businesses, the security and safety of transition are major concerns when shifting toward digital banking (Ramesh et al., 2021). This is because perceived security is important for building consumer trust in banks (Ramesh et al., 2021; Khatoon et al., 2020; Shaikh et al., 2023; Ul Haq & Awan, 2020). We suggest that easy and unproblematic websites may help customers build trust in their banks, thereby enhancing their perceived security. We therefore hypothesize the following:

H4. 

A functional and easy-to-use website is positively associated with customer satisfaction.

H5. 

Perceived security is positively associated with customer satisfaction.

H6. 

A functional and easy-to-use website and security are intercorrelated.

Service convenience refers to consumers’ perceptions of time and effort regarding buying or consuming a service. Convenience is one of the primary elements in crafting customer experiences (Garg et al., 2014). Service convenience can increase consumer satisfaction (Kaura, 2013; Kaura et al., 2015). However, customers who prioritize convenience over other quality antecedents may be less likely to build satisfaction and loyalty.

The interest expense associated with banking services such as loans and deposits is an essential factor in customers’ decision-making about loans (Keisidou et al., 2013). Price is likely an important choice factor in consumer markets with high brand parity. Vanniarajan and Manimaran (2008) found that female customers in retail banks tend to prioritize price and convenience over quality in their evaluation. On the one hand, customers weighing price as the most important factor may be more satisfied with low-priced banks. On the other hand, like convenient locations, customers prioritizing price over other quality dimensions may be less likely to build satisfaction and loyalty. We therefore suggest the following:

H7. 

A convenient location is directly associated with customer satisfaction.

H8. 

Interest rates are directly associated with customer satisfaction.

H9. 

Interest rates are positively associated with customer loyalty.

2.2. Customer Satisfaction and Customer Loyalty

The relationship between customer satisfaction and loyalty has been extensively tested in the extant literature. Customer loyalty results from customer satisfaction (Boonlertvanich, 2019; Cronin & Taylor, 1994; Egala et al., 2021; Haron et al., 2020; Keisidou et al., 2013; Moosa & Kashiramka, 2023; Omoregie et al., 2019; Ozkan et al., 2020; Taoana et al., 2022; Ul Haq & Awan, 2020). Customer loyalty to a bank encompasses both attitudinal loyalty, which reflects behavioral intentions, and behavioral loyalty, which reflects actual behavior (Dick & Basu, 1994; Oliver, 1999). Research has suggested that customer satisfaction with the delivered products and services influences consumers’ decisions to continue (or maintain) a relationship. Ladeira et al. (2016) and Buhler et al. (2024) found a positive correlation between satisfaction and loyalty in meta-analyses. Thus, it can lead to long-term profits by gaining customer loyalty (Fornell, 1992; Khatoon et al., 2020; Omoregie et al., 2019). Furthermore, the marketing literature suggested that satisfied customers are likely to spread positive word of mouth about the service providers and maintain a stable relationship over time (Sirdeshmukh et al., 2002). Notably, although loyal customers are most likely satisfied, satisfaction does not universally lead to customer loyalty (Oliver, 1999; Dawes, 2024). Thus, we suggest the following:

H10. 

Customer satisfaction is positively associated with customer loyalty.

2.3. The Influence of Switching Barriers on Customer Satisfaction and Loyalty

A highly satisfied customer is less likely to switch to another service provider. The relationship between customer switching and loyalty has been widely tested (El-Manstrly, 2016). Switching barriers—defined as any costs (monetary or non-monetary) that customers associate with changing service providers—can serve as an obstacle to exit, preventing customers from leaving their current providers even when dissatisfied (Jones et al., 2007). Jones et al. (2000) suggested that switching barriers encompass not only financial expenses but also time, effort, and psychological factors (e.g., anxiety about adapting to a new system) that the customers perceive as associated with service providers. Research has shown that switching barriers are a primary driver of customer lock-in in banking, reducing competitive dynamics and reinforcing market power within the retail banking sector (Gerrard & Cunningham, 2004). We posit that in such lock-in scenarios, customers may experience cognitive dissonance (Festinger, 1957), i.e., being dissatisfied but perceiving the high costs of switching to another bank. To reduce dissonance, customers may adjust their evaluation and perception of their bank in a positive direction. Thus, we hypothesize the following:

H11. 

Switching barriers have a positive impact on customer satisfaction.

H12. 

Switching barriers have a positive impact on customer loyalty.

Figure 1 presents an overview of the literature related to the constructs included in the proposed model.

3. Methods

3.1. The Questionnaire and Measurement

The data used in this paper were obtained as part of a large online survey on bank choice, satisfaction, and loyalty. The questionnaire consists of two sections. Section one deals with the demographic profile of the respondents, including gender, age, education level, dwelling type, and income. Section two addresses the recent experiences of consumers with banks and questions concerning consumer attitudes towards and preferences for banks. The measurements of customer loyalty and its antecedents were adjusted from previous research by Jensen (2011) and Bhat et al. (2018). Other items in the questionnaire were adapted from the extant literature, together with results from brainstorming: personnel (Garg et al., 2014), convenience (Garg et al., 2014), the bank’s website (Garg et al., 2014), security (Vyas & Raitani, 2014), interests (Vyas & Raitani, 2014), banking ethics and reputation (Mulki & Jaramillo, 2011), switching barriers (Vyas & Raitani, 2014), and customer satisfaction (Vyas & Raitani, 2014). Completing the questionnaire takes about 15–20 min. Table 2 provides a detailed description of the measurement items.

3.2. Data Collection and Sample

The target population consists of adult Danish bank account holders. Data were collected through convenience sampling. Specifically, invitations to participate in the study were disseminated by asking all undergraduate BSc students in Economics and Business Administration enrolled in a quantitative data analysis class at the University of Southern Denmark Business School to share a link to the online survey via email and social media platforms. While this approach allowed for a diverse range of participants, it may have introduced potential biases related to age, education level, and other factors. To increase the heterogeneity of the sample, students were encouraged to share the link with as broad a response group as possible in terms of gender, age, and occupation. After eliminating non-completed questionnaires, 1132 cases were available for further analysis.

Table 1. Descriptive statistics of the respondents.

Demographic Variables	in %
Gender
Female	63.1
Male	36.9
Age
18–29 years old	38.6
30–49 years old	27.0
40–49 years old	15.4
50–59 years old	22.7
Over 60 years old	11.7
Personal annual income before tax
DKK 199,999 or less	32.7
DKK 200,000–399,999	33.0
DKK 400,000–599,999	19.8
DKK 600,000–799,999	4.9
DKK 800,000–999,999	1.6
DKK 1 million or more	1.3
Don’t know/Prefer not to say	6.7

displays the major socio-demographic profile of the sample. The sample is skewed by gender, age, and income, with 63.1% of the total sample being female, 65.6% under 40 years old, and 65.7% having a personal income below DKK 400,000 (approximately EUR 54,000). This reflects the fact that university students distributed links to our questionnaire through their social media and emails. Furthermore, 79.1% of the respondents had only one bank, and 59.1% had maintained the same bank over the past five years.

4. Results

The proposed model was tested using a two-stage approach (Anderson & Gerbing, 1988). First, the measurement was tested by conducting a confirmatory factor analysis of the applied multi-item scales. Next, the measurement model and the structural equation paths were estimated simultaneously to test the proposed model.

4.1. Estimation of the Measurement Model

The results of the parameter estimates are presented in

Table 2. Parameter estimates for the measurement model.

Construct	Standardized Factor Loading	Standard Error	t-Value	Construct Reliability	Extracted Variance
Competent Personnel (ξ1)				0.75	0.51
X1: Personnel in the bank can help/guide me with the financial services I need	0.731
X2: The bank’s advisers provide competent advice	0.746	0.050	19.726
X3: Personnel in the bank show respect and understanding for the customers	0.654	0.048	18.274
Bank Ethics and Reputation (ξ2)				0.76	0.52
X4: The bank has a good reputation/image	0.734
X5: The bank is known for its high level of social responsibility	0.695	0.052	18.738
X6: The bank has high ethical standards, for example, not being involved in money laundering	0.730	0.061	19.199
Website (ξ3)				0.82	0.60
X7: The bank’s website is easy to navigate	0.752
X8: The bank’s website is up-to-date and unproblematic	0.861	0.049	24.392
X9: The website of the bank is easy and clear	0.707	0.045	22.117
Security (ξ4)				0.65	0.49
X10: The bank handles personal information confidentially/safely	0.574
X11: The security standard of the bank’s website is high	0.801	0.125	13.484
Convenient Location (ξ5)				0.78	0.55
X12: The bank is near my domicile	0.807
X13: The bank is easily accessible from my workplace/study location	0.783	0.047	19.646
X14: The bank has branches easily accessible across almost the whole of Denmark	0.596	0.042	17.501
Interest Rates (ξ6)				0.74	0.50
X15: The rates of lending are among the lowest in the market	0.836
X16: The rates of deposit are among the highest in the market	0.511	0.041	14.076
X17: It is possible to get a loan under good conditions	0.677	0.049	16.522
Switching barriers (ξ7)				0.68	0.59
X18: I believe that it will be pretty difficult for me to get used to being a customer in another bank than my current primary bank	0.675
X19: I feel that I will waste the time I have spent familiarizing myself with my primary bank if I switch to another bank	0.754	0.085	12.716
Customer satisfaction (η1)				0.79	0.56
Y1: I am confident that my current primary bank meets my needs better than others	0.740
Y2: How do you evaluate your primary bank in general, as compared to other banks in Denmark	0.769	0.057	23.956
Y3: How satisfied are you with your current primary bank	0.735	0.043	22.976
Customer loyalty (η2)				0.79	0.55
Y4: Have you ever considered choosing another bank as your primary bank in the past three years	0.743
Y5: How likely is it for you that your current primary bank will continue to be your primary bank in the next five years	0.790	0.032	24.766
Y6: Even if my family or friends recommend switching to another bank, I will definitely choose to remain with my current bank	0.688	0.037	21.736

Notes: Except for Y₂,Y₃, Y₄, and Y₅, a five-point Likert Scale, ranging from “1 = strongly disagree” to “5 = strongly agree”, was used as a measurement scale. The measurement scale for Y₂ is a 7-point scale, in the range of “1 = much worse than other banks, 4 = equally good, and 7 = much better than other banks”. Y₃ is assessed on a five-point scale, ranging from “1 = Very unsatisfied to 5 = Very satisfied”. Y₄ is measured on a five-point scale, in the range of “1 = No, it has never been in my consideration, 2 = Yes, but the thought comes out of my mind immediately, 3 = Yes, but I figured out quickly that it was better for me to keep my current primary bank, 4 = Yes, I am still thinking about switching to another bank, and 5 = Yes, I thought so that I will switch my current bank soon”. The measurement scale for Y₅ is a five-point scale, ranging from “1 = totally unlikely ” to “5 = totally likely”.

. Eight items have a factor loading below the recommended threshold of 0.7, suggesting they may warrant consideration for removal (Hair et al., 2021). Nonetheless, Hair et al. (2021) advise caution when removing items with loadings of 0.40–0.708, recommending deletion only if it leads to an improvement in internal consistency reliability or convergent validity. Otherwise, these items can be retained (Hair et al., 2021). Furthermore, Howard (2023), through a systematic literature review of exploratory factor analysis in management, found that factor loadings deemed moderate were between 0.50 and 0.64, those deemed large were between 0.64 and 0.77, and those deemed very large were above 0.77. Consequently, Howard (2023) proposed a cutoff of ≥0.50. In light of this, we decided to retain these eight items. Moreover, all items were statistically significant, indicating that the chosen generic items for each latent variable represent a unidimensional construct. All composite reliabilities and extracted variances were above or very close to the 0.70 and 0.50 thresholds (Hair et al., 2006), respectively, demonstrating that convergent validity was achieved. The fit measures, CFI, GFI, and NFI, were all above 0.90 and close to 0.95, indicating a good fit of the model. RMSEA was below 0.05, supporting that the model fits the data well (Hair et al., 2006).

The discriminant validity of the applied constructs was tested using the method proposed by Fornell and Larcker (1981).

Table 3. Discriminant validity results.

	ξ1	ξ2	ξ3	ξ4	ξ5	ξ6	ξ7	η1	η2
ξ1	0.71
ξ2	0.41	0.72
ξ3	0.40	0.30	0.77
ξ4	0.57	0.41	0.57	0.70
ξ5	0.14	0.20	0.20	0.10	0.74
ξ6	0.61	0.20	0.22	0.24	0.14	0.71
ξ7	0.14	0.37	0.10	0.10	0.26	0.10	0.77
η1	0.26	0.17	0.10	0.00	0.00	0.00	0.44	0.75
η2	0.14	0.10	0.10	0.00	0.10	0.17	0.53	0.67	0.74

displays the square root of the average variance extracted (AVE) for each construct on the diagonals and the inter-construct correlations in the non-diagonals. The results confirm that the square root of the AVE for each construct exceeds its highest correlation with any other construct, satisfying the Fornell–Larcker criterion. This demonstrates that each construct shares more variance with its own indicators than with others, which empirically supports discriminant validity.

4.2. Hypothesis Testing

The structural model was estimated using SPSS AMOS version 29. The estimation results are presented in

Table 4. Estimation of the structural model.

Hypotheses	Construct Regression Relationships	Std. Estimate	S.E.	C.R.	Decision
H1	Competent Personnel → Customer Satisfaction	0.250	0.061	5.72 ***	Accept
H2	Bank Ethics and Reputation → Customer Satisfaction	0.104	0.048	2.49 *	Accept
H4	Bank’s Website → Customer Satisfaction	0.010	0.044	0.268	Reject
H5	Security → Customer Satisfaction	0.055	0.035	1.17	Reject
H7	Convenient Location → Customer Satisfaction	−0.112	0.032	−3.00 ***	Accept
H8	Interest Rates → Customer Satisfaction	−0.119	0.036	−2.56 ***	Accept
H9	Interest Rates → Customer Loyalty	−0.165	0.032	−5.84 ***	Accept
H10	Customer Satisfaction → Customer Loyalty	0.830	0.058	18.60 ***	Accept
H11	Switching Barriers → Customer Loyalty	0.163	0.038	4.749 ***	Accept
H12	Switching Barriers → Customer Satisfaction	0.423	0.039	9.127 ***	Accept
Hypotheses	Construct Correlation Relationships	Std. Estimate	S.E.	C.R.	Decision
H3	Competent Personnel ←→ Bank Ethics and Reputation	0.418	0.016	9.25 ***	Accept
H6	Security ←→ Bank’s Website	0.375	0.018	12.49 ***	Accept
Modification indices	Convenient Location ←→ Switching barriers	0.249	0.034	5.828 ***	Accept
Modification indices	Bank’s Website ←→ Convenient Location	0.152	0.018	4.602 ***	Accept
Modification indices	Bank Ethics and Reputation ←→ Interest Rates	0.362	0.024	8.534 ***	Accept
Modification indices	Competent Personnel ←→ Interest Rates	0.381	0.020	8.979 ***	Accept
Explained proportion of construct variance		R2
Customer satisfaction		0.248
Customer loyalty		0.845

The model fit indices λ²(260) = 1174.7; λ²/df = 4.514, p < 0.05; CFI = 0.902, GFI = 0.922, AGFI = 0.902; RMSEA = 0.056, HI(90) = 0.059. Note: *, ***denotes the significance levels at 0.05 and 0.001, respectively.

.

The initial test of the conceptual model presented in Figure 1 revealed that the model did not fit the data as well as desired, based on model fit criteria (i.e., several fit measures were below the 0.90 threshold). However, modification indices suggested including four additional correlation paths in the model (Convenient Location ←→ Switching Barriers, Convenient Location ←→ Bank’s Website, Interest Rates ←→ Bank Ethics and Reputation, and Competent Personnel ←→ Interest Rates). The suggested correlations are theoretically plausible. First, customers who find their bank’s location convenient may worry that switching could lead to a less convenient option. Similarly, those who find the bank’s website easy to navigate may see learning a new site as a switching barrier. Customers who view their bank as ethical are less likely to believe it will charge high interest rates. Finally, competent bank staff can foster customer trust, reducing perceptions of unfavorable interest rates.

Based on the above considerations, it was decided that the four suggested correlations can be added to the proposed model. Table 4 shows the overall results of the modified model. The model fit indices CFI, GFI, and AGFI are all above the 0.90 threshold level (0.902, 0.922, and 0.902, respectively), and RMSEA is below the 0.60 threshold level (RMSEA = 0.056; HI(90) = 0.059).

Starting with H1, competent personnel are significant and positively correlated with customer satisfaction (β = 0.25; p < 0.001), suggesting that personnel contact is an important factor in building customer satisfaction. Bank ethics and reputation is significant and positively correlated with customer satisfaction (β = 0.10; p < 0.05), indicating that good ethics and reputation are important for building customer satisfaction in the banking sector. Thus, H2 is supported. Surprisingly, neither banks’ websites (β = 0.01; p > 0.05) nor security (β = 0.06; p > 0.05) is significant for satisfaction, meaning that H4 and H5 are rejected. This may be because Danish banks’ websites have high security standards, leading consumers to share similar perceptions and a general sense of satisfaction. A convenient location is significant and negatively correlated with customer satisfaction (β = −0.11; p < 0.001), supporting H7 and suggesting that individuals choosing their bank primarily for convenience are less likely to build satisfaction with their choice. Interest rates are significant and negatively correlated with customer satisfaction (β = −0.12; p < 0.001), thus supporting H8. Likewise, interest rates are also significant and negatively correlated with loyalty (β = −0.17; p < 0.001), thus supporting H9. This suggests that in price-sensitive markets with high levels of parity, it is difficult to build customer satisfaction and loyalty. Switching barriers are significant and positively correlated with satisfaction (β = 0.42; p < 0.001), supporting H12. It is also significant and positively correlated with loyalty (β = 0.16; p < 0.001), supporting H11 and indicating that customers facing high switching barriers are more likely to remain with their current bank. Interestingly, the effect of switching barriers on satisfaction is significantly greater than on loyalty. However, this difference should be viewed in light of the fact that satisfaction has a very strong effect on loyalty (β = 0.83; p < 0.001), supporting H10.

Competent personnel and banking ethics and reputation are significantly and positively correlated (r = 0.42; p < 0.001), supporting H3. This suggests that personnel play a crucial role in persuading customers about the bank’s ethics and reputation. Security and bank websites are significantly and positively correlated (r = 0.38; p < 0.001), supporting H6. This suggests that the bank’s website is an important means of assuring customers about security.

Turning to the four correlations suggested by the modification indices, convenience is significant and positively correlated with switching barriers (r = 0.25; p < 0.001), suggesting that customers perceive switching from a bank to a more convenient location as a potential barrier to switching. The bank’s website is significantly and positively correlated with convenience (r = 0.15; p < 0.001), indicating that an easy-to-navigate website reduces customers’ concerns about the convenience of their bank’s location. Bank ethics and reputation are significant and positively correlated with interest (r = 0.36; p < 0.001), suggesting that if customers trust that their bank has good ethics, they are more likely to view the interest rates as fair. Finally, competent personnel are significant and positively correlated with interest (r = 0.38; p < 0.001), suggesting that competent personnel help convince customers that the interest rates are fair.

5. Conclusions and Discussion

The purpose of this study was to investigate the influence of key antecedents on customer satisfaction and loyalty in the evolving Danish banking sector. We proposed a conceptual model with seven antecedents and hypothesized their relationship with customer satisfaction and loyalty. Five antecedents were found to be significantly related to customer satisfaction and loyalty, as expected. Competent personnel and the bank’s ethics and reputation were positively associated with customer satisfaction, with competent personnel having the strongest influence. Personal interactions shape perceptions of institutional ethics. Moreover, those two antecedents were positively intercorrelated, suggesting that personal contact shapes customers’ perception of the banks’ ethics and reputation. This finding is consistent with van Esterik-Plasmeijer and van Raaij (2017), who noted that employee behavior has a direct impact on corporate image in high-trust environments. It also aligns with previous studies that emphasize employee competence and ethical branding as key drivers of customer satisfaction (Omoregie et al., 2019; Ozkan et al., 2020). Similarly, Sekhon et al. (2014) found that customers often link employee skills with trust-building in service industries. The influences of banks’ websites and security were not supported; however, both were significantly and positively intercorrelated, indicating that a functional website may enhance feelings of security. The lack of impact on customer satisfaction from security may be due to Denmark’s high levels of bank trust, which are among the highest in the world (Anneli Järvinen, 2014). Similarly, the absence of influence from a bank’s website on customer satisfaction may stem from a well-regulated and generally high level of web security standards across Danish banks, making this tangible factor more of a hygiene factor that could lead to dissatisfaction rather than the motivational factor to create satisfaction (Herzberg et al., 1959). This aligns with Shaikh et al. (2023), who found similar results in technologically advanced markets. Additionally, the negative and significant correlations between the two tangible factors—convenient location and interest rate—may be explained by Hertzberg’s two-factor theory. The absence of physical offices triggers frustration, but their presence is expected. The shift towards more online banking has made physical branches less important, while intense competition among Danish banks has led to similar interest rates, creating high parity in these areas.

In line with past research, customer satisfaction significantly and positively correlates with customer loyalty. Switching barriers are significantly and positively correlated with both customer satisfaction and customer loyalty. This supports prior research indicating that when customers perceive a high switching barrier, they are more likely to remain with the current service provider rather than switch to another, even if they are less satisfied (Burnham et al., 2003). The positive relationship between switching barriers and customer satisfaction can be explained by Festinger’s cognitive dissonance theory, which states that when individuals experience discrepancies between their beliefs, attitudes, and behaviors, they will strive to reduce these conflicts by changing one of these factors (Festinger, 1957). Bank customers who feel locked into their current bank due to high perceived switching barriers may increase their satisfaction levels to reduce the conflict between their satisfaction and loyalty to the bank. The significant positive correlation between convenient location and switching barriers suggests that bank customers may perceive this factor as a potential barrier to switching when considering a replacement for their current bank. This extends Burnham et al. (2003)’s work by showing that barriers influence both behavior and perception, particularly in locked-in markets.

The results of this study provide several theoretical and practical implications, which are outlined below.

5.1. Theoretical Contributions

This study focuses on the Danish banking sector, known for its high competitiveness and widespread adoption of digital banking, providing new insights compared to research in less competitive and less digitalized markets. Our study refined Herzberg’s theory for digital services by demonstrating that websites and their associated security function as hygiene factors in a highly digitalized market. Few studies have included websites and bank ethics as antecedents of customer satisfaction and loyalty, and even fewer have investigated the influence of switching barriers on these factors. The dual impact of switching barriers (on loyalty and satisfaction) advances the cognitive dissonance theory, suggesting that locked-in customers rationalize their satisfaction to align with behavioral inertia. We suggest that some observed relationships can be explained by Herzberg’s two-factor model and Festinger’s cognitive dissonance theory, underscoring the need for future research to investigate the specific roles of these theories in customer satisfaction and loyalty. In high-trust markets, ethics/reputation may not differentiate satisfaction but remain critical for maintaining trust—a nuance absent in prior studies (Omoregie et al., 2019). Finally, the study reveals that some of the antecedents may be intercorrelated; therefore, future research should not assume that these factors are independent of each other.

5.2. Managerial Implications

The study results have important implications for retail banks. Firstly, the absence of significant differences related to gender and generation suggests that Danish bank managers should treat their markets as relatively homogeneous. Secondly, the findings suggest that banks should differentiate themselves from competitors by emphasizing non-financial and intangible factors, such as competent personnel and their commitment to ethical and social responsibility. Banks should prioritize investing in employee training to improve advisory skills and provide empathetic service, as these factors directly influence satisfaction and ethical perceptions. Banks should also engage in socially responsible activities and offer personalized services to enhance branding. Banks can also design a customer loyalty program by focusing on non-monetary barriers (e.g., customized service) rather than financial costs. The results also suggest that tangible factors such as convenient location, bank website, and interest rates do not drive satisfaction and should not be considered primary differentiators. However, it is important to note that experiencing problems with the website may lead to dissatisfaction.

Finally, research also suggests that effective retail bank management is crucial for fostering an ethical climate (Jaramillo et al., 2006; Mulki & Jaramillo, 2011). Management that supports the employees’ well-being can help create an organizational culture that operates at a high ethical level. Thus, the results also have implications for the creation of bank leadership.

5.3. Limitations and Future Research Directions

This study has limitations that suggest directions for future research. Firstly, the data were collected from a convenience sample of bank customers in Denmark. Although we encouraged our students to share our survey with a broad audience in terms of demographic background, this approach can introduce limitations due to selection bias—our sample overrepresents individuals such as students at higher educational institutions in Denmark, who tend to be younger, have higher education, and report relatively lower income compared to the general population. As a result, convenience sampling limits the generalizability of the findings. Future research could adopt a stratified sampling technique, systematically selecting participants from key demographic subgroups to ensure proportional representation. This method could enhance external validity and mitigate biases inherent in convenience sampling. Secondly, future studies could extend the current model by incorporating other antecedents of customer satisfaction, such as service failures and service recoveries. These factors play an important role in shaping customer experiences and perceptions, ultimately influencing their overall satisfaction levels. The extension of this model would enhance our understanding of the factors that contribute to positive customer experience and loyalty. Thirdly, research has shown that ethical perceptions are closely associated with cultural backgrounds (Armstrong, 1996). Thus, it may be worthwhile to test the transferability of the current model in different cultures and/or markets that are less digitalized. Such comparisons could reveal how ethical standards and expectations vary among societies and how these variations influence customer satisfaction and loyalty. Such research would not only enhance the model’s applicability in different markets but also contribute to a more nuanced understanding of the interplay between ethics, culture, and consumer behavior in various business environments.