Application of Commitment Trust Theory and Relationship Marketing to the MMO Gaming Context

The concept of being entertained, of spending (possibly limited) personal time in a way that brings personal emotional satisfaction is at the heart of the entertainment industry (Vogel, 2007). However, the continuing relationship that an individual has with his local golf club and paying his monthly dues is very different from the continuing relationship that the same individual may have with his local energy supplier; though both may entail monthly payments. Unsurprisingly as a result, in practice, different types of companies treat customers of their products in different ways, some even rejecting customer relationship development as an option due to the nature of purchase decisions or the income model used by the business (Palmer, 1996, p.20).

Commitment Trust Theory as posited by Morgan and Hunt (1994a) is defined by the theoretical assertion that

…trust and relationship commitment play key mediating roles in the process of relationship development and relationship performance…

that (1) relationship benefits and termination costs relationship influence commitment, (2) shared values influence both commitment and trust, and (3) communication and opportunistic behaviour directly influence trust (and, through trust, indirectly influence commitment)…Morgan and Hunt propose qualitative outcomes and theorize that these outcomes promote relationship success. (Goo and Huang, 2008, p.217)

The theory centres around the identification of key characteristics or variables, called constructs in the marketing literature, that have causal relationships with each other. Morgan and Hunt’s (1994a) “mediating variable approach” (p.22) focuses on the antecedents that impact on the decision making processes of relationship commitment and trusting behaviours, and associations between these variables, and the relationship outcomes. This

approach uses positivist deductive research strategies and methods, including the sampling of a given population (usually with survey questionnaires, including online ones), and statistically tests a number of investigative hypotheses for associations between the framework’s antecedents and outcomes, usually using Structural Equation Modelling techniques (Holdford and White, 1997; MacMillan et al., 2005; Goo and Huang, 2008).

Commitment Trust Theory is a predictive and causal theory. It establishes that in particular contexts, certain antecedents influence the mediating variables of Commitment and Trust which are psychological outcomes of the antecedents. These mediating variables moderate and influence the behavioural outcomes of the relationship; propensity to leave, stay, become frustrated and so forth (Figure 2 below).

Figure 2: Morgan and Hunt (1994a) Original Framework

The theoretical assertions that Morgan and Hunt (1994a) present in their framework are based on established conceptual models (MacMillan et al., 2005, p.808), with Commitment Trust Theory “founded upon empirical findings” (Friman et al., 2002, p.404) and representing a hybrid of theories for the Relationship Marketing field. This is conceptually common approach in a

discipline which regularly integrates and hybridises theories drawn from economics, political science, organisational sciences, sociology and social psychology and law (Eiriz and Wilson, 2006).

Social Exchange Theory (Blau, 1964; Chadwick-Jones, 1976; Heide and John, 1988) is used by Morgan and Hunt (1994a) as the basis of the prediction that relationship benefits and termination costs drive relationship commitment. The positive contribution of trust towards relationship commitment is also based on Social Exchange Theory with Morgan and Hunt (1994a) drawing upon empirical contributions of Achrol (1991) and Moorman et al. (1992) which were developed from interpersonal sociological theories of long term exchange (Perlman and Duck, 1987). Similarly, the link between shared values and commitment is based on the social–psychological theory of attraction based on similarity (Berscheid, 1985), and the theories of the development of commitment norms and opinions in business and interpersonal interactions (Kelman, 1961; Mathieu and Zajac, 1990;

Chatman, 1991; Heide and John, 1992).

The contribution of shared values, communication and (a lack of) opportunistic behaviours towards trust have been identified as all having

“…the same theoretical foundations and are based on socio-cognitive learning theories which assume that individuals are rational and learn from the past.” (MacMillian et al., 2006, p.808). Conceptually these theories are drawn from established interpersonal learning literature (Heider, 1958; 1980), with Morgan and Hunt (1994a) drawing upon the empirical studies of Dwyer et al. (1987), Heide and John (1992) and Moorman et al (1993) as reinforcement to their claims.

The rationale for the selection of Commitment Trust Theory is that it is focused on understanding the key drivers of a successful relationship.

Morgan and Hunt (1994a, p.31) are explicit that their aim is to provide managers, marketers and researchers with clear guidance as to not only which antecedents are relevant, but also which antecedents are important.

This investigative approach synergises well with the principal research aim of

understanding the re-subscription decision drivers in MMO customer relationships.

Commitment Trust Theory has mainly used the technique of Structural Equation Modelling as the analysis method for simultaneous testing of the construct inter-relationships, and for examining a series of predictive hypothesis. Morgan and Hunt’s (1994a) initial study predicted 13 relational hypotheses between the variables, though other studies have presented more and less depending on how they have adapted the model to the environment that they are examining. Commitment Trust Theory has used Structural Equation Modelling in its empirical testing of these hypotheses due to a number of interconnected reasons.

Firstly, Commitment Trust Theory’s use of latent psychological variables, such as Trust and other abstract psychological variables, rather than more easily measurable manifest variables, means that the measures themselves are error prone (Garson, 2009) due to the nature of what they are trying to measure. Commitment Trust Theory’s use of Structural Equation Modelling, and multiple measures per latent variable, means that measurement error can be reduced through Structural Equation Modelling’s use of confirmatory factor analysis. This technique concentrates on the convergent and discriminant validity of the measures which constitute the latent variable.

Secondly, Structural Equation Modelling allows for the measurement of models which use mediating variables rather than being limited to an additive linear regression model. This allows for antecedent constructs such as Relationship Benefits to contribute positively towards the mediating variable of Commitment, which itself contributes to a behavioural outcome. A linear approach would have to apply both Commitment and Relationship Benefits additively, ignoring the mediating effect of the Commitment Construct.

Morgan and Hunt (1994a) even go as far as in their initial study to use an additive linear “Rival Model” (p.30) as a comparator with which to establish that a mediating approach is more valid than a linear additive approach for

replicated by this study similarly to add rigour to the interpretation, and validity to the claim, that Structural Equation Modelling provides a more valid interpretation of the research data than an additive linear model.

Lastly, it is the concentration of Structural Equation Modelling on comparing alternative rival models to assess model fit (Garson, 2009) that adds robustness and validity to the results of Structural Equation Modelling. While Morgan and Hunt’s (1994a) initial research concentrated on establishing that the relationship was better modelled with Commitment and Trust as the key psychological mediators, subsequent examinations have gone on to compare this baseline model against a number of alternative explanations.

This study furthers this approach by including the concept of effect size (Cohen, 1988) in the analysis, comparing the explanatory power and goodness of fit of this study’s relational model against a model which includes only those relationships which display a significant effect size.

While the Commitment Trust framework has been applied to a number of different business relationship environments, as of yet it has not been applied to the business of MMO games. Morgan and Hunt’s (1994a, p.33) concluding comments are clear though that “…the model needs further explication, replication, extension, application, and critical evaluation” with a subsequent call for future application and development tied to marketing business practice. Commitment Trust Theory (Morgan and Hunt, 1994b) is a contextual theory, with the original study conducted using a national survey of more than 200 independent automobile tyre retailers to examine the key drivers in those ongoing business relationships. Subsequently the theory has been used in a wide range of diverse situations, from examining the relationship between pharmacy students and their programme of study (Holdford and White, 1997), to the relationship between a not-for-profit organisation and its organisational funders (MacMillan et al., 2005) and the relational governance in an outsourcing engagement (Goo and Huang, 2008), to cite but a few examples.

Perhaps as influential to this study as the original Morgan and Hunt (1994a) Commitment Trust Theory research itself, Gruen’s (1995) adaptation of the conceptual framework to mass market business-to-customer relationships is fundamental to this study’s approach. Firstly, Gruen’s (1995; 2000) concepts of the business-to-customer relationship as a membership are the foundation of this study’s operational definition of the relationship. Second, Gruen’s (1995, p.451) characteristics of business-to-customer relationships influence and contribute towards this study’s operational definitions of the domain of the communication construct. Linked to this, Gruen’s (1995, p.455) inclusion of Expectancy Theory in the operational definition and domain of the Trust construct forms the basis of this study’s delineation of past satisfaction competence-based expectancy from Trust benevolence-based expectancy, which is a fundamental premise of this analysis. Thirdly, Gruen’s (1995, p.453) adaption of the conceptual framework links Satisfaction to both Commitment and Trust and links Commitment to Propensity to Leave (Future Intentions). Finally, the analysis by Gruen (1995, p.464) of the interaction of the psychological outcomes forms the basis of this study’s evaluation of the results.

To conform with the characteristics of an online gaming entertainment product, with a subscription based renewal method, this study adds such constructs as customer satisfaction, past satisfaction and other relevant contextual antecedents. These contribute towards ongoing relationship commitment (See Figure 3 below) in keeping with the findings of Gruen’s (1995) adaptations of Commitment Trust Theory to the business-to-customer market. This is reflecting the emergent properties from the literature, which differentiate an entertainment product with an ongoing subscription element from other types of product relationships. These constructs are further discussed in Chapter 3 which details the construct domain and the reasons behind construct selection and delineation.

Figure 3: A Model of Commitment and Trust in the MMO game setting Having established a rationale for the choice of this study’s conceptual framework, Commitment Trust Theory, this study now moves on to consider the wider question of why Relationship Marketing as a discipline is applicable to the MMO re-subscription business-to-customer relationship. A relationship which is an income generating one for the product generating company that is based on on-going relational exchanges, rather than discrete one-off payments.

Relationship Marketing itself has been defined in many ways over the years.

While Berry’s (1983, p.25) definition may be the most recognised it is also generalistic and as such this study takes Gronroos’s (1996, p.11) definition as its basis due to its comprehensiveness. In particular Gronroos’s (1996)

definition focuses on the role of relationship marketing to “identify and establish” (p.11) relationships, which is seen to synergise well with this study’s approach to the investigative problem:

identify and establish, maintain and enhance and, when necessary, terminate relationships with customers and other stakeholders, at a profit so that the objectives of all parties involved are met; and this is done by mutual exchange and fulfilment of promises (Gronroos, 1996, p.11)

Marketing research differentiates clearly between discrete exchanges and relational exchanges. Discrete exchanges are seen to have a beginning, an end, a short duration and involve usually anonymous parties, with little or no history, and little or no future interaction intention (Egan, 2008, p.38).

Companies that evidence these characteristics are seen to find traditional transactional marketing more useful to meeting their goals and requirements, with it perhaps even “unrealistic” (Palmer, 1996, p.20) to examine these types of businesses or products using relational approaches. In contrast, relational exchanges are based around the concept of “…a series of exchanges over a long (or indefinite) period of time..” (Hunt et al., 2006, p.77) in which the very first sale is perceived as just the beginning of a long term relationship. Subsequent research into successful relational exchanges (Hunt et al., 2006, p.77) has found six factors which are repeatedly evidenced and cited in the literature; trust (Dwyer et al., 1987; Morgan and Hunt, 1994a; Wilson, 1995; Smith and Barclay, 1997; Sivadas and Dwyer, 2000), Commitment (Anderson and Weitz, 1992; Moorman et al., 1992; Day, 1995; Geyskens et al., 1999), communication (Mohr and Nevin, 1990; Mohr et al., 1996), keeping promises (Gronroos, 1990, 1994), shared values (Morgan and Hunt, 1994a; Yilmaz and Hunt, 2001; Brashear et al., 2003) and cooperation (Anderson and Narus, 1990; Morgan and Hunt,1994a), with knowledge of these relational success factors seen to “…urge marketers to develop and nurture the characteristics of relationships that are associated with successful relational exchange...” (Hunt et al., 2006, p.78) in order to maximise the effect of a relational marketing strategy.

The theoretical reasons for the use of Relationship Marketing to examine MMO games stem from the very nature of the product as an on-going entertainment expenditure by the customer. Research has shown (Egan, 2008, p.117-18) that the greater the emotional involvement from the customer in the transaction, the greater the drivers promoting the use of examining the transaction from a relationship marketing perspective. With a survey by the Interactive Software Federation of Europe (2008) finding that 80% of respondents said they played computer games for fun and enjoyment, 55% for relaxing and de-stressing and 41% as a relief from boredom, certainly substantial evidence exists that emotional involvement is at the core of a MMO gaming experience (Yee, 2004; 2006). In addition, the emotional dynamics of an entertainment subscription relationship can be considered a “higher level” motivational investment on the part of the customer (Dwyer et al., 1987) with comparable previous Relationship Marketing research investigating football season-ticket sales (Gruen, 2000, p. 356), TV channel subscriptions (Burez and Van den Poel, 2007) and similar entertainment products.

Relationship marketing theorists, coming from an applied discipline perspective, have discussed the “considerable evidence that organisations are increasingly applying relationship marketing concepts in mass markets”

(Bhattacharya and Bolton, 2000, p.327), with one of the key issues behind this seen to be the importance of customer retention. Discussing the necessary pre-conditions for relationship marketing both Bhattacharya and Bolton (2000, p.328) and Egan (2008, p.108) discuss the impact of customer acquisition costs and retention upon business strategies, with those businesses who incur high acquisition costs relative to retention costs seen to pursue more relational strategies with their customers, with even small increases in retention seen to have large beneficial effects on business profitability (Fornell and Wernerfelt, 1987, 1988; Reichheld and Sasser, 1990). Especially apparent though is the need for the mass market product or service to have attributes of “customer intimacy” (Bhattacharya and Bolton, 2000, p.330) to form a relationship. The invoking of an emotional

response, “…happiness, pride, achievement…” (Bhattacharya and Bolton, 2000, p.330), which leads to a sense of identification or affinity developed over an extended time period is seen as an important part of the

“…necessary conditions for an exchange relationship to exist” (Bhattacharya and Bolton, 2000, p.329) in a mass market. Correspondingly, Egan (2008) discusses how “high emotion involved in the exchange” (p.117) can be a driver for a relational approach, and especially when attachment, membership or affinity are involved might go beyond the business-to-customer channel and the business-to-customer may start to incur emotional or social costs for leaving the product (p.87).

The practice-based reasons for using Relationship Marketing come from the practises that companies who run MMO games actually utilise; with usage of loyalty schemes (sometimes with exclusive in-game items as loyalty bonuses for long time subscribers), using customers to sell to their friends (recruit-a-friend schemes) and using current customers to try and persuade former customers who have quit to return (Activision Blizzard’s “Scroll of Resurrection” scheme for example). These are all classic examples of

“leveraging” (Hart et al., 1999, p.544) the customer relationship, as seen in a wide number of industries and products. Industry and practice papers (Doorselaer and Coopens, 2003), clearly discuss that game publishers of MMO games are not just creating a disc and some computer code which they can publish, market and then walk away from (a viable business model in non-subscription based video games for many consoles) they are instead

“game service providers” (p.118). The emphasis is on the ongoing service element, as the majority of the revenues generated by a MMO game are being generated not by the initial purchase decision, but by the ongoing re-subscription decision made every month (Figure 4 below). Therefore there is an economic imperative, innate to the MMO business model, to engender retention and loyalty strategies; strategies which are seen as core outcomes of Relationship Marketing (Palmatier et al., 2006). Hence, when examining the decision by the customer to keep subscribing to a massively multiplayer

game there are both theoretical and practice based reasons behind why an examination from a Relationship Marketing perspective is valid.

Figure 4: A comparison of revenue streams over time (adapted from Doorselaer and Coopens, 2003)

Definitions of domain Relationship Marketing vary and some academics would even contend that Relationship Marketing should be used to examine all companies and products alike (Rogers and Peppers, 1994). A more balanced, and more mainstream, approach (Christy et al., 1996; Voss and Voss, 1997) would be that different products and different types of customer relationships create different types of dyadic business relationships; including ones in which Relationship Marketing perhaps should not be used.

Consequently with the dyadic relationship being a core principle of the Relationship Marketing approach, and of Commitment Trust Theory, this study now sets out its definition of how the dyadic relationship channel is viewed.