Banks’ distributional channel strategies, customer satisfaction and mediating role of trust

Размещено на http://www.allbest.ru/

Размещено на http://www.allbest.ru/

Modibbo Adama University, Yola. Adamawa State-Nigeria

Department of Business Administration

Banks' distributional channel strategies, customer satisfaction and mediating role of trust

Haruna Isa Mohammad

Kamarudeen Babatunde Bello

Abstract

The purpose of this paper is to investigate the impact of perception of banks' distributional channel strategies (Human Teller, ATM-Banking, POS-banking, Mobile banking, internet banking) on customer satisfaction; and the mediating effects of trust in these relationships. Cross-sectional (survey) data was obtained from a sample of 352 retail bank customers in north-east region of Nigeria, using pre-validated measures. The ADANCO 2.0.1 analytical package was employed in conducting the tests of hypotheses. Results indicate that only the human teller and internet banking channels influence customer satisfaction directly and indirectly through customer trust; while the influence of mobile banking channel on customer satisfaction is fully mediated by trust. The research findings have further confirmed the irreplaceability of the human element in the delivery of banking services. Bank managers should emphasize continuous training and orientation of employees towards effective service delivery and relationship building. The findings also point to the need for bank managers to devise innovative means of improving the quality of services through self-service channels (ATM, POS and Mobile), and communicate such to customers; while putting measures in place to win the trust of customers for the channels and the entire services of the banks. This paper contributes to the body of knowledge on customer perceptions and response in a multichannel banking environment, and the mediating role of trust, particularly from developing country perspective.

Key words: customer satisfaction, distributional channel strategies, trust, banks, Nigeria.

Стратегії банків щодо каналів розподілу, задоволеності клієнтів і посередницька роль довіри

Анотація

bank service customer satisfaction

Мета даної роботи - дослідити вплив сприйняття стратегій банківських каналів розподілу (людський касир, банкомат-банкінг, POS-банкінг, мобільний банкінг, інтернет-банкінг) на задоволеність клієнтів; а також опосередковуючий вплив довіри в цих відносинах. Дані крос-секційного (опитувального) дослідження було отримано на вибірці з 352 роздрібних банківських клієнтів у північно-східному регіоні Нігерії з використанням попередньо перевірених заходів. Для перевірки гіпотез використовувався аналітичний пакет ADANCO 2.0.1. Результати показують, що тільки канали банківського обслуговування за участю операційістів та інтернет-банкінгу впливають на задоволеність клієнтів прямо та опосередковано через довіру клієнтів; тоді як вплив каналу мобільного банкінгу на задоволення клієнтів повністю опосередковано довірою. Результати дослідження ще раз підтвердили незамінність людського фактора у наданні банківських послуг. Керівники банків повинні приділяти особливу увагу постійному навчанню та орієнтації працівників на ефективне надання послуг та побудову відносин. Результати дослідження також вказують на необхідність для менеджерів банків розробляти інноваційні способи підвищення якості послуг через канали самообслуговування (банкомати, POS та мобільні пристрої) та повідомляти про це клієнтів; водночас необхідно вживати заходів для завоювання довіри клієнтів до каналів та всіх послуг банків. Ця стаття робить внесок у скарбничку знань про сприйняття та реакції клієнтів у багатоканальному банківському середовищі, а також про посередницьку роль довіри, особливо з погляду країн, що розвиваються.

Ключові слова: задоволеність клієнтів, стратегії розподільчих каналів, довіра, банки, Нігерія.

Introduction

Developments in the banking and financial services sectors have seen the deployment of multiple channels for service delivery as a means of creating value and satisfaction for customers. These channels, comprising both the traditional human teller and the automated delivery channels, have been predicted as capable to “dramatically change the distributional structure of retail banks” (Mols, 1998:331), through their influence on the nature and extent of competition for customer patronage. In the Nigerian banking industry, deployment and availability of multiple channels for service delivery have long become a major selling point for banks in their communication efforts to customers. Evidences to the differentiation powers of multiple distribution channels in the Nigerian banking industry began manifesting around 1990s when the deployment of automated service channels (such as the automated teller machines (ATM), Telephone banking, and internet banking) became a basis for classifying banks into old generation vs new generation banks dichotomies (Bello et al., 2014). Researchers are unanimous in declaring that the emergence of the new generation banks, with their multiple distribution channel strategy marking a new turning point for the industry, as it not only forced the older banks to deploy the same technologies, but also made the banks compete against one another in being the first to deploy subsequent channels as they evolve.

Available statistics reveals a large-scale adoption of multichannel banking by bank customers in Nigeria. A report by KPMG (2018) reveals that 65% of bank customers make use of multichannel distribution, as against only 2% who use electronic banking exclusively. To sustain this patronage, bank managers have been paying considerable attention to customer satisfaction with the quality and performance of the channels. In the same vein, several researchers (e.g., Adewoye, 2013; Adeyemi, Ola & Oyewole, 2014; Tijani Ilugbemi, 2015; Oluwagbemi, Abah & Achimuga, 2011) have conducted researches in distributional channels of banks in Nigeria. However, a major observation on these studies is that they have adopted a disaggregated approach by focusing mostly on one (or a combination) of the automated channels, without considering the traditional channel of human tellers. Thus, creating the need for a study that captures both the traditional and automated channels, to provide a holistic view of customers' perceptions of banks' distributional channel strategies in Nigeria.

From the international perspective, Al-Hawari, Ward and Newby (2009) have also observed the dearth of studies examining customer service quality responses to traditional and automated channels in one model. Hence, their study proposed a “comprehensive” model encompassing the ATM, Telephone banking, internet banking, and traditional channel service attributes (p. 456). However, the time-lag between 2009 and the present has seen the deployment of additional channels such as mobile banking and the POS-agent banking, which have their own unique features; thus, creating a lacuna in the knowledge-base on consumer response to banks' channel strategies. Similarly, Patricio et al (2003) had earlier assessed the contribution of individual channels to customer satisfaction in a multi-channel setting. However, mobile banking was not included and the authors used interview as a means of data collection with small sample size. Furthermore, the authors called for quantitative studies in relation to multi- distributional channel.

Beyond the number of distribution channels studied, this study also contributes to the literature by including trust as a mediating variable. Consumer trust is defined as a psychological state relating to the customer's belief that a product or service provider can be relied on and to exhibit good behavioral intentions in serving the long-term interests of the customer (Crosby et al., 1990; Morgan & Hunt, 1994), particularly in situations where the customer is exposed to some degree of risks (Delgado-Ballester, 2004). Pemartin and Rodriguez-Escudero (2020) state that such positive psychological state “can either trigger or result from” specific actions of transacting partners; thus, placing trust as an effective mediator of customers' attitudinal/behavioral intentions. Trust connection is particularly important to the context of this study, given the reported incidences of frauds across the traditional and automated banking channels in Nigeria (Bello & Alkali, 2020). Hence, given the reported growth in customers' use of the multiple service distributional channels, it is logical to examine the level of customers' trust in these channels, and its intervening role in relation to customer satisfaction with the banking distribution channels.

Based on the foregoing, this study proposes an integrative model to explore customers' perceptions of banks' distributional channel strategies, customer satisfaction and the mediating effect of trust in the Nigerian banking industry. The subsequent section presents a review of literature focusing banks' distribution channel strategies, customer satisfaction and trust, along with the hypotheses development and conceptual framework in section 3. This is followed by discussion of the study methodology in Section 4, research results and discussion in sections 5 and 6 respectively; and lastly, implications, conclusions and future studies in Section 7.

Literature review

Distributional channel strategy for banks could be conceptualized as any means used to increase the availability and/or convenience of services, which help to maintain existing bank users or increase their use among current and potential clients (Osuagwu, 2008). Besides the provision of quality services, banks are also concerned about customer convenience - in terms of location and timing - of obtaining services. Traditionally, banks' distributional decisions were centered on factors like location of headquarters and its proximity to the network of branches, the proximity of branches to customers, and operating hours (Murowaniecki, 2015). Over the years, innovations in information and communication technologies have created opportunities for additional channels for distributing services to customers, via such platforms as the Automated Teller Machine (ATM), the Point of Sale (POS) Terminals, the mobile banking and internet banking. This development has given rise to the need for banks to evolve strategies for choosing and combining these channels along with the traditional human tellers for service distribution, in a manner that ensures customer positive attributions and satisfaction.

Customer satisfaction

Customer satisfaction is an important issue for marketing organizations (Delacroix and Guillard, 2016). Marketing literature defines customer satisfaction as reflecting the degree of positive/negative feelings consumer experiences after using a product or service encounter (Kotler, 2003). Thus, customer satisfaction is said to have taken place when customers' expectations are met, or exceeded (Anderson & Fornell, 1994); and the opposite (i.e., dissatisfaction) occurs when the product or service encounter has fallen short of customer's expectations. Generally, customer satisfaction has always been regarded as one of the most important goals/performance measures of business (Anderson, Fornell & Lehmann, 1994). It is a variable that has significant impact on customers' positive post patronage behavior (Islam et al., 2021), culminating into loyalty and customer retention; and also on business growth and profitability (Dauda & Lee, 2016).

Conceptualizations of customer satisfaction have revealed two major aspects: transaction- specific and cumulative (Kaura & Sharma, 2015). Transaction-specific satisfaction is allied to the evaluation of particular service encounter, and tends to vary from one experience to another (Veloutsou et al., 2005). Cumulative satisfaction on the hand is associated with the overall evaluation of service provider to date (Kaura & Sharma, 2015). It is often based on all past encounters with the product or service provider, and usually links the customer's satisfaction with service encounters to other facets of the service provider (Veloutsou et al., 2005). For this reason, most researchers consider customer satisfaction as a `snowballing' construct arising from a universal assessment of all the features that make up the customer association with the service supplier rather than being a transaction-specific phenomenon' (Anderson, Fornell, & Lehmann, 1994; Manon, Sandrine, Isabelle, & Lova, 2017, Thakur, 2014).

Arising from the arguments above, this study aligns with the cumulative perspective of the satisfaction construct, and views consumer satisfaction as encompassing the totality of consumer's emotions/feelings regarding all encounters with the distributional channels to date. Furthermore, the study also adopts the expectancy-disconfirmation theory of satisfaction (Oliver, 1980), regarded as the most widely applied theory in satisfaction studies. The theory postulates that customers would typically “compare a new service experience with some standard that they have developed” (Barsky, 1992: 53). This process takes place through the interplay of customers' expectations regarding the service, its perceived performance, disconfirmation (or confirmation) - resulting from comparison between prior expectation and performance, leading to satisfaction or dissatisfaction (Oliver, 1980). Hence, the extent to which the service experience measured up to the preconceived standard determines whether the customer is satisfied or not with the encounter (Barsky, 1992;

Holjevac, Markovic & Raspor, 2013). To this end, this study proposes that consumers' satisfaction with the respective distribution channels will be based on the comparison between the preconceived expectations they have formed regarding the channels and the performance they have experienced while using the channels. The associated hypotheses are presented in the subsequent subsections.

The Human Teller

Human tellers (HT) otherwise known as branch-based-banking or mortar and brick banking is the first physical point of banking contact. Human tellers are the human element in the banking hall (Moutinho & Smith 2000). Human tellers help customers withdraw, deposit, transfer money, pay bills, attend to customers' problems and facilitate loans among others. Despite proliferations of different technology oriented bank distributional channels, human tellers have remained an important channel of distribution. Osuagwu (2008) noted that, high branch-banking network proved the relevance of human tellers and the channel appeals to large number of customers in Nigeria. Report from KPMG (2017) shows that 75% bank customers in Nigeria mostly visit bank branches to lay complaints or engage in financial transactions. A major unique attribute of the Human teller is its social orientation through face-to-face interaction. It is this social orientation attribute that perhaps makes most bank customers perceive human teller channel as less risky, since problems arising from service encounters can be resolved at the point of transaction (Patricio et al., 2003) indicated that human teller has less risk, because problems are resolved at the point of transaction. On this basis, the study hypothesizes as follows:

H1: Human teller channel positively influences customer satisfaction.

Point Of Sale (POS) Terminals

The POS Terminal is an electronic device that could be used for verifying and processing credit card transactions. A Retail Point of Sales system typically includes a computer, monitor, cash drawer, receipt printer, customer display, a barcode scanner, and a debit/credit card reader. In some instances, the POS system may also include a weighing scale, integrated credit card processing system, a signature capture device and a customer pin pad device (Okechi & Kepeghom, 2013). This combination of accessories elevates the POS as a form of electronic banking platform that allows customers handle financial transactions without visiting the bank or even the ATM. Various uses of this form of e-banking include cheque verification, credit authorization, cash deposit and withdrawal and cash payment, among others. Recently, the POS has gained prominence as means for actualizing funds transfer and payment at the point of sales such as supermarkets, petrol stations, hospitals, restaurants, among others. The main advantage of POS is that it enables customers pay for goods and services without cash, as the purchase price would be debited on the buyer's card and credited on the seller's account (Olanipekun, Brimah & Ajagbe, 2013). In addition, POS-banking is also made available through agents who set up kiosks in strategic locations in major towns and cities across the country; and providing easy access to cash for individuals in areas too far from bank branches and ATMs. Hence, it is expected that customers would derive immeasurable values from the use of the POS system, thus, increasing the chances for satisfaction. To this end, the paper proposes the following hypothesis.

H2: The POS banking channel positively influences customer satisfaction.

The Automated Teller Machine (ATM)

ATM is one of the electronic distributional channels deployed by banks that have reduced banking hall congestion and made it easier for customers to withdraw cash and make other bank transaction. Narteh (2013) defined ATM as an electronic device, which permits customers deposit, withdraw, transfer money, pay bills and engage in other financial transaction without being present in the banking hall. The ATM is widely acknowledged as having achieved large-scale acceptance and favorable attitude from customers. Leonard and Spencer (1991) associated ATM with improving the image of banks, just as Moutinho and Brownlie (1989) reported that customers place high values on the locational advantages of ATM, leading to higher levels of satisfaction. Similarly, Adeyemi, Ola and Oyewole (2014) in comparing of ATM and human teller, indicated that ATMs are mostly located in areas easy to reach for customers, compared to the human teller that needs banking hall for transactions to take place.

The popularity of the ATM has made it a focus of numerous studies. Hence, despite its positive links with customer satisfaction in prior studies, some scholars have indicated problems with ATMs, which could reduce or prevent customer satisfaction. Murdock and Franz have highlighted some difficulties and security risks associated with ATM, while Mark and Luiz (1995) reported that customers' needs more functionality on ATMs to enable them enjoy banking services. Within the Nigerian context, Alaba (2011) pointed out issues of fraud and lack of knowledge in using ATMs as affecting its adoption in the country. Other problems with ATMs are high bank charges, failure of machines and the long process of recovering the stock cash. Nevertheless, evidences abound that the ATM has remained the most prominent customer-interfacing channel within the Nigerian banking system (Bello, Danjuam & Udo-Imeh, 2014). Thus, the following hypothesis is proposed.

H3: ATM banking channel positively influences customer satisfaction.

Mobile banking

Mobile banking (m-banking) refers to all forms of services offered by financial institutions that make use of technology-enabled portable devices such as mobile phones, smart phones, or tablets (Makanyeza, 2017; Tam & Oliveira 2017). Mobile banking is regarded as one of the important recent strategic innovations in retail banking (Tam & Oliveira 2017); and its use cuts across developed and developing countries (Makanyeza, 2017). Mobile banking has significantly improved banking delivery through varieties of banking features, such as alerts, notifications and geographical benefits (Ha, Canedoli, Baur & Bick, 2012), which allows customers to enjoy bank services anywhere and anytime (Tam & Oliveira 2017). These multiple benefits derivable from using mobile banking through cell phones increases the likelihood for customer satisfaction with the banking channel.

M-banking became a prominent feature of the Nigerian financial system due to the proliferation of Global system for mobile communication (GSM), (Adeyemi, Ola & Oyewole, 2014)

. Recent data from the Nigerian telecommunication commission indicates that there are about 199,863827 active mobile lines in the country as at January 2021 (www.ncc.gov.ng), representing over 90% of the country's population. In the same vain, reports from the Nigerian interbank settlement scheme (NIBSS) have indicated that mobile channel have attained the status of most preferred channel having recorded consistent growth over the years; culminating into an increase of 84.6% from 506.16 million transactions in 2019 to 933.66 million transactions in 2020 (The Punch.com, 2021). This seeming largescale acceptance of mobile banking is consistent with Adewoye's (2013) assertion that the channel helps bank to improve their services, thus leading to customer satisfaction. Therefore, this study hypothesizes as follows.

H4: Mobile banking channel positively influences customer satisfaction.

The Internet Banking

Internet banking refers to banking process that enables a customer to engage in banking activities with electronically connected internet through bank website provided by the financial firms, and normally accessed via a the computer and smart phone devices (Shaikh & Karjaluoto, 2015)

. It is generally regarded as a channel with wide acceptability among customers (Moghavveni, Lee and Lee, 2018); because it has changed the way customers transact with banks and made it easier for them to access account, transfer money, pay bills, engage in investment and establish inquiries as long as there is internet connectivity, from their comfort zones (Lee and Chung, 2009; Mols, 1998).

The ubiquitous nature of internet banking has attracted research interests on its adoption/acceptance and customer attitudes. Countless researchers have reported large scale acceptance as well as a positive relationship between internet banking and customer satisfaction. A sizeable number of these studies have focused on quality of internet banking as major drivers of customer satisfaction (e.g., Amin, 2016; Ankit, 2011; Harington and Weaven, 2009; Jun and Can, 2001; Pikkarainen, Pikkarainen, Karjaluoto, and Pahnila, 2004; Rod, Ashill, Shao and Carruthers, 2009); while some studies have established a link between certain attributes of the internet banking and customer satisfaction. A typical example of the latter group of studies is Ramseook-Muhuerrum and Naida (2011), who found that ability of customers to access and perform banking anywhere as some reasons that attract and satisfy customers.

Despite the wide acceptance and satisfaction with Internet as channel of banking service distribution, it is not without its share of reported challenges that could reduce or prevent customers from getting the desired satisfaction. For example, securities issues have been reported as greatly affecting the utilization and satisfaction of customers (Jun & Cai, 2001; Liao & Cheug 2002; Ramseook-Munhurum & Naido 2011). Similarly, Chen, Husia, and Hwang (2012) in their study of evaluating the satisfaction of customers using the internet in Taiwan, recommended that banks should design a friendly user internet and ensure strong internet security. Notwithstanding these challenges, the current study aligns with the prevalent position that internet banking is directly linked to customer satisfaction, to the extent that it is considered to influence customer satisfaction better than human teller (Polatoglu & Ekin 2001). Hence, the following is hypothesized.

H5: Internet banking channel positively influences customer satisfaction.

Mediating Role of Trust on the Relationship between Distribution Channels and Customer Satisfaction

Trust is generally conceived as an idividual's belief in the ability of another, to deliver on expectations, as well as his/her honesty to act in a manner that would not jeopardize one's interest. In business relationship, trust is viewed as customer confidence in vendor's openness and honesty in transactions (Morgan & Hunt, 1994; De Wulf, Odekerken-Schroder & Iacobucci, 2001). Trust reduces risk and uncertainty associated with commercial transaction (Huurne, Ronteltap, Corten, & Buskens, 2017). Delgado-Ballester (2003:574) specifically conceptualize trust as the “the confident expectations of the brand's reliability and intentions in situations entailing risk to the consumer". This definition identifies two important ingredients associated with trust in business relationships in general, and banking relationships in particular. These are the `confidence' that the buying party has in the selling party, and the recognition of `potential risks' that typically pervades every level of such relationships. Thus, it emphasizes the necessity of the trust element at every stage of business relationships (Urban, Sultan, & Qualls, 2000). Kim, Ferrin and Rao (2003) underscored this point when they posited that even for the first time buyer, with no prior direct experience with the seller, an “initial trust ... formed by indirect experiences such as reputation, recommendation, information quality of the seller's websites and so on” are necessary for the buyer to initiate the relationship (p.312). For this reason, trust is considered as the foundation of any business relationship (Berry, 1995; Kim, Ferrin and Rao, 2003; Urban, Sultan, & Qualls, 2000), and a major influencer of the consumer's cognitive, emotional and behavioral responses towards the firm and its offerings.

The importance ascribed to the trust variable has inspired researchers' interests in studying its effects in business. Most previous studies have conceived trust as a relationship variable (Kim, Ferrin and Rao, 2003), and have situated it within the social exchange theory (SET), which emphasizes that exchanges are mostly based on the expectations of trust and reciprocation between parties (Blau, 1964). Firms are naturally interested in building long-term profitable relationships; and the trust element is confirmed as major connection that leads to lasting relationships between customer and clients (Dwyer, Schurr, and Oh, 1987; Urban, Sultan, and Qualls, 2000), especially when customer expectations are fulfilled (Urban, Sultan, and Qualls, 2000). The nature of banking relationships further underscores the importance of trust in service relationships. Banks serve as custodians of peoples' funds, and a good measure of trust is required for customers to feel comfortable `safe-keeping' their hard earned funds with the banks. Zsofia (2009) further likned the importance of trust in banking relationships to the experiences of financial and economic crises that have affected customers' confidence in banks globally. Hence, S/he noted that trust has become a crusial element and differentiating factor, necessary to create a balanced and persistent relationship.

The current study proposes trust as mediator between distribution channel usage and customer satisfaction. This proposition is in line with the popular conception in the literature, where trust has been widely acknowledged as mediating variable across several disciplines, and most particularly in the service evaluation (Vlachos, et al., 2009; ). Further more, within the context of banking services, research findings have confirmed trust as a major influencer of customer satisfaction with banking services (Lee and Chung, 2009; Kim et al., 2003). More specifically, previous studies on customer evaluations and response regarding banks' distributional channels have confirmed trust as effective mediator between automated service quality and customers' commitment (Al-Hawari, 2011); and between service quality and loyalty in e-banking context (Chu, Lee and Chao, 2012). To this end, this study hypothesize as follows.

H5a: Trust mediates the relationship between human teller (HT) distributional channel strategy and customer satisfaction

H5b: Trust mediates the relationship between point of sales (POS) distributional channel strategy and customer satisfaction

H5c: Trust mediates the relationship between automated teller machine (ATM) distributional channel strategy and customer satisfaction

H5d: Trust mediates the relationship between mobile banking (MB) distributional channel strategy and customer satisfaction

H5e: Trust mediates the relationship between internet (INTNB) distributional channel strategy and customer satisfaction

1. Conceptual Model

In regard to prior theoretical review propose the conceptual model was form. As presented on Figure 1, the proposed model indicated that the latent exogenous variables affect directly the latent endogenous variable. Therefore, distribution strategy as target constructs is independent variable, which is represented by five dimensions (HM, POS, ATM, MB and INTNB). Trust is the mediating variable and customer satisfaction is the dependent variable.

Figure 1. Conceptual Framework Source: Authors conceptualization

Methodology of the study

This study adopted the survey research design, with personally administered questionnaire as research instrument. A total 450 retail bank customers comprising academic and non-academic staff from six tertiary institutions in North East Nigeria participated in the study. The survey questionnaire consisted of 34 items, measured on a five-point Likert scale, ranging from “strongly disagree (1), to “strongly agree (5)”. Respondents were required to tick appropriate answers corresponding to their opinions/evaluations. At the end of the survey, an initial set of 400 questionnaires were retrieved, and subjected to data screening procedures involving missing data analysis and treatment of outliers. Missing data were identified and treated using the recommended expectation maximization (EM) technique (Karanja et al., 2013). Similarly, extreme case scores (univariate outliers) that may have substantial negative impacts on the analysis were removed, and the Mahalanobis Distance (D) was used to determine and handle multivariate outlying cases (Hair et al., 2010). Subsequent preliminary analyses to ensure that all data treatment and multivariate assumptions were satisfied (i.e., normality, homoscedasticity, linearity, test for independence of the error terms, and multicollinearity) were conducted, leading to the exclusion of 48 responses. Therefore, a final set of 352 observations were used to detect R2 values around 0.005, assuming a significance level of 5%, and a statistical power of 95%. Moreover, 92 observations are required to detect a medium effect size of 0.15 with five predictors (as in the case of the current study), assuming the same significance level of 5% and statistical power of 0.80 (Nitzl, 2016).

Measures

Variables in the study were measured using multi-item, previously validated scales. Consumer perceptions of distributional channel strategies (independent variable) and trust (mediating variable) were measured by a set of questions inspired by the intensive study of the literature (Mohammad, 2020; Strydom & Fourie, 2018). To this end, Human teller and internet banking channels were measured using five items each, ATM-banking and consumer trust were measured through four items each; while POS and Mobile banking channels were measured through six and three items in the questionnaire respectively. Sample item for the distributional channels reads: “The internet banking services are available 24/7”. The dependent variable (customer satisfaction) was measured through seven items adapted from Mohammad (2015). Sample item reads “The distributional channels of my bank meet my expectations”. The questionnaire was pretested through a pilot test using 30 respondents to ensure that the items were reliable, after which adjustment was made to strengthen the instrument.

Common method Variance (CMV) or Bias

After primary data is collected, it is required to test the common methods variance or bias to insure whether orderly bias is influencing the data that have been collected (Podsakoff, MacKenzie, Lee, & Podsakoff 2003). CMV is a common bias in self-administered survey method. According to Bagozzi, Yi, and Phillips, (1991), CMV appears to be possible when correlation between constructs is more than 0.9. Table 1 shows the correlation between satisfaction and trust to be the highest with value of 0.6294. Therefore, it signifies no common method variance or bias from the data used.

Table 1. Inter-Construct Correlations

2. Partial least squares (PLS) Path Modeling Techniques

ADANCO 2.0.1 (Henseler, 2017) was employed to conduct the PLS-PM based analysis. The software is a variance-based analytical tool reputed for its `superiority' over covariance-based SEM counterparts in handling non-parametric and non-normally distributed data (Hair et al., 2014). The analysis follows the three-stage recommendations of Hair et al. (2017), comprising assessment of the model fit for the global (estimated) model, followed by evaluation of the measurement (outer) model - to determine the extent to which the indicator variables align with the theoretical concepts they are meant to represent. The third stage entails the assessment of the structural (inner) model to ascertain if the proposed relationships between constructs are in line with the proposed hypotheses. Details of these analyses are presented below.

Overall Model Fit and Measurement Model results

The major indicator of overall or global model fit in PLS is that the standardized root mean square residual (SRMR) should not be greater than 0.08 (Henseler, 2016). As indicated in Table 2, the SRMR for the current model is 0.067. Since this value is within the required cut-off point, the proposed model has satisfied the fitness criteria.

Upon confirming the model fitness, the study proceeded to assess the measurement model to ascertain the validity of the research items and constructs they represent. Results presented in table 2 reveals that all measurement items have factor loadings greater than 0.6 (Hulland, 1999), hence, confirming their reliability. Similarly, estimates from the relevant parameters for construct validity (i.e., Dijkstra-Henseler's rho (pA), Joreskog's rho (pc), Cronbach's alpha (a)) are all greater than 0.6, which is the minimum acceptable threshold for SEM analysis in social researches (Bagozzi and Yi, 1998). Likewise, the average variance extracted, AVE for all the main constructs are greater than 0.5, thus confirming the convergent validity of the constructs (Fornell and Larcker, 1981). Subsequently, discriminant validity was assessed to examine whether the study constructs are divergent from one another (Mohammad, 2019). Henseler (2015) recommends the HTMT test as the best for establishing discriminant validity, and Teo et al. (2008) recommends a threshold of not more than 0.9 to be used for measuring HTMT. Results in Table 2, indicates that all the HTMT values for the study constructs are within the acceptable threshold.

Figure 2. Reflective Measurement Model Source: Output from ADANCO software

Table 2 Model fit and Measurement Model Results

Structural Model

To assess the structural model (i.e., relationships among the endogenous latent variables), Hair et al. (2014) recommends checking for collinearity issues among constructs. Results of the collinearity analysis revealed that tolerance (VIF) values are generally below the threshold of 5 (tolerance higher than 0.20). Therefore, collinearity among the predictor constructs is not an issue in this structural model. Having certified the data as collinearity free, the analysis proceeded with checks on the model's predictive capability. For this, Chin et al. (2003) recommends assessing the R2 value, which indicates the sum of variance in the endogenous constructs explained by all of the exogenous constructs connected to it (Hair et al., 2017). Generally, Falk and Miller (1992) recommend that the variance explained, or R2s for endogenous variables should be greater than 0.1., while Hair et al. (2017) classify R2 values of 0.75, 0.50 or 0.25 for the endogenous constructs as corresponding to `considerable', `moderate', and `weak' predictive capabilities. The R2 values for the endogenous variable in this study are 0.381 and 0.497 for Trust and ST respectively, see figure 3 and table 3. These values are well above 0.1 recommended by Falk and Miller (1992), and also greater than 0.25 classified as weak by Hair et al. (2017). Thus, certifying the model as having valid predictive capability, and suitable for confirming hypotheses.

Source: Output from ADANCO software

Tests of hypotheses were conducted based on the recommended non-parametric bootstrap analysis with 5000 samples (Hair et al., 2017) to assess the path coefficients and t-statistics. Results presented in Table 3 reveals that only two of the five hypothesized direct effects are significant and positive. Human teller (P=0.2586; t=5.1908; p=0.0000) and INTNB (P=0.1896; t=3.6842; p=0.0002) are proven to be positively associated with customer satisfaction; thus providing support for hypotheses H1 and H5. The test statistics for hypotheses H2, H3, and H4 fell short of the acceptance criteria; hence, they were not supported.

Similarly, the non-parametric bootstrap procedure was applied in testing the significance of the indirect effects for the mediated paths in the model. This process is highly favored above the Baron and Kenny's (1986) causal step approach, due to its convenience and versatility in detecting mediation effects, particularly in SEM analysis (Nitzl et al., 2016; Zhao et al, 2010). Results from the bootstrap procedure reveal that four out of the five mediated paths are significant. Consumer trust significantly mediates the relationship between HT and satisfaction (P= 0.1291; t= 4.8790; p= 0.000); as well as the relationship between INTNB and satisfaction (P= 0.1368; t= 4.6614; p= 0.000). In both cases, the 95% bias corrected confidence intervals were above zero. Therefore, hypotheses H5a and H5e are supported. Furthermore, given that the direct effects remained significant even with the presence of the mediating effects, indicates partial mediation.

For the indirect path between MB on satisfaction via trust, estimates reveal significant positive and full mediation (P= 0.0373; t= 1.9339; p= 0.0266). This is indicated by the absence of significant effect on the direct path, and the 95% bias corrected confidence intervals, which were above zero. Thus, confirming hypothesis H5d. However, the results from the tests of significance on the indirect paths for POS^ST and ATM^ST through TRUST revealed insignificant effects as their associated t-values and p-values were below the threshold for acceptance. Also, the 95% bias corrected confidence intervals for the two paths all included a zero value. As such, the study found no support for hypotheses H5b and H5c.

Table 3. Structural Model Assessment and Hypotheses Testing

Discussion

The results of this study extend the understanding on distributional channel strategy and customer satisfaction, mediating role of Trust in a banking context. The study established that Human teller and internet banking significantly affect customer satisfaction, while POS, mobile banking and ATM do not exert significant direct effects on customer satisfaction. Positive influence of human teller on customer satisfaction is in line with the position of Seck and Phillipe (2003), and particularly Osuagwu, (2008) who stated that most Nigerian's rely on face to face banking hall transaction and have a strong confidence in it. Similarly, significant positive effects of internet banking on customer satisfaction is consistent with those of previous studies, i.e., Raza and Hanif (2013), Jun and cai (2001), Pikkarainen et al (2004), Herington and Weaven (2009), Road et al (2009) and Ankit (2011) who found that quality internet banking services leads to customer satisfaction across several contexts.

Findings from this study have not confirmed the direct effects of POS, ATM and Mobile banking on customer satisfaction. Although, these findings are largely inconsistent with those of previous findings (i.e., Sampio et al., 2017; Seck and Phillipe, 2003), they provide indications that customers may not be so happy with the performance and quality of services received through these distributional channels. The reasons for this lack of satisfaction may not be unrelated to what previous researchers have pointed as problems associated with ATMs and other automated service channels. For instance, Murdock and Franz highlighted issues of the difficulty of usage and risk, while Sanda and Arhin (2011) raised issues of customer complaints of high bank charges, unreliable service, and faulty machines. Specific to this study's context, Alaba (2011) characterize ATMs in Nigeria with fraud and other unwholesome practices. This phenomenon is basically same with other automated service delivery channels, and has become a constant news item on the traditional and social media.

Findings from the mediation analyses also provide support for the importance of trust in influencing customer behavior. Results indicated that trust is partial mediator between two distributional channels (human teller and internet banking) and customer satisfaction, while it fully mediates that between mobile banking and customer satisfaction. These findings are consistent with Rizan, Warokka & Listyawati (2014), who found customers' trust as important mediator of the effect of relationship banking tactics on loyalty towards the banking services. More specifically, the full mediation effect of trust between mobile banking and customer satisfaction conforms to Al- Hawari (2011), who observes that most factors of automated banking channels only have indirect effect on customer commitment through customer trust and delight.

Implications, conclusions and study limitations

Findings from this study have significant implications for banks managers. First and foremost, the study findings have provided additional confirmation to the importance of the human teller in bank services, particularly in the study context. As indicated in the results, the channel maintained significant positive effects on customer satisfaction, both directly and indirectly, with trust only exerting a partial mediating effect. This finding underscores the irreplaceability of the human element in banking service relationships (Al-Hawari et al., 2009). Bank managers will need to implement continuous training and orientation of employees towards effective service delivery and relationship building.

Similarly, the results further emphasize the need for Nigerian banks to pay more attention to distributional channel strategies, particularly those aimed at self-service (ATM, POS, Mobile banking). The fact that these channels do not directly influence customer satisfaction point to the need for bank managers to device innovative means of improving the quality of services from these channels. Additional efforts would also be required to communicate these measures to customers in

order to overcome the learning difficulties that likely impact on customers' perception and attitudes towards these channels (Al-Hawari, 2011).

Furthermore, the study findings also highlight the imperative for bank managers to win the trust of customers towards the banking channels and their services in general. As observed by Kim et al. (2003), trust is a crucial factor in banking relationships, and it plays important role at every stage of banking relationship. Consumers need to develop some measure of trust before even deciding to use any of the distributional channels, and this trust is further strengthened when the quality and performance levels of the channels meet or exceed customer expectations. Bank managers should consider instituting security and safety measures, as well as customer feedback mechanisms to help alleviate all negative issues impacting on customers' trust for the channels.

The study acknowledges the following limitations. First, the study sampled customers from the north-east region of Nigeria. Although, the respondents are considered cosmopolitan, extended studies covering more regions and other territories may provide additional insights and improve generalizability of findings. Secondly, the study focused on retails customers only. Future studies may examine the relationships from corporate customers' viewpoints to account for differences in usage patterns. Finally, even though this study has assessed most of the distributional channel strategy in retail banking. It will be appropriate for future studies to approach it from Omnichannel strategy perspective (Mainardes, Rosa and Nossa, 2020).

References

1. Adewoye, J.O. (2013). Impact of Mobile Banking on Service Delivery in the Nigerian Commercial Banks. International Review of Management and Business Research, 2, (2), 333343.

2. Adeyemi, O.A., Ola, O.S. &Oyewole, F.A (2014). Internet Banking Functionality in Nigeria and Outcomes of Customer Satisfaction: An Empirical Investigation. International Journal of Academic Research in Business and Social Sciences,Vol. 4, No.8, pp.195-205.

3. Amin, M.(2016). Internet banking service quality and its implication on e-customer satisfaction and e-customer loyalty. International Journal of Bank Marketing, 34, (3).280-306 Ankit, S. (2011). Factors influencing online banking customer satisfaction and their importance in improving overall retention levels: an Indian banking perspective. Information and Knowledge Management, 1(1), 45-55.

4. Alaba, S.O. (2011). The adoption of automated teller machine in Ille-Ife, Osun state: factors, rate, problems and challenges. Journal of Basic and Applied Scientific Research, 1(9), 1089-1093. Al-Hawari, M., Ward, T, and Newby, L. (2009) The relationship between service quality and retention within the automated and traditional contexts of retail banking, Journal of Service Management 20 (4), 455-472.

5. Al-Hawari, M.A., (2011),"Automated service quality as a predictor of customers' commitment", Asia Pacific Journal of Marketing and Logistics, 23 (3), 346 - 366 Anderson, E.W. & Fornell, C. (1994). New Directions in Theory and Practice. Thousand Oaks, Ca: Sage Publications

6. Anderson, E.W., Fornell, C. & Lehmann, D.R. (1994). Customer satisfaction, market share, and profitability: findings from Sweden. Journal of Marketing, 58 (3), 53-66.

7. Bagozzi, R. and Yi,Y. (1988) On the Evaluation of Structural Equation Models. Journal of the Academy of Marketing Sciences, 16, 74-94.

8. Bagozzi, R.P., Yi,Y. & Phillips, L.W. (1991). Assessing construct validity in organizational research. Administrative Science Quarterly, 36(3), 421-458.

9. Barsky, J.D. (1992), Customer Satisfaction in the Hotel Industry: Meaning and Measurement Journal of Hospitality & Tourism Research 1992 16: 51-73.

10. Baron, R.M. & Kenny, D.A. (1986).The moderator-mediator variable distinction in social psychological research: Conceptual, strategic, and statistical considerations. Journal of Personality and Social Psychology, 51(6), 1173-1182.

11. Bello, K.B., Alkali, A.U. (2020), “Customer rights knowledge in technology-enhanced banking environment: empirical evidence from bank customers in Yola”, ATBU Journal of Accounting and Finance, 1 (1), 92-102.

12. Bello, K.B, Danjuma, I, and Udo-Imeh Phillip. (2014) “Innovative ICT-based Services: Drivers of transition to...