Word of mouth communication: A mediator of relationship marketing and customer loyalty

Abstract

A cross-sectional and quantitative study design, with the aim of explaining the relationship between relationship marketing and customer loyalty, and the mediating role of word of mouth in this relationship was undertaken. A sample of 384 was determined from a population of the mobile telecommunication users based on Krejcie & Morgan sampling framework. The study utilized 384 questionnaires. The authors conducted confirmatory factor, correlation, regression, mediation and SEM for analysis, interpretation and results.

While trust, relationship satisfaction and reciprocity components of relationship marketing have been found to be significant predictors of customer loyalty in other studies, our study findings reveal contrasting results. This is a unique finding in our study. However, in line with earlier studies, our findings indicate a positive significant relationship between relationship marketing components of communication, commitment and customer loyalty. The study also finds a significant positive relationship between the relationship marketing components of communication and commitment and word of mouth and a significant positive relationship between word of mouth and customer loyalty. Telecommunication companies should pay attention to relationship encounters that build commitment, should develop targeted communication channels which build positive word of mouth communication. This will ultimately create loyal customers for mobile telecommunication companies.

Keywords:

• Uganda’s ICT sector
• mobile telecommunications
• word of mouth
• relationship marketing

PUBLIC INTEREST STATEMENT

The 21st-century business environment is highly competitive and volatile. As a coping strategy, organizations have resorted to employee cuts, leaner organizations, and heavy investment in sales promotions. Unfortunately, these strategies have not helped today’s organizations in meeting the desired market and profitability needs. This has made business survival very difficult. Scholars and practitioners have been looking for affordable strategies that can make organizations survive the competitive business environment. They agree that creating loyal customers is one of the most effective strategies of remaining ahead of the competition, become profitable and survive the volatile environment. As a result, companies are continuously looking for innovative ways to engage their customers and keep them for a long period of time. In this paper, we indicate that companies can create loyal customers through clear communication, trust and developing a positive word of mouth.

1. Introduction

Ever since Uganda gained its independence in 1962, the country experienced poor governance and leadership, political and economic instability which put the economy to its lowest from 1962 to 1986. Most of the economic and social infrastructures such as telecommunications, trade, education, health and civil service had died until a new government akin to the economic and political development of the country took over power in 1986 (Sejjaaka & Kyeyune, 2013). The new government’s liberalization policy saw a number of mobile telecommunication companies enter the Ugandan market. Currently, the total number of mobile telecom users in Uganda stands at 24.8 million, a 70.9% penetration rate (Uganda Communication Commission, 2018). The technological changes in the telecommunication sector have created a major shift of the customers from the use of landlines to increased use of mobile phones for business, social and political agendas. The boom in Uganda’s mobile market as a result of the clear policy of liberalization has brought in a number of market players, thereby creating a competitive business environment in the telecom sector. This has significantly lowered market prices, increased consumer knowledge and subsequently lowered the average revenue per user. The companies that operate in the Ugandan market are MTN Uganda, Airtel Uganda, Uganda Telecom, Africell, Vodafone, Smile Telecom, Smart Telecom, Sure Telecom and K2 telecom. While the mobile telecom companies cover both urban and rural areas, they are most prominent in the urban areas. Some of the telecom companies like Smile, Smart, Sure and K2 have a smaller coverage, but all the companies at least cover Kampala. Kampala is the capital, industrial and business hub hence the most affluent city in Uganda.

The mobile telecommunication investors come in to meet customer needs but also to earn a profit. The profitability of the telecom companies depends on their customer base, and yet the recorded number of loyal customers out of the total subscribers for each of the telecom companies is low (UCC, 2014). While the numbers of subscribers on each individual network have been growing for the past five years, usage period for the majority of subscribers does not go beyond 12 months. Given that the telecom industry in Uganda is very dynamic, vibrant and competitive, the telecom customers are driven by short term rewards and incentives that come from different promotional offers. The customers are equally smarter and more informed with many choices with little or no switching costs. This has made customers switch from one telecom to another looking for better offerings at lower prices. The moment the promotional offers end, subscribers switch to another network with better rewards and promotions. A number of mobile telecom users in Uganda own more than one telephone lines from the different mobile telecommunication companies and keep switching from one network to another. This phenomenon raises the question as to whether the customer loyalty in the mobile telecommunication sector in Uganda is real or spurious. Companies that are able to retain their current customers will succeed in long-term. This phenomenon has pushed the mobile telecom companies to shift from short-term and transactions focus to longer-term and relational focus, which drives a positive word of mouth and creates customer loyalty. Relationship management and positive word of mouth (WOM) are low-cost strategies that can yield a strong loyal customer base (Barreda, Bilgihan, Nusair, & Okumus, 2015).

In a market where the customer acquisition rate is slowing, increasing customer loyalty remains the biggest challenge faced by the telecom companies. Ugandan being a collective society where collective conscience, common beliefs and feelings are cherished in interactions and relationship building, customers tend to be loyal to a service if others in society recommend it. The recommendation constitutes a positive word of mouth. Although academic research on relationship marketing has received considerable attention, some of the claims of its influence on customer loyalty seem to be inflated. We have not found sufficient evidence of what mediates the relationship between the two constructs. Most studies have tended to focus on measurement issues (Alan and Basu, 1994) and direct effects of relationship marketing on customer loyalty. It is further true that scholars and practitioners have relied on evidence from studies conducted on fast moving goods and more so in developed economy perspective to make conclusions on services in least developed economies, and yet relationship marketing and customer loyalty are context-specific (Mitchell et al., 2018;Tabrani, Amin, & Nizam, 2018). Evidence of what drives customer loyalty in a least developed economy context remains scanty.

This study is an attempt to bridge the gap in the empirical and theoretical literature on relationship marketing and customer loyalty in the service sector in a least developed economy perspective. Most importantly, the study presents word of mouth as a mediator between relationship marketing and customer loyalty. To this extent, the study contributes to the knowledge of how customer loyalty comes to be in a service sector in a least developed economy, which can be the starting point to develop a framework or set of criteria to be taken into account to explain service loyalty in a least developed economy context. Since the number of social network users in Uganda is still small, the study was confined to the traditional WOM as opposed to eWOM.

2. Literature review and hypotheses

Customer loyalty has become an important variable in marketing and particularly in the field of customer relationship management. Loyalty is likely to lead to positive attitudes and behaviors such as repeat patronage and purchases and positive recommendations which may influence other actual or potential customers. Generally, service industries are characterized by high competition and relatively low switching costs. As competition intensifies, many firms are developing or improving their loyalty programs to deter customers from defecting to their competitors (Ho & Wu, 1999). In particular, a loyal customer base will generate more predictable sales, steady cash flow and an improved profit stream while counteracting any competitive agitation (Aaker, 1991). For the mobile telecom companies whose competitive landscape has become volatile, building loyalty of customers will provide business sustainability. While customer loyalty can be attained in various ways, most marketing scholars emphasize the influence of relationship marketing as an important strategic tool for attaining customer loyalty.

The concepts of relationship marketing and customer loyalty have been studied and widely accepted as important aspects in successful business operations. A firm that has a relationship marketing orientation has higher chances to succeed in business through the creation of positive word of mouth, and customer loyalty. According to Arnett and Badrinarayanan (2005), different marketing studies suggest various factors that make up relationship marketing. However, Alrubaiee and Al-Nazer (2010) mention the three consistently identified important factors as trust, relationship commitment and communication. The other factors include relationship satisfaction and reciprocity among others. All these factors have been argued to drive customer loyalty.

For relationship marketing to drive customer loyalty, there should be interactions between the organization and the customers. As businesses interact with their customers, they transact and exchange value as a way of creating utility and mutual satisfaction. Scholars like Mizra (2018) and Yang (2018) have considered relationship marketing (and customer loyalty) to be context specific, dynamic across contexts. Thus, the interaction theory (Gallagher, 2001) that focuses on both behaviors and environmental contexts becomes an appropriate anchor for this study, especially where word of mouth comes into play. The theory is also appropriate to the Ugandan context since it does not assume rationality, and is dynamic. When deciding on mobile telecom services, customers largely rely on recommendations by others in society. It is a common phenomenon to find a Ugandan mobile telecom customer subscribing to more than one network or owning mobile telephone lines of all the existing companies, including the ones that they do not use frequently. Their purchase decision is rather whimsy, sometimes based on impulse.

The interaction theory places much emphasis on communication and mutual influence as well as commitment to the interactions between parties (Kutschler, 1985), which are dynamic practices and key elements in relationship marketing. Customer loyalty and success of businesses in collectivist societies like Uganda is largely driven by successful relationships. Since people tend to work together, the needs and goals of a group are superordinate to the individual needs. As such, collective conscience, common beliefs and feelings develop in the society. Relationships with other members of the group and the interconnectedness between people play a central role in each person’s identity; hence, people tend to consult each other in their purchase decision. In the process, they encounter the word of mouth from those they consult from, which becomes their basis for building service loyalty. The sections that follow present the theoretical relationship that exists between the relationship marketing components and customer loyalty.

2.1. Trust and customer loyalty

Hunt and Morgan (2005) contend that trust exists when one party has confidence in the partner’s reliability and integrity. Trust is the belief that a partner’s word or promise is reliable, and a party will fulfill his/her obligations in the relationship (Schurr & Ozanne, 1985). Trust is an important construct in relational exchange because relationships characterized by trust are so highly valued that parties will desire to commit themselves to such relationships (Ting, 2004) and are likely to stay loyal. Trust attracts new customers on each individual network and helps in retaining customers for a longer period. To support this notion, trust has been posited as a major determinant of customer loyalty. Thus, creating trust in customers’ minds is of great importance for companies to achieve customer loyalty (Ball, Coelho, & Machás, 2004).

Trust plays a significant role in interactions, it actually drives word of mouth and customer loyalty. Developing, improving and practicing trust are considered important aspects of investing in a dyadic and affective relationship between the parties in the relationship. Increased trust between the customer and the business is cited as critical for relationship success (Kara, Lonial, Tarim, & Zaim, 2005). In collectivist societies like Uganda, trust is normally the basis for sharing and accepting information and drives competition. In interactions, the relational exchange is characterized by relationships based on trust. Customers prefer to continue interacting with organizations and services they trust. Trust is thus construed an important factor that determines relationship commitment and customer loyalty. We, therefore, hypothesize as follows:

H1: There is a positive relationship between trust and customer loyalty.

2.2. Commitment and customer loyalty

While some scholars contend that there is no difference between commitment and loyalty (Henning and Klee, 1997) the majority of researchers suggest that these two constructs are related but different and that commitment is vital in building successful relationships, which ultimately leads to loyalty (Berry, Carbone, & Haeckel, 2002). Indeed, affective commitment in interactions creates emotional attachment which results in loyalty. Although loyalty simply was considered as repeat purchase at the beginning, some researchers realized that repurchase alone is not sufficient evidence of loyalty (Newman & Patel, 2004). Loyalty should be conceived as the commitment to the product/service/firm stimulated by certain positive attitudes. Commitment indicates the motivation to maintain a relationship. When customers are committed, their turnover decreases (Gounaris, 2005). Commitment has therefore been considered as one of the key factors affecting customer loyalty (Rauyruen & Miller, 2007). Positive intentions to maintain and strengthen a business relationship and thus stay loyal are created by commitment. A number of relationship marketing studies have shown that these two constructs of commitment and loyalty seem to be crucial in influencing one another (Evanschitzky, Iyer, Plassmann, Niessing, & Meffert, 2006; Fullerton, 2005). Relationships are built on the foundation of mutual commitment, and the commitment level has been found to be the strongest predictor of the voluntary decision to pursue and stay in a relationship (Ibrahim & Najjar, 2008).

The Ugandan mobile telecommunication companies have invested a lot in customer commitment so that they get customers to develop a strong attachment to or engagement with, their companies and services. Events like marathons and client parties have become a popular thing. These companies have strived to follow through their promises, for instance when they guarantee quick, efficient and reliable network and services they try to fulfill. Some of the companies have introduced rewards points in loyalty plans like bonus airtime, bonus data, free calls and the like to boost customer commitment. Each mobile telecom company has come up with at least a unique offer(s) for which customers are convinced that no other offer or company could do a better job of meeting that need than them. Others have tried to align their values with their customer values to make customers more committed. Following this line of argument, we hypothesize as follows:

H2: There is a positive relationship between commitment and customer loyalty.

2.3. Communication and customer loyalty

Communication in relationship marketing involves all the formal and informal exchanges that result into meaningful and timely exchange of information between buyer and seller (Ranjbarian & Berari, 2009). Providing timely, accurate and reliable information to customers, including information on new services and promises is considered key in building customer relationships. Open communication channels between the organization and customers will ensure smooth information exchanges between both parties. Therefore, communication between the company and the customers should be solid and predictable if both sides are to be aware of the mutual benefits of the relationship. This way, the customers and the company will be willing to commit to long term relationships. Communication has also been argued to increase the level of trust between partners, as well as improving the partners’ ability to align their expectations and perceptions (Alrubaiee & Al-Nazer, 2010). Appropriate communication from service provider can be helpful, positive, useful and pleasant, and has important implications for customer behavior. To ensure effective communication with their customers, the mobile telecommunication companies have come up with help lines or codes that customers can dial to speak to customer care. The dial codes for customer care have further been designed to take care of different languages. The customer is given the opportunity to choose a language of their choice by dialing a number that corresponds to that language prompted by the voice machine. There are also customer relationship managers who ensure that communications flow smoothly between the customers and the organization. Since Uganda is a multi-lingual country, mobile telecommunication companies have deliberately recruitment call center staffs from different parts of the country to cater to the different communication needs of their customers. While recruiting customer service personnel, these companies consider communication skills as a key requirement. Indeed, the ability to speak at least three to four languages (including English) is a parameter that these companies consider during recruitment of call center staff. All this is done to ensure effective communication with their clients. Following the above argument, we hypothesize as follows:

H3: There is a positive relationship between Communication and customer loyalty.

2.4. Reciprocity and customer loyalty

Reciprocity is regarded as one of the key ingredients that can cement a lasting long-term relationship between an organization and the customer (Baggozi, 1995; Fournier, Dobscha, & Mick, 1998). Scholars such as Schultz and Bailey (2000) and Morais, Dorsch, and Backman (2004) used reciprocity as a theoretical base for building their respective theories of brand loyalty and customer loyalty. To maintain social harmony, there are reciprocal obligations expected of all the parties in an interaction, especially in a service sector like the mobile telecom services where revenues are greatly dependent on how long a customer stays with the company and the time span that the customer uses the service. Customers want to see that the resources they invest in the seller are being reciprocated in an equitable fashion. This creates satisfaction with the transaction, generates and supports an ongoing relation between the exchange partners (Dorsch & Carlson, 1996). In a collectivist society like Uganda, and drawing from the interaction theory, reciprocity in an interaction is crucial. This is so because in such societies relational transactions are more important than transactional relationships; thus, reciprocity improves relationships which makes customer loyal. Willingness to engage in a reciprocal and mutually beneficial relationship between customers and service providers increases the chances of creating customer loyalty. Going by this line of argument and debate, we, therefore, hypothesize that;

H4: There is a positive relationship between reciprocity and customer loyalty.

2.5. Relationship satisfaction and customer loyalty

In relationship marketing, it is important that the customer is satisfied with the relationship. Alrubaiee and Al-Nazer (2010) define relationship satisfaction as a customer’s affective or emotional state toward a relationship. A high-quality relationship where there are trust, commitment and mutual communication will be satisfying to the customer and a customer who is satisfied with a relationship is likely to talk good about the organization, recommend the services to others and be more loyal. Relationship satisfaction is influenced by the customer’s expectations of the relationship and the perceived quality of the relationship (Hu, Kandampully, & Juwaheer, 2009). Satisfaction can, therefore, be taken to mean a customer’s emotional response when evaluating the discrepancy between expectations regarding the relationship and the perception of the actual relationship. If a customer feels that he has a satisfying relationship with the business, he may perceive the business to have a high level of service and will ultimately become loyal (Rootman, 2006). It is therefore natural to hear mobile telecommunication users in Uganda sharing their good and bad experiences during their interactions with the company staff or their bad encounters with the company’s services. Mobile telecommunication companies have gone ahead to send birthday messages to their customers as a way of cementing their relationships with the customers. The events such as marathons, client parties and sponsorships of shows and celebrities by mobile telecom companies are also meant to cement their relationships with the customer. Following this logic, we, therefore, hypothesize as follows:

H5: There is a positive relationship between relationship satisfaction and customer loyalty.

2.6. Customer Relationship Marketing (CRM) and Word of Mouth (WOM)

It is expected that strong relationships between customers and employees foster positive word of mouth among customers; thus, many companies focus on establishing a relationship with customers to foster positive word of mouth. Word-of-mouth is defined as any positive or negative statement made by customers experiences about a product or company (Henning-Thurau & Klee, 1997). Anderson and Sullivan (1993) described WOM as vivid and novel, adding that, while positive WOM refers to pleasant experiences, negative WOM includes product denigration, unpleasant experiences, rumor and private complaining. The usefulness of a product/service, its major advantages and features are easily spread through word of mouth by customers to others. Customers who have trust in, and are committed to the relationships with the company are expected to talk positively about the company and its products (Kumar, Petersen and Leon, 2010). For example, in Uganda, it has been a practice for customers to talk positive and recommend to others particular mobile telecommunication services based on their past experiences with these services in terms of their trust and commitment to the services. Customers will also promote a positive word of mouth about telecom companies that keep the communication lines open and effective—for example, communicating frequently to customers when the services breakdown and rectify problems quickly. In the same vain, customers will talk negative and even dissuade others from some mobile telecom companies based on their past experience. Given the low penetration of social media in Uganda, this study considers the traditional word of mouth as opposed to the eWOM. Following this debate, we, therefore, hypothesize as follows:

H6: Customer relationship marketing positively relates to word of mouth.

2.7. Word of mouth and customer loyalty

To attract and retain customers, word of mouth is arguably an important factor. What customers talk about a company will shape customers’ attitudes about an organization and its’ services, manifested by intentions and behaviors of re-patronization and recommendation. While WOM can be defined as Oral or written recommendation by a satisfied customer to the prospective customers of a good or service, customer loyalty is the Likelihood of previous customers to continue to buy from a specific organization (Businessdictionary.com). Customer loyalty is an attitude about an organization and its’ services that is manifested by intentions and behaviors of re-patronization and repeat purchase (Oliver, 1999).

Word-of-Mouth communication is a good way for enterprises to catch the attention of new customers as these will largely rely on what the existing customers say about a company and its services. The new customer will ultimately get glued to the company, its services and product leading to loyalty. Word of mouth communication has been recognized as a particularly valuable vehicle for promoting a firm’s products and services and indeed driving customer loyalty. Positive WOM recommendation removes doubt, creates customer excitement and may even cause a switching barrier in that it prevents the customers from breaking the relationship. It boosts customer confidence in the company and its services and makes customers feel they made the right choice. Indeed, Fornell (1992) thinks that loyalty is the function of satisfaction, switching barriers and word of mouth. The more companies motivate customers to speak good about their offers, the more they will create loyalty among their customer base. Customers no longer trust marketing messages from companies but would like to listen to experiences of real customers. It is these experiences that will attract and keep them loyal to the company. By speaking positively about the company and its offers, customers can promote new and repeat purchase. Positive WOM can indeed be a powerful marketing tool in creating customer loyalty (Ferguson & Paulin, 2006). Thus, positive word of mouth is likely to increase the customers’ trust of an organization and its services, which will increase customer loyalty. Although word of mouth communication can be very influential in any purchase decision, it is particularly important for services in aiding repeat purchases and ensuring customer loyalty. Therefore, people who spread positive word of mouth to others about particular goods/service, and those that receive that positive word of mouth tends to exercise some loyalty in their consumption patterns. In a country like Uganda where the culture encourages collectivism as opposed to individualism, people tend to work together. Thus, customers are likely to build collective conscience, common beliefs and feelings. Customers will, therefore, listen to and believe in the views of those that they take as their opinion leaders. When such influential customers speak negatively about the services, the rest will shun away from the services and the company will lose customers. Given that mobile telecom services are technology driven, customers will tend to listen and act to the word of mouth of those that they perceive to be more knowledgeable. We, therefore, hypothesize as follows:

H7: There is a positive relationship between word of mouth communication and customer loyalty.

Since the reviewed literature suggests a positive relationship between relationship marketing and word of mouth, and a positive relationship between word of mouth and customer loyalty, it follows that the indirect relationship between relationship marketing and customer word of mouth could be driven through word of mouth. By inference, we further hypothesize that

H8: Word of Mouth mediates the relationship between relationship marketing and customer loyalty.

Conceptual framework:

Following the above literature debate, we develop the conceptual framework in Figure A1. In this conceptual framework, we argue that there is a relationship between the components of relationship marketing (that is trust, commitment, communication, and reciprocity and relationship satisfaction) and customer loyalty. We also argue that there is a relationship between relationship marketing and word of mouth, word of mouth and customer loyalty. We further argue that word of mouth mediates the relationship between relationship marketing and customer loyalty.

Figure A1. Conceptual framework.

3. Methodology

This study adopted a cross-sectional and quantitative study design, with the aim of explaining the relationship between relationship marketing and customer loyalty, and the mediating effect of word of mouth in this relationship. The total number of mobile telecom users in Uganda stands at 24.8 million, a 70.9% penetration rate. (UCC, 2018). These include all mobile phone users registered and operating mobile phone lines with all the telecom companies operating in Uganda. From this population, we drew a sample of 384 respondents based on Krejcie and Morgan (1970) sample size determination criteria. Evidence from previous customer surveys conducted in Uganda indicates an average nonresponse rate of 30% for well-planned and administered surveys. Thus, the sample was increased to 499 (30%) to cater for the non-response. The study utilized the 384 usable questionnaires collected, representing a response rate of 77%. Since the mobile telecommunication phone users are spread all over the country and there was no accurate record of the users, it was not easy to establish the sampling frame from which we would draw a random sample. For this reason, the convenience sampling method was used to select the respondents. Data was collected from the telecommunication mobile phone users using a self-administered questionnaire. The survey was well planned and the research assistants were trained to avoid any bias. The questionnaire was anonymous and the answer options were not leading.

Measures for the variables were obtained and adapted from studies undertaken by previous scholars and anchored on a five-point likert scale to capture the opinions and attitudes of respondents. Customer relationship marketing was measured using an instrument developed by Sharma and Patterson (2000) which measures relationship marketing on commitment, communication, trust, reciprocity and satisfaction. Word of Mouth was measured using Lang’s (2009) tool with two factors, i.e., WOM activity (how much WOM one is likely to engage in) and WOM Valence (how positive or negative one’s WOM is likely to be). Customer loyalty was measured using Anderson and Sullivan (1993) instrument with dimensions of repeat purchase, referrals, and retention.

The scales were tested for reliability using the Cronbach Alpha Coefficient. All the alpha coefficients of the individual constructs were above the recommended cut-offs and were considered good according to Nunnally (1978) (see Table A1).

For construct validity, we carefully selected the already used and validated instruments with items that measure the distinct concepts. We also ensured theoretical operationalization by carefully reviewing related literature. To establish whether the instrument measured what it intended to measure, we further tested for convergent and discriminant validity, especially for word of mouth and customer loyalty constructs, using methods adopted by Ritter, Wilkinson, and Johnston (2002), Gaski (1984), Field (2005) and Hair, Anderson, Tatham, and Black (2010). Going by Ritter et al. (2002), the correlations between each of the measurement items for WOM and the total/overall WOM rating as well as the correlations between the individual measurement items for customer loyalty and overall customer loyalty rating were highly significant. This indicated that the instrument had convergence or is a good predictor of what it intended to measure.

The correlations between WOM measurement items and the total customer loyalty rating and the correlations between customer loyalty measurement items and total WOM rating were largely insignificant suggesting that the measurement scales ably discriminate between the two measures of WOM and customer loyalty. The few that were significant had a correlation of less than 0.3 which is still considered a weak correlation with the scale (Cohen, 1988; Field, 2005), suggesting that the measurement scales ably discriminate between the two measures of WOM and customer loyalty and that the two constructs are distinct. The results are indicated in Table A2.

We also found that no correlation coefficient between the two constructs exceeded the constructs’ reliability as shown in Tables A1 and A4, providing additional evidence of discriminant validity of the constructs (Gaski, 1984). Gaski (1984) suggests that if the correlations between two composite constructs are not higher than their respective reliability estimates; then, discriminant validity exists. The discriminant validity results, therefore, indicate heterogeneity between the two constructs that is WOM and customer loyalty, further supporting the fact that the two constructs are distinct.

To establish the quality of the scales, we conducted explanatory factor analysis (EFA) using SPSS software to establish the factor loadings. EFA was carried out on the multiple items that measured each variable to check if they loaded together on the same factor. Hair, Anderson, Tatham, and Black(1998) contend that when the item value loads on only one factor with a factor loading larger than or equal to 0.5; then, it is accepted. Table A3 shows the number of items that loaded on each related factor with the value ranges >0.5, as well as the total variance explained.

With the help of the SPSS data editor, data were cleaned and made ready for analysis through sorting and editing. Missing value analysis was done so as to establish the extent of the missing data. Following the recommendation by Tabachnick and Fidell (2001), we first examined the amount of the missing data, and then the pattern of the missing data. The frequency runs indicated that 8% of the data was missing overall. Further analysis of the missing data with the help of the expectation maximization (EM) method with Little’s missing completely at random (MCAR) test revealed that the data were missing completely at random. Since the missing data did not relate to specific items but was randomly distributed among the questionnaires, we were comfortable that the missing data would not lead to biased estimates of results (Pallant, 2005). We, therefore, went ahead to estimate and fill the missing data using the mean score as recommended by Mertler and Vannata (2002).

The nature and strength of the relationships between the variables were tested using the zero-order bivariate correlation analysis, while regression analysis was conducted to determine the prediction level of the models. The regression analysis also helped determine which predictor variables contributed more or were more important/significant in predicting the dependent variable. Specifically, regression analysis was used to determine the variance in customer loyalty explained by relationship marketing and word of mouth.

The hierarchical regression approach, a more elaborative approach which enters the variables according to their order of importance (Field, 2006) was also conducted. The hierarchical regression procedure indicated precisely what happened to the model as different variables were introduced in the model, which provided the researcher with the opportunity to systematically follow the contribution of the independent constructs to the explanatory power of the model.

The mediator effect was tested using multiple regression analysis, following a method recommended by Baron and Kenny (1986), later updated by Kenny, Kashy, and Bolger (1998). For a mediator effect to be tested, there should be a significant association between an independent variable and the outcome variable, that is there should be a relationship to mediate. Baron and Kenny (1986) suggested an estimation of a series of regression models to test for mediation. To establish whether there was partial or full mediation, the researcher derived the Sobel test. The mediation test results were diagrammatically presented in a Medgraph.

4. Results and discussion

Correlation analysis was performed to determine the strengths and direction of relationships between the independent and dependent variables of the study variables (Hair, Black, Babin, Anderson, & Tatham, 2006).

4.1. Correlation analysis results

The results of the correlation are presented in Table A4.

4.2. Customer relationship marketing and customer loyalty

The correlation results revealed positive and significant relationships between customer relationship marketing components and customer loyalty as follows: trust and customer loyalty (r = .414, p-value < 0.01), commitment and customer loyalty (r = .547, p-value < 0.01), communication and customer loyalty (r = .474, p-value < 0.01), reciprocity and customer loyalty (r = .454, p-value < 0.01), relationship satisfaction and customer loyalty (r = .469, p-value < 0.01), supporting hypotheses H1, H2, H3, H4, H5. This means that a positive change in customer relationship marketing components results in a positive change in customer loyalty. A company that builds strong relationships with its customers is likely to have the customers committed to that company and its products/services. This finding is consistent with Castaneda (2011) who argued that without customers trusting and being committed to the service provider, loyalty will never be achieved. Roland and Werner (2010) also agree that service providers can only survive in volatile conditions once they create trust, commitment and effective communications amongst its customers for they will be assured of long-lasting business relationships. When a customer trusts a business or brand, that customer is willing to form a positive buying intention towards the business. When service providers act in a way that builds customer trust, the perceived risk with the service provider is reduced, thus enabling the customer to make confident predictions about the service provider’s future dealings. Our results are largely in line with earlier scholars that have found trust, commitment, relationship satisfaction, communication and reciprocity components of customer relationship marketing to have an influence on customer loyalty (Aydin, Özer, & Arasil, 2005; Bettencourt, 1997; Du Plessis, Jooste, & Strydom, 2005; Luarn & Lin, 2003).

4.3. Customer relationship marketing and word of mouth

The results indicate that all the relationship marketing components of trust, commitment, communication, reciprocity and satisfaction positively and significantly correlated with word of mouth (r = .333, p-value < 0.01; r = .448, p-value < 0.01; r = .399, p-value < 0.01; r = .523, p-value < 0.01; r = .357, p-value < 0.01, respectively) supporting hypothesis H6. This means that a positive change in the customer relationship marketing components results in a positive change in word of mouth. A company that builds strong customer relationships is likely to have its customers talk good about that company and its products/services. This finding is consistent with Babin, Lee, Kim, and Griffin (2005) who found that relationship marketing components of trust, communication, commitment and satisfaction positively influenced word of mouth amongst the customers.

4.4. Word of mouth and customer loyalty

There was also a significant and positive correlation between word of mouth and customer loyalty (r = .447, p-value < 0.01) supporting hypothesis H7. This means that a positive change in word of mouth will result in a positive change in customer loyalty. When customers speak positively about a company and its services/products, then customers are likely to be loyal to that organization. This finding is in line with Osmonbekov and Czaplewski (2006) who argued that a significant result of word-of-mouth is the customer purchase, repurchase and loyalty.

4.5. Regression analysis results

In order to determine the effect of independent variables on customer loyalty, we performed regression analysis. Specifically, the hierarchical regression approach was used to determine the effect of each individual construct of relationship marketing and word of mouth on customer loyalty. Each construct was entered one at a time, creating six (6) models to explain the variation in customer loyalty. The results of hierarchical regression are shown in Table A5.

In model 1 we entered trust, and the results indicate that trust is a significant predictor of customer loyalty. Trust accounts for 17.1% of the variation in customer loyalty. Model 1 is statistically significant (sig. = 0.000, p < 0.01, F = 79.077).

In model 2 commitment was added and the results indicate that commitment is a significant predictor of customer loyalty. The R square increased to 29.9%. Thus, commitment accounts for 12.8% increase in the variation in customer loyalty. Model 2 is statistically significant (sig. = 0.000, p < 0.01, F = 69.685).

In model 3 communication was added, and it also significantly predicts customer loyalty. The R square in model 3 increased to 31.3%, implying that communication accounts for 1.4% increase in the variation in customer loyalty. Model 3 is also significant (sig. = 0.005, p < 0.01, F = 8.039).

In models 4 and 5, reciprocity and relationship satisfaction were added, respectively. These two came out as non-significant predictors of customer loyalty. In model 4, the R square increased to only 31.6%, and in model 5 the R square did not change at all (it remained as 31.6%). Thus, reciprocity and relationship satisfaction account for 0.3% and 0% increase (respectively) in the variation in customer loyalty. Models 4 and 5 are not statistically significant.

When word of mouth was added in model 6, it came out as a significant predictor of customer loyalty. The R square in model six increased to 35.9%, implying that word of mouth accounts for 4.3% increase in the variation in customer loyalty. Model 6 is statistically significant (sig. = 0.000, p < 0.01, F = 25.464).

The overall model is statistically significant (sig. = 0.000, p < 0.001, F = 25.464). The independent variables (trust, commitment, communication, reciprocity, relationship satisfaction and word of mouth) explain 34.9% of the variation in customer loyalty. In the final model, commitment, word of mouth and communication are the only significant predictors of customer loyalty (sig. = 0.000, p < 0.01; sig. = 0.000, p < 0.01 and sig. = 0.047, p < 0.05, respectively). A unit increase in commitment increases customer loyalty by 0.312 (Beta = 0.312). A unit increase in word of mouth increases customer loyalty by 0.247 (Beta = 0.247) while a unit increase in communication increases customer loyalty by 0.129 (Beta = 0.129). Thus, commitment has the strongest impact on customer loyalty, followed by word of mouth and then communication. This finding is consistent with Fullerton (2005) who revealed that commitment (especially based on shared values and identification) has a positive impact on customer loyalty. Other studies like Evanschitzky et al. (2006) confirm a significant interaction of affective commitment and continuance commitment on loyalty.

4.6. Model fit

Enders and Hoover (2012) avers that in a regression setting, coefficient of determination (R2 or r2) does not only assess the ability of a model to predict an outcome but generally indicates whether or not a model is a good fit of the data. Field (2006) also asserts that the overall coefficient of determination can provide a relative measure of fit for the model/regression equation, while the standardized estimation coefficients (betas) can closely approximate the magnitude of the effect. For example, a value of 1 means every point on the regression line fits the data; a value of 0.5 means only half of the variation is explained by the regression. Enders and Hoover (2012) further states that an R2 of 0.35 is a very high portion of variation to predict in a field such as the social sciences, and therefore is a good indicator that the model fits the data. In this study, we generate an R2 of 35.9% and adjusted R2 of 34.9%. Going by Enders and Hoover (2012) and Cohen (1988), this value indicates that the model in this study is a good fit of the data.

To further test the model fit, we went ahead to run a structural equation model (SEM) using the Amos for SPSS software Version 19. This software enables the building of models more accurately than with standard multivariate statistics techniques. The model fit indicators, their benchmark values as well as the values derived from the research model are shown in Table A6. The final structural equation model is indicated in Figure A2. Both the absolute and incremental fit indicators show that the model had a good fit of the data.

Figure A2. The structural equation model.

4.7. Mediation tests

The relationship marketing components were aggregated to form one latent variable called customer relationship marketing. We followed Baron & Kenny’s (1986), later updated by Kenny et al., (1998) approaches for testing mediation. We regressed customer loyalty on customer relationship marketing and word of mouth. Results indicate that customer relationship marketing affects word of mouth (Beta = 0.485, p < 0.01). Also, relationship marketing affects customer loyalty (Beta = 0.550, p < 0.01), which confirms that there is actually an effect to mediate. When customer loyalty is regressed on both word of mouth (mediator variable) and customer relationship marketing (independent variable) as predictors in the same model, word of mouth affects customer loyalty (Beta = 0.235, p < 0.01), with the coefficient of word of mouth as the mediator being significant. In this model, the absolute effect of customer relationship marketing on customer loyalty reduces from Beta = 0.550 to Beta = 0.436. Hence, the results confirm that word of mouth partially mediates the relationship between customer relationship marketing and customer loyalty (Sobel z-Value = 3.47781 Sig., 0.000506), supporting H8. The mediation analysis results are shown in Table A7 and diagrammatically represented in the MedGraph in Figure A3.

Figure A3. Medgraph for word of mouth as a mediator of customer relationship marketing and customer loyalty.

5. Conclusions and implications

The study finds a significant positive relationship between relationship marketing and word of mouth, and a significant positive relationship between word of mouth and customer loyalty. In our final model, only commitment and communication components of relationship marketing were significant predictors of customer loyalty. Trust, relationship satisfaction and reciprocity components were not significant predictors of customer loyalty contrary to some earlier studies. This unique finding can largely be explained by contextual factors such as the differences in demographics, the strength and length of the relationships and switching costs (Mittal & Kamakura, 2001; Ou, de Vries, Wiesel, & Verhoef, 2014). The collectivist nature of the Ugandans could also explain the variation in results. Due to collective conscience, customers are likely to ride on the trust and satisfaction of others in society, and may not care whether reciprocity exists as long as others trust the services after all, people tend to believe in common feelings and beliefs. Sysoev and Neiman (2004) also confirm that trust and satisfaction do not necessarily yield loyalty. Word of mouth was found to be a significant predictor of customer loyalty, and a partial mediator of the relationship between relationship marketing and customer loyalty. We confirm that relationship marketing is important in driving service loyalty, and is equally important in providing foundation for sound business practices in a least developed economy context where customer purchasing power is relatively low, with little saving to spend on such services like mobile telecommunication that are beyond the basic needs.

5.1. Theoretical implications

Scholars such as Alrubaiee and Al-Nazer (2010), Fournier et al. (1998) and Baggozi, (1995) have theoretically conceptualized customer loyalty as consisting of trust, satisfaction, communication commitment and reciprocity. Recent researchers such as Moretta Tartaglione, Cavacece, Russo, & Granata (2019) and Ou, Verhoef, and Wiesel (2017) have found trust, relationship satisfaction, commitment and reciprocity to be significant predictors of customer loyalty. To the contrary, this study finds communication and commitment as the only significant components of relationship marketing that predict customer loyalty. The mediating effect of word of mouth on the relationship between relationship marketing and customer loyalty is very evident and empirically supported. Empirical studies explaining the role of word of mouth in creating customer loyalty are not well documented. Previous studies on relationship marketing and customer loyalty have considered either direct effects or have placed much emphasis on measurement issues. In this study, we further posit that word of mouth is crucial in customer interactions. It tends to build collective conscience which will likely drive customers to buy a service on recommendation by others in society. This study confirms that relationship marketing is not only a philosophy but also a set of practices, prominent among these practices is communication and commitment. Thus, the study contributes to the understanding of the relationship marketing concept. The study strengthens the existing body of knowledge by providing some empirically tested insight in a least developed country context. The study confirms that in a collectivist society like Uganda, the interaction theory is more appropriate in explaining relationships. Lastly, we prove that service loyalty is a function of communication, commitment and word of mouth, hence relationship marketing and word of mouth are crucial in a service setting. To this extent, this study contributes to the existing body of knowledge.

5.2. Practical implications

In this study word of mouth was found to be a significant predictor of customer loyalty. Word of mouth was also found to be a partial mediator of the relationship between relationship marketing and customer loyalty. With markets becoming increasingly competitive and less profitable, companies have resorted to cutting promotional expenditure but remain competitive in the market. Companies should consider investing their marketing efforts in word of mouth. Word of mouth is a low-cost investment promotional tool that yields customer loyalty. Companies should also direct their marketing efforts in strengthening their interactions and relationships with their customers through proper and effective communication channels. Companies should further demonstrate commitment to their relationships with their customers. Companies should create and intensify activities that increase their interactions with the customers as a way of cementing their relationships as well as create a sense of relationship satisfaction. Companies should also look for influential people in the communities to speak good about their products and services as a way of spreading the good word of mouth. As competition in the telecom sector intensifies due to deregulation and technological developments, keeping existing customers becomes more challenging as they keep switching. One way of countering this is to devote considerable effort or investment in low-cost promotional tools like the spread of word of mouth. Thus, marketers need to concentrate their marketing efforts on the relationship marketing components of communication and commitment, as well as creating a positive word of mouth to create service loyalty. This will drive marketers to shift from short term and transaction focus to long term and relational focus, which they should consider as one of their deliberate prominent strategy.

6. Limitations of the study and areas for further research

Our sample was limited to mobile telecom customers in Kampala city since all the telecom companies have at least coverage in Kampala. While telecom companies provide a wide range of services including fixed telecom and internet services among others, our study was limited to the mobile telecommunications service. Thus, generalizing and applying the results of this study across other industry sectors and other geographical locations should be done with caution. Future studies may explore the extent to which word of mouth mediates the relationship between relationship marketing and customer loyalty among other industry settings beyond the mobile telecom. Given the differences in geographical settings and location advantages, it is pertinent for future research to examine if the concept of loyalty to a mobile telecom network is plausible in rural areas or areas away from the central cities where some mobile phone networks cannot be tapped. The specific issues to explore would include situations where customers may repurchase or continue subscribing to a mobile telecom network (with or without positive WOM) due to situational factors or technical hitches like inability to access/tap other networks, absence of viable alternatives or vendor lock-in among others. This will help rule out the possibility of spurious loyalty.

Additional information

Funding

The authors received no direct funding for this research.

Notes on contributors

Muhammed Ngoma

Assoc. Prof. Muhammed Ngoma is the Dean, Faculty of Graduate Studies & Research, Makerere University Business School. His research interest includes; Institutional networks, Entrepreneurial training and development, Venture creation, Family businesses, and general management.

Peter Dithan Ntale

Peter Dithan Ntale is the Deputy Director, Department of Doctoral Training in the Faculty of Graduate Studies & Research, Makerere University Business School. His research interests include; Inter-agency collaboration, Collaborative Intelligence, Collaborative governance, Institutional leadership & management, Business Psychology and entrepreneurial training and development.

Appendix

Table A1. Cronbach alpha (α) coefficients

Construct	Number of scale items	Cronbach alpha (α)
Trust	8	.844
Commitment	10	.880
Communication	4	.798
Reciprocity	3	.712
Relationship Satisfaction	8	.778
Word of Mouth	13	.653
Customer Loyalty	11	.761

Table A2. Item to total analysis

Items	WM1	WM2	WM3	WM4	WM5	WM6	WM7	WM8	WM9	CL1	CL2	CL3	CL4	CL5	CL6	CL7	CL8	CL9	CL10	CL11	TCL	TWM
WM1	1
WM2	.222	1
WM3	.377	.274	1
WM4	.197	.309	.499	1
WM5	.216	.141	.235	.220	1
WM6	.065	.205	.103*	.105*	.288	1
WM7	.080	.135	.088	−.019	.205	.241	1
WM8	.147	.129*	.243	.202	.192	−.027	.059	1
WM9	.064	−.098	.089	.020	−.037	.012	−.010	.227	1
CL1	.230	.054	.220	.086	.161	.026	.021	−.099	.114	1
CL2	.158	.172	.294	.381	.118	.027	.000	.238	.117	.236	1
CL3	.152	.041	.201	.117	.114	.367	.029	−.118	.075	.106	−.026	1
CL4	.292	.154	.314	.231	.029	.139	.214	.095	.033	.419	.424	.056	1
CL5	.320	.213	.116*	.284	.024	.048	.084	.104	.020	.005	.168	.143	.093	1
CL6	.320	.203	.350	.391	.076	.076	−.063	.267	.071	.344	.216	−.069	.485	.119*	1
CL7	.063	−.037	.166	.000	.267	.117	.112	.069	.113	−.005	.137	.278	−.117*	.037	−.216	1
CL8	.290	.272	.301	.360	.093	−.004	.211	.236	.006	.254	.202	−.067	.361	.374	.449	−.010	1
CL9	.363	.310	.368	.352	.210	.118*	.125	.165	−.132	.215	.109*	.144	.377	.340	.311	.162	.446	1
CL10	.353	.327	.394	.552	.182	.065	.114	.285	.046	.222	.315	.011	.449	.382	.567	−.026	.529	.546	1
CL11	.405	.220	.296	.367	.148	−.136	.117	.306	.095	.219	.226	−.019	.331	.417	.373	−.045	.506	.365	.540	1
TCL	.035	.067	.032	.109*	.063	.180*	.101*	.152*	.019	.410**	.365**	.209*	.415**	.410**	.420**	.697**	.502**	.509**	.603**	.523**	1
TWM	.430**	.201*	.584**	.542**	.542**	.481**	.384**	.429**	.295**	.026	.187*	.097	.036	.147	.062	.066	.020	.019	.178*	.194*	.246*	1

**.correlation is sign. at the 0.01 level (2-tailed): *.correlation is sign. at the 0.05 level (2-tailed): CL = cust. loyalty; WM = word of mouth; TCL = total cust. loyalty; TWM = total word of mouth

Table A3. Major indicators of the factor analysis

Constructs	No. of items that loaded	Factor Loadings	Total Variance Explained (%)
Trust	5	.51 to .83	49
Commitment	5	.58 to .81	49.2
Communication	4	.75 to .83	62.4
Reciprocity	3	.77 to .83	64
Word of Mouth	5	.51 to .69	40
Relationship satisfaction	5	.50 to .75	39
Customer Loyalty	6	.51 to .76	41

Table A4. Correlation coefficients

	Mean	S.D	Min	Max	1	2	3	4	5	6	7
1. Trust	3.73	.682	1.00	5.00	1.00	–
2. Commitment	3.59	.715	1.00	5.00	.734**	1.00	–
3. Communication	3.32	.996	1.00	5.00	.508**	.713**	1.00	–
4. Reciprocity	3.30	.860	1.00	5.00	.520**	.719**	.679**	1.00	–
5. Relationship satisfaction	3.40	.691	1.00	4.00	.669**	.800**	.697**	.662**	1.00	–
6. Word of mouth	3.22	.508	2.00	4.00	.333**	.448**	.399**	.523**	.357**	1.00	–
7. Customer Loyalty	3.68	.521	2.00	5.00	.414**	.547**	.474**	.454**	.469**	.447**	1.00

**. Correlation is significant at the .001 level (2-tailed)

Table A5. Hierarchical regression results

	Model 1 Beta Sig.	Model 2 Beta Sig.	Model 3 Beta Sig.	Model 4 Beta Sig.	Model 5 Beta Sig.	Model 6 Beta Sig.
Trust	.414 .000	.027 .673	.032 .605	.033 .599	.030 .635	.020 .744
Commitment		.527 .000	.400 .000	.363 .000	.357 .000	.312 .000
Communicatn			.172 .005	.145 .025	.142 .034	.129 .047
Reciprocity				.078 .229	.077 .242	.027 .685
Satisfaction					.013 .866	.046 .537
Word of mouth						.247 .000
F	79.077	69.685	8.039	1.450	0.028	25.464
Sig. F	0.000	0.000	0.005	0.229	0.866	0.000
R	41.4%	54.7%	56%	56.2%	56.2%	59.9%
R^2	17.1%	29.9%	31.3%	31.6%	31.6%	35.9%
R^2 change	17.1%	12.8%	1.4%	0.3%	0.0%	4.3%
Adjusted R^2	16.9%	29.5%	30.8%	30.9%	30.7%	34.9%

Table A6. Model fit indicators

Measure Best Range (Garson, 2008; Hu & Bentler, 1999)
	Absolute Fit Measurement (AFM)	Model Reported Values
Chi-Square [X^2 (p)]	P ≥ 0.05	15.474 (0.162)
Goodness-of-Fit Index (GFI)	>0.9	0.945
Root Mean Squared Error of Approximation (RMSEA)	<0.08	0.033
Incremental Fit Measures (IFM)
Comparative Fit Index (CFI)	>0.9	0.965
Incremental Fit Index (IFI)	>0.9	0.984
Tucker-Lewis Fit Index (TLI)	>0.9	0.987

Table A7. Mediating effect of word of mouth

Dependent Variable: Customer Loyalty
	Equation/Model I:			Equation/Model II:			Equation/Model III:
	B	SE	Beta	B	SE	Beta	B	SE	Beta
Constant	1.955	.119		2.208	.116		1.736	.148
CRM	.365	.034	.485	.424	.033	.550	.336	.037	.436
Word of Mouth							.241	.049	.235