Customer Advocacy: A Distinctive Form of Word of Mouth

Abstract

Enthusiastic customer endorsements can significantly influence buying decisions and drive sales. In service contexts, advocates are especially important because the specific and complex characteristics of services make personal recommendations very effective. Customer advocacy communications differ from other positive word of mouth (PWOM), though literature on advocacy is surprisingly sparse and inconsistent. Notably, advocacy is strong, passionate, explicit, and ongoing, with an explicit goal of positively influencing others’ views. As its central theoretical contributions, this article defines advocacy; identifies advocacy as a distinctive form of PWOM; conceptualizes advocacy according to a hierarchy of behaviors, which increase in intensity and effort; and develops a robust, reliable advocacy scale. By investigating positive behavioral outcomes of different levels of advocacy, this study also reveals the importance of identifying strong advocates, rather than just those who give PWOM, as well as salient drivers of advocacy. Accordingly, this article establishes a platform for further investigations of the importance of advocates, as well as recommendations to help managers identify these potentially valuable advocates.

Keywords

advocacy positive word of mouth Rasch modeling advocacy profile

Customers have grown less receptive to mass marketing (Nielsen 2007), forcing organizations to reevaluate their expenditures on advertising and reconsider the potential of customer referrals (Kilby 2007). When customers enthusiastically tell others about their experiences with a firm’s services, such communications can influence nearly 70% of all buying decisions (Balter 2008); they represent a driving force in two thirds of industries (Dye 2000). Such advocates are unlike customers who engage in general word of mouth (WOM), by which we mean informal communication, such as recommendations and evaluations of goods and services (Zeithaml, Berry, and Parasuraman 1996). Advocates are passionate in their endorsements of products, so they help secure new customers, encourage repeat purchases, and increase brand awareness. Despite recognition that advocates can contribute to firm success, marketing has yet to conceptualize customer advocacy clearly or examine how it relates to antecedents and loyalty behaviors. In particular, unclear or simplistic conceptualizations of WOM (Mazzarol, Sweeney, and Soutar 2007) lead to inaccurately synonymous uses with related terms, such as referrals (Garnefeld et al. 2013) or opinion leadership (Rogers 1995). The term positive word of mouth (PWOM) is a general description of all positive customer-to-customer communication (Swan and Oliver 1989); we lack a sufficient clear distinction of weak versus strong, implicit versus explicit, or plain versus passionate PWOM.

Instead, we argue that advocacy has important distinguishing characteristics, compared with more general PWOM, in that it entails a strongly expressed recommendation (Wilson 1994), delivered in a forceful manner (Krapfel 1985), which reflects a more entrenched relationship, such that advocates even defend the organization against rivals (Walz and Celuch 2010). The behavioral outcomes associated with advocacy in turn should be more extreme than those associated with general PWOM, so approaches to identify and encourage valuable customer advocates should offer significant benefits for firms. With this theorizing, we address a research gap identified by Walz and Celuch (2010) and Payne and Frow (2013), related to the need for specific empirical research on advocacy, by examining it in service contexts, in which recommendations tend to be more salient (Brown et al. 2005) and personal sources of information are often critical input to consumption decisions (Murray 1991). That is, it is relatively more difficult for service providers, compared with goods providers, to offer consistent customer experiences (Esquivias et al. 2013), so reassurance through recommendations is especially helpful in these settings, and service companies are particularly likely to benefit from advocacy.

In turn, this study makes five specific contributions: We (1) define advocacy, as a distinctive form of PWOM; (2) conceptualize advocacy according to a hierarchy of behaviors that increase in intensity and effort; (3) develop a robust, reliable advocacy scale that reflects this conceptualization; (4) examine the positive behavioral outcomes of customers engaged in different levels of advocacy; and (5) examine salient antecedents or drivers of advocacy. To achieve these contributions, we first discuss relevant WOM and advocacy literature. Next, we report on three studies: a qualitative study exploring the concept of advocacy and identifying a set of 12 advocacy items; a large-scale quantitative study, which uses a Rasch modeling approach (Andrich 2011) to develop a valid and reliable advocacy scale; and a quantitative study that identifies antecedents and outcomes of advocacy. Finally, we discuss our findings, along with their implications and some research directions.

WOM and Related Terms

Our purpose is not to review the entirety of the extensive stream of WOM research; comprehensive examinations are available elsewhere (e.g., Berger 2014; East, Hammond, and Wright 2007; Keiningham et al. 2018; Sweeney, Soutar, and Mazzarol 2012). In these contemporary views of WOM activity and behavior, WOM often functions as an umbrella term, spanning both positive and negative forms. We seek a more granular approach to advocacy to identify the characteristics that distinguish it from PWOM and related terms. Specifically, consumers frequently use WOM to share their experiences, explain their preferences, and seek opinions. Broadly, WOM communications include any informal discussion or recommendations of products through noncommercial, person-to-person, and online channels. In investigating specific aspects of WOM, Sweeney, Soutar, and Mazzarol (2012) study the structure of powerful WOM messages and identify three communication style dimensions pertaining to how it gets conveyed, rather than actual behaviors of WOM givers, which represent the focus of this study.

The umbrella concept of WOM also encompasses several related terms used to describe WOM givers. Doss and Carstens (2014) usefully distinguish several key terms. For example, market mavens display a comprehensive understanding of multiple products and brands but do not necessarily seek to influence consumption behavior (Feick and Price 1987). In contrast, opinion leaders have in-depth insights into a specific product category in the marketplace, through their long-term involvement with it, such that they become established as experts (Bloch and Richins 1983). They seek recognition as sources of useful product information (Summers 1970), which may motivate their advocacy, though it is not advocacy itself. The term evangelist evokes values that underpin religious evangelists (Kawasaki 1992); in a marketing context, it refers to customers who promote a company’s products by “trying fervently to convince or persuade others” to buy (Matzler, Pichler, and Hemetsberger 2007, p. 27), using zeal, guts, and cunning (Kawasaki 1992). Evangelists are rare, and firms need to exert much effort to create them (Collins and Murphy 2009). Their excessive behavior also transcends “loyalty beyond reason” (Roberts 2005, p. 66). Finally, a promoter can be defined according to the net promoter score (NPS) developed by Reichheld (2003), which is based on the response to a single question: “How likely is it that you would recommend our company/product/service to a friend or colleague?” This measure has been subject to considerable criticism, due to the arbitrary boundaries around scores (Keiningham et al. 2007; Zaki et al. 2016) and its focus on intentions, such that it is a loose proxy for actual behavior (Raassens and Haans 2017). Lowenstein (2011) reports that only 56% of high NPS customers of a U.S. bank actually function as advocates.

Advocacy

The term “advocacy” derives from the Latin advocare (to avow, vouch, or justify). Broadly, it refers to highly influential communication, generated with the explicit goal of influencing others. Advocacy is poorly defined though, especially relative to the extensive body of work on WOM, and we find surprisingly little conceptual and empirical research on this topic. Wilder (2015) asserts that this paucity results from the lack of consensus regarding its definition across both scholarly and practitioner literature.

In Table 1, we identify some illustrative definitions and descriptions of advocacy from scarce prior literature, which reveals the confusion of perspectives available. In particular, advocacy takes different meanings, often used interchangeably with WOM or PWOM. For this study, we focus on advocacy initiated by a customer, or customer advocacy, as an extreme form of PWOM. An alternative usage would pertain to organization-initiated advocacy, or organizational advocacy for customers (e.g., Urban 2004). In this latter case, organizations provide customers with objective advice, which means they even might recommend a competitor, according to customers’ needs and to represent their best interests. We adopt the first perspective of customer advocacy for this study.

Table 1.

Illustrative Definitions and Descriptions of Advocacy.

Source	Definition/Description
Cross and Smith (1996)	Advocacy involves promoting a service to others and even defending it against detractors.
Bettencourt (1997)	Advocacy is a customer citizenship behavior, relating to positive WOM and customer recommendation.
Bendapudi and Berry (1997)	Advocacy is the ultimate test of a customer’s willingness to promoter service to others including defense of the service against detractors.
Urban (2004)	Advocacy involves a company acting on behalf of its customers, representing them within the firm and protecting their best interests with the objective of earning customers’ trust loyalty and future purchases.
Fullerton (2005)	Advocacy includes identification with a company, positive WOM, and recommending the firm to others.
Park and MacInnis (2006)	Advocacy is at the highest level of a consumer behavioral hierarchy and reflects consumer investment of their resources (e.g., time, money, and credibility).
Keller (2007)	Advocacy is a form of WOM, but concepts are used interchangeably.
Roy (2015)	Advocacy involves a firm creating mutual dialogue with customers. If a company acts as an advocate for its customers, they will reciprocate with a higher share of wallet, greater trust, and their loyalty.
Bhati and Verma (2017)	Advocacy is an extreme form of WOM, which is not only positive but also proactive.

Note. WOM = word of mouth.

Another potential confusion results from the insufficient distinctions among advocacy, PWOM, and willingness to recommend. Marketing scholars often use these terms synonymously. For example, Keller (2007) describes advocacy as PWOM that drives recommendations, and Reichheld (2003) views it in terms of willingness to recommend. We instead propose that advocacy constitutes a distinctive, granular form of PWOM, distinguished by the strength of the message and its influence on an audience. We concur with Wilder’s (2015) argument that using PWOM as a proxy for advocacy is misleading, because PWOM does not encompass the personal, relationship-based nature of advocacy. Positive comments about or mere recommendations of a service do not necessarily represent advocacy (Hill, Provost, and Volinsky 2006); rather, advocacy must include strong recommendations and praise (Fullerton 2011). It involves a proactive, strongly expressed recommendation, delivered in a more forceful manner than PWOM (Wilson 1994). Advocacy also involves persuasive effort, whereas PWOM may involve positive comments without any aim to influence others (Mazzarol, Sweeney, and Soutar 2007).

The concept of advocacy also can be linked to relationship marketing. Christopher, Payne, and Ballantyne (1991) distinguish advocates who are deeply involved in a relationship with the service provider from supporters who represent a weaker relationship level and only offer PWOM. Bendapudi and Berry (1997, p. 30) also conclude that “The ultimate test of the customer’s relationship with the service may be whether the customer is willing to become an advocate for the service.” The relational characteristics of advocacy thus may mean that customer advocates spend more for a given service than do those who just provide PWOM. However, a poor understanding of why some customers are proactive, passionate, and persistent in their recommendations, whereas others offer only general PWOM, persists due to the lack of research into advocacy specifically.

With the assertion that advocacy is an extreme, potentially beneficial form of PWOM, we seek to conceptualize its richness and depth by proposing a hierarchy of advocacy behaviors. For example, a moderate level of advocacy might be associated with a few behaviors that distinguish it from more general PWOM; strong advocacy should feature additional distinguishing behaviors. Strong advocacy and moderate advocacy also might share some parallels with the concepts of delight and satisfaction. Customer delight refers to an extremely positive emotional state, resulting from having expectations exceeded to a surprising degree (Rust and Oliver 2000). The difference between satisfaction and delight thus can be depicted in terms of concentric rings: The innermost ring refers to basic “must-haves,” the next ring contains embellishments to the basic product or service, and a third ring represents attributes that are highly and unexpectedly enjoyable (Rust and Oliver 2000; Rust, Zahorik, and Keiningham 1995). If we apply this concentric ring framework to advocacy, we can predict that all advocates, at a minimum, provide strong recommendations and say positive things about the service provider. Stronger advocates likely defend the provider (Cross and Smith 1996). As advocacy levels continue to rise, advocates may become proactive in their promotion and initiate informative messaging (Huffaker 2010). These advocates in turn offer additional benefits to the firm due to their very loyal and committed behaviors. In our proposed hierarchy, more easily initiated behaviors should be common to all advocates, but more challenging behaviors are carried out only at higher levels of advocacy. In summary, we predict that advocates exist to varying degrees that correspond to their specific behaviors.

Empirical Studies

We report on three studies: Study 1 is a qualitative study that defines advocacy, Study 2 identifies a hierarchy of advocacy behaviors and an advocacy scale, and Study 3 develops a conceptual framework of customer advocacy, including its antecedents and important outcomes. In this way, we confirm its distinctiveness from general PWOM.

Study 1: Exploratory Qualitative Study

Research Approach

We explore the conceptual domain of advocacy by first developing a set of initial items that we subsequently test in an advocacy scale. We conducted semistructured interviews with service consumers who were potentially advocates because they had issued a “very strong recommendation” (score 6 or 7 on a 7-point scale, where 1 = very weak, 7 = very strong) for a service provider in the past 6 months. The interviewees, recruited through convenience sampling, all indicated that they had offered PWOM about multiple service providers; however, they had provided very strong recommendations for just a few, select providers. A total of 14 interviews were conducted, leading to saturation, such that no new information was forthcoming. The diverse sample ranged in age from 22 to 64 years; was broadly balanced in terms of gender (eight men, six women), education, and nationalities; and represented various occupations (e.g., dental technician, IT director, and defense force officer). The in-depth interviews lasted 60–90 minutes, focusing on (1) the context, (2) the interviewee’s thoughts and feelings about the provider, (3) encounters leading to these recommendations, (4) specific circumstances, (5) how interviewees made their recommendation, and (6) reactions of listeners. As Miles and Huberman (1994) recommend, we coded verbatim transcripts of the interviews and field notes to identify recurring themes. For the purposes of this discussion, we refer to interviewees who offered advocacy as advocates.

Findings

First, advocacy involves a proactive initiation of discussions about the service, suggesting an active communication style. Advocates do not wait to be asked for advice but rather feel a motivation to mention the service provider proactively:

Any time the topic came up you would need to say it. You would be compelled in some way to add, “This was really good,” and not be able to keep quiet about it. (Interviewee 8)

Proactive promotion, especially if the conversation is about another topic, demonstrates stronger advocacy. Conversely, PWOM tends to be generated by direct inquiries from listeners (Mangold, Miller, and Brockway 1999).

Second, advocates defend the focal service provider against criticisms and dismiss competitors when faced with disagreement. For example, a British interviewee asserted:

Don’t ever go for BT [British Telecom]…. I think that this [smaller company] is the best provider, and they are much cheaper, and they can answer your calls at 11:30 at night. (Interviewee 11)

Third, advocates exhibit a judgmental, forceful communication style, such that they are explicit and enthusiastic, such as:

This guy is absolutely brilliant. I couldn’t recommend him more highly. Here is his email address; why don’t you get in touch with him? (Interviewee 1)

In contrast, the delivery of general PWOM communication tends to be more subdued:

There is a different excitement and energy about [advocacy] as opposed to when I used to recommend other companies in the past. It was—“Yeah they’re good, they did the job.” Whereas for this one it’s like—“It is superb.” (Interviewee 5)

Fourth, advocates recognize that they act as promoters for the service provider, urging listeners to use the provider’s service, following up, and even contacting the service provider on behalf of their friends to encourage a deal:

So, I rang up [the company] and said, “this is what my friend has been offered, what can you do?” (Interviewee 6)

Fifth, the certainty associated with advocacy stems from the advocate’s experience, such that advocates recognize their own conviction:

For someone to strongly recommend, they need to believe in it. They need to have experienced it.…That’s how strong I feel about it. (Interviewee 12)

In summary, this exploratory study suggests that advocacy has distinguishing characteristics, compared with more general PWOM. Therefore, we adopt the following definition of advocacy:

Advocacy is a customer’s ongoing, proactive, and voluntary promotion of a product or service compared with market alternatives through strong, passionate, and explicit recommendations, with the goal of positively influencing others’ views of the product or service.

Item Generation

We integrated these qualitative results with prior theoretical work to develop an initial set of 15 items that might represent the conceptual theme of advocacy. Four judges, academic experts in services, evaluated the face and content validities of these items. To confirm the items were representative of our conceptual domain, each judge received our newly developed definition of advocacy. Through a standard procedure (Zaichowsky 1985), the judges agreed that 11 items represented the domain of interest and suggested one further item, related to providing feedback about the service provider through blogs and review sites. These 12 items, as listed in Table 2, then were included in Study 2.

Table 2.

Advocacy Items, Rasch Scale Values, and Fit Statistics: Study 2.

Item #	Advocacy Item	Location on Rasch Scale	Item Fit	χ² Probability*
1	When I recommend this service provider, I always do so strongly (Score 6 or 7 out of 7).	−1.99	10.79	.21
2	I am enthusiastic in my recommendations of this service provider.	−1.26	18.41	.02
3	I have only good things to say about this service provider.	−1.05	8.94	.35
4	When discussing this service provider, I urge people to consider using it.	−0.90	23.40	.01
5	Whenever there is a conversation about this type of service, I usually strongly recommend this service provider, without being asked.	−0.83	13.00	.11
6	I would defend this service provider if people were to give negative comments about it directly to me.	−0.32	8.88	.35
7	I describe this service provider as the best of its kind.	−0.17	2.11	.01
8	I have told more people about my positive experience with this service provider than I have with most other service providers.	0.43	14.23	.08
9	When talking about this service provider, I usually compare it to its competitors, explaining why competitors are not as good.	0.46	5.07	.75
10	I take the initiative to actively promote this service provider (e.g., passing on details of this service provider, or contact the service provider on behalf of others if needed).	1.46	24.79	.01
11	Even when there is no conversation, but if I think there are people who have an interest in this type of service, I strongly recommend this service provider, without being asked.	1.51	13.00	.11
12	When practical, I provide positive written feedback on this service provider (e.g., recommendation blogs, ratings, and comments on review websites).	2.67	25.46	.01

Note. The shaded area, Items 10–12, represents “strong advocates” and the unshaded areas, Items 1–9, represents “moderate advocates.”

* After taking into account the Bonferroni adjustment, all probabilities were above the .01 level, showing a good fit of the individual items to the scale (Salzberger 2009).

Study 2: Measuring Advocacy

To determine whether some advocacy behaviors occur more readily than others, we develop a hierarchy of advocacy behaviors, as well as a new advocacy scale. For generalizability, we consider consumer services that involve varying levels of customer-employee contact and experience versus credence properties (Keh and Pang 2010). The extent of customer-employee contact is an important differentiator across service categories (Bowen 1990) and for consumer evaluations (Keh and Pang 2010). The service categories we consider span experience and high-contact services (hairdressing, beauty salon, and childcare), experience and low-/medium-contact services (telecommunication, basic banking, and dry cleaning), credence and high-contact services (doctors, dentists, and physiotherapists), and credence and low-/medium-contact services (financial advisors, car servicing, equipment repairs, and legal and accounting services). Respondents focused on a specific service provider, and we used quotas to ensure that the services reported were relatively evenly spread across the four categories.

Research Approach

This large-scale, quantitative study relied on a national online panel in Australia. Respondents, at least 18 years of age, were recruited on the basis of whether they had strongly recommended a service provider in the previous 6 months. Thus, we ensured a specific focus on strong recommenders, who are potential advocates, rather than consumers who provide more moderate PWOM. Using a 7-point screening question—“If you have recommended this service provider to others, how strongly was the recommendation expressed?”—we identified those who scored above the midpoint (>4). We contacted 6,937 respondents, of whom 4,235 agreed to participate and 1,045 passed the screening criteria. Data quality checks for completeness and quality (i.e., response patterns and completion speed) reduced the sample to 975 respondents. The sample constitution was as follows: gender (48% men, 52% women), age (18–29 years 23%, 30–39 years 22%, 40–49 years 24%, and 50 years+ 31%), income (below AUD30K 34%, AUD30K–AUD80K 48%, and AUD80K+ 18%), and education (year 12 or less 31%, technical/further education 28%, and degree 41%).

These respondents then evaluated the extent to which the 12 advocacy items from Study 1 described their past recommendation behaviors, using a 7-point Likert-type scale (1 = does not describe me at all, 7 = describes me very well). In return, respondents received a small cash credit, offered through the online panel’s standard procedure. Quotas helped ensure that the sample was representative of each Australian state. The mean scores revealed substantial variation in the extent to which respondents agreed about their own behaviors, which suggests initial support for our assertion that there are different levels or degrees of advocacy behaviors. For example, almost all respondents provided a strong recommendation (mean of 5.83 on a 7-point Likert-type scale), but “defending the service provider when hearing negative comments about it” was somewhat less common (5.27), and “posting feedback and comments on review sites and blogs” was among the least prevalent behaviors (3.27).

Next, in line with our objective of developing a hierarchy of behaviors, we applied Rasch modeling (Andrich 2011; for more detail, see the Online Appendix), the use of which is well established in marketing (e.g., Laurent, Kapferer, and Roussel 1995; Sweeney, Danaher, and McColl-Kennedy 2015). A Rasch model recognizes that items have different intensity levels, so each item in a scale represents a different level of the core construct—in this case, advocacy. Items on the higher end of the scale are more difficult to endorse or carry out, while items at the lower end are easier. An item’s intensity and the relative intensity of all items are central to capturing the construct. Therefore, the scale is defined in terms of its ordering of the constitutive items. Then, a Rasch approach orders respondents according to the same scale, according to their probability of agreeing or disagreeing with each item along the scale (Andrich 2011). Sufficient variation in the range of item endorsability is critical to ensure the Rasch Scale takes the full range of the focal construct into account.

Findings

We entered the 975 qualified responses into RUMM 2030 (Andrich, Sheridan, and Luo 2011), a widely used software package for Rasch analysis. Following Ewing, Salzberger, and Sinkovics (2005), we examine the threshold ordering for each item post hoc. Disordered thresholds arise if there are more response categories than respondents recognize. We iteratively collapsed adjacent categories of the 7-point Likert-type scales if we found any evidence of such disordering. Recoding was necessary for 10 of the 12 items. Some items could be collapsed into three to six categories, instead of seven, but for 2 items, we had to collapse the scores into just two coding categories: agree (scores of 1–4) or disagree (scores of 5–7). For clarity and measurement consistency, we recoded all items using this dichotomous coding, which also is consistent with Rasch measurement theory (e.g., Salzberger and Koller 2013).

Next, we divided the sample according to the four service categories to test for generalizability. We found very little difference in the fit of the Rasch model across these subsamples though, in terms of uniform or nonuniform differential item functioning (DIF). Thus, the groups were similar in both the patterns of their responses to items and the levels of these responses (p > .05 for all 12 items for both uniform and nonuniform DIF). The locations of the 12 items on the Rasch Scale also were very similar across the four subsamples, according to the Pearson correlations of pairs of subsamples (ranging from .91 to .97). The specific order of the scale items is virtually identical across subsamples, with a Spearman’s nonparametric correlation of .89 to .96. Thus, for our purposes, the operationalization of the Rasch Scale is equivalent for all four groups, so we combined the samples. The fit for this overall sample was good (item-trait interaction: χ² = 187.10, df = 96, p < .01, person separation index = .73, reliability = .84). Table 2 presents the ordering of the Advocacy Construct Scale items.

The Pearson item distribution on the Rasch Scale (see Figure A1 in the Online Appendix) confirms that the spread of items across the range of the construct is excellent, from −1.99 to 2.67 (M = 0, SD = 1.36). The respondents’ scores ranged from −3.53 to 3.83 (M = 1.21, SD = 1.77). The spread of the items thus is less than the spread of the sample, which is typical for Rasch measurement modeling (Salzberger and Koller 2013). The good fit of the scale, the lack of significant difference in fit across subsamples, and the spread of items support its psychometric soundness.

The order of the items in Table 2 further reveals that the lower end of the scale generally represents the strength and enthusiasm of the recommendation; in line with our proposed definition of advocacy, Items 1 and 2 suggest minimum requirements or hygiene factors for advocacy. A recommendation cannot constitute advocacy if it is subdued or lacks passion. The center of the scale includes strong service provider support, defense against criticisms, and comparisons with competitors to highlight the provider’s strengths (Items 6–8). At the upper end of the scale, Items 10–12 indicate increasingly proactive behaviors, such as recommending the provider even if the discussion is about other topics. The most extreme item, Item 12, refers to endorsing the service provider on blogs and forums.

Consistent with our definition, all advocates offer strong, passionate, and explicit recommendations (i.e., lower end of the scale). Moderate advocates also defend the provider against market alternatives (midsection of scale). Strong advocates conduct these same advocacy behaviors but also voluntarily promote the provider, even without prompting, and take the initiative, such as by posting positive comments on blogs or contacting the provider on behalf of others, as represented by upper end items of the scale. Even as we refer to weak, moderate, and strong advocates, we note that the Rasch Scale indicates a continuum of these degrees of advocacy.

Study 3: Advocacy Antecedents and Outcomes

With Study 3, we develop a conceptual framework and investigate what drives advocacy, as well as its valuable outcomes relative to the benefits of general PWOM. In line with our definition, advocacy characteristics are evident in the proposed measurement (Table 2). In contrast, general PWOM implies saying positive things about an entity, including recommendations. Our framework of the antecedents and outcomes of advocacy reflects the advocates’ perspective, which makes it helpful for designing effective advocacy programs. That is, we highlight valuable outcomes of advocacy compared with weaker WOM, then propose and test its antecedents using relevant theoretical models (Berger 2014; Stokburger-Sauer, Ratneshwar, and Sen 2012). We consider perceived service quality as a further brand-related antecedent (Roy 2015) in the conceptual framework in Figure 1. Details about the constructs, their measurement, and related items are available in the Online Appendix.

Figure 1.

Customer advocacy: A conceptual framework.

Outcomes

Our review of WOM literature reveals many benefits of verbal and online discussions about providers, but as a stronger, more passionate expression of WOM, we suggest that advocacy should have even more powerful and beneficial outcomes than PWOM. Our literature review and conceptual definition highlight the ongoing nature of advocacy; advocates express beliefs that become internalized over time, unlike the potentially external, reactive, and immediate form of PWOM behavior. According to Cialdini (2009), people prefer to behave consistently to enhance their self-perception and beliefs, so they actively work to be consistent with traits that they ascribe to themselves through their recent actions (e.g., “I did X, so I am helpful”), as well as with their previous behaviors (Cialdini and Goldstein 2004). The drive to be consistent is stronger and more persistent if the commitment is active; public WOM is a form of commitment or public stance (Garnefeld et al. 2013), and as we show in Studies 1 and 2, advocacy is a strong, active form of WOM. By voicing a particular view, people can align their attitudes and behaviors to reinforce and remain consistent with their stated opinion. Following Garnefeld, Helm, and Eggert (2011), consistency principles help explain why we expect stronger, more consistent beneficial outcomes for advocacy than for PWOM, as manifested in commitment and loyalty intentions. In addition, self-perception theory suggests that articulating a point of view provides personal insights about an attitude or belief (Bem 1965). By advocating an inner belief strongly, the advocate becomes even more firmly committed to the line of reasoning (Garnefeld, Helm, and Eggert 2011), reducing the likelihood that he or she will switch suppliers or complain about a service. These two theoretical perspectives suggest that stating an opinion passionately increases the person’s commitment and loyalty, such that we argue:

Hypothesis 1: Greater advocacy increases the likelihood of the following, compared with greater general PWOM: (a) commitment, (b) loyalty, and (c) willingness to pay a higher price, but it reduces the likelihood of (d) customer switching and complaining intentions.

Antecedents

We consider several potential antecedents and derive the final model by reviewing robust prior theoretical explanations but excluding conceptualizations that remain confusing.¹ In particular, our proposed model extends Stokburger-Sauer, Ratneshwar, and Sen’s (2012) consumer brand identification (CBI) framework. CBI refers to “a consumer’s perceived state of oneness with a brand” and identifying with the brand’s values (Stokburger-Sauer, Ratneshwar, and Sen 2012, p. 407). Identifying with the personal relevance of a brand helps clarify unique self-aspects and results in a positive sense of self-identification (Kunda 1999). Lam et al. (2010) similarly conceptualize CBI as a customer’s psychological state of perceiving, feeling, and valuing his or her sense of belonging with a brand. These higher order motivations are enduring and relate to powerful, intrinsically held beliefs. Brand identification thus should link to ongoing, passionate expressions of endorsement, and we predict:

Hypothesis 2a: Greater CBI is related to greater advocacy.

Hypothesis 2b: CBI has a stronger relationship with advocacy than with PWOM.

The concept of customer delight also may be relevant for understanding advocacy. Delight emerged in the late 1990s as a useful extension of mere satisfaction or service quality assessments, and higher levels evoke important behavioral consequences (Rust and Oliver 2000). Two views of delight emerge from prior literature. First, delight is a mix of emotion and cognition, such that it combines high pleasure with high arousal (Oliver 2010) and results from joy (Kim, Vogt, and Knutson 2015); in contrast, satisfaction is capable of producing pleasure but not arousal (Wang 2011). Second, delight might refer to an extremely positive emotional state that results when expectations are exceeded, such that it implies an extreme level of satisfaction. In this view, delight is a positive, nonlinear response to customer satisfaction, and the relationship of satisfaction with intentions increases beyond the upper limit of a zone of tolerance, into a “zone of delight.” Finn (2012) offers compelling empirical evidence that delight is distinct but simultaneous with satisfaction (see also Kim, Vogt, and Knutson 2015). When CBI is achieved by brands, because they have elicited feelings of warmth by establishing a central role in consumers’ autobiographic memories and self-relevant experiences (Stokburger-Sauer, Ratneshwar, and Sen 2012), due to the significant emotive component of CBI (Davvetas and Diamantopoulos 2017), we expect:

Hypothesis 3: Greater CBI is related to greater delight.

We further argue that service quality is an antecedent of both satisfaction and delight. Innumerable studies demonstrate the former link (e.g., Cronin and Taylor 1992), but we seek to establish that service quality is related to customer delight, especially if it is unexpected or surprising (Wang 2011). Magnini, Crotts, and Zehere (2011) identify a range of service quality elements linked to customer delight, such as the physical environment or staff behaviors. Accordingly, we hypothesize:

Hypothesis 4: Greater service quality is related to greater satisfaction.

Hypothesis 5: Greater service quality is related to greater delight.

Perceived service quality also may be an important antecedent of both advocacy and WOM. Service quality enhances PWOM behavior (e.g., Harrison-Walker 2001; Sweeney, Soutar, and Mazzarol 2012). A high level of service quality that exceeds expectations likely generates strong PWOM and even advocacy. Service quality is highly functional; it reflects a judgment about the superiority or excellence of the offering (Zeithaml 1988), in contrast with CBI, which refers to affective and evaluative aspects of identifying with a brand (Stokburger-Sauer, Ratneshwar, and Sen 2012). Lam et al. (2013) propose service quality as an antecedent of CBI, but the association is indirect and not particularly strong. Therefore, we argue instead that service quality is more likely to generate general PWOM, rather than specific, strong, passionate advocacy behavior. We model service quality as a separate antecedent of advocacy and include it in our CBI-advocacy model:

Hypothesis 6a: Greater service quality is related to greater advocacy.

Hypothesis 6b: Service quality has a stronger relationship with PWOM than with advocacy.

Satisfaction also relates to positive behavioral intentions, such as repurchasing, likelihood of revisiting, and WOM (Finn 2012), so customer satisfaction might generate both PWOM and advocacy. The scope of satisfaction is relatively limited, compared with delight, in that it incurs positive affect but not arousal or joy (Wang 2011). Using our conceptualization of advocacy (Table 2), we predict that satisfaction has less explanatory power with respect to advocacy than it does for WOM. Therefore,

Hypothesis 7a: Greater satisfaction is related to greater advocacy.

Hypothesis 7b: Satisfaction has a stronger relationship with PWOM than with advocacy.

If we extend these arguments with respect to advocacy and recognize that delight results from the interrelated components of arousal and positive affect, we can also predict:

Hypothesis 8a: Greater delight is related to greater advocacy.

Hypothesis 8b: Delight has a stronger relationship with advocacy than with PWOM.

Finally, we consider interpersonal communication. Berger (2014) identifies five transmitter functions associated with WOM communication: impression management, persuasion of others, emotion regulation, social bonding, and information acquisition. Impression management is motivated by a desire to demonstrate an aspired self to others, and it relates to persuasion of others (Petty, Wheeler, and Tormala 2003), social sharing of emotions (Rimé 2009), bonding with others (Rimé 2007), and seeking information to confirm an outside perspective (Fitzsimons and Lehmann 2004). These fundamental human motives suggest the generation of deeply held, ongoing, emotive expressions of endorsement (i.e., advocacy). We also argue that they may be more associated with advocacy than with less deeply held, less internalized, more reactive, and immediate PWOM behavior. That is:

Hypothesis 9a: Greater impression management, emotion regulation, information acquisition, social bonding, or dispositions to persuade others are related to greater advocacy.

Hypothesis 9b: Impression management, emotion regulation, information acquisition, social bonding, and the disposition to persuade others have stronger relationships with advocacy than with PWOM.

The conceptual model that summarizes these hypotheses is in Figure 1.

Research Sample and Method

To test these hypotheses, we surveyed respondents available through an Australian online national panel. We focused on high-contact services in this study to minimize the impact of the range of industry types on antecedents of WOM (Harrison-Walker 2001). To ensure that the sample included a range of typical consumers, we initially imposed no restriction with respect to whether they gave WOM or not or on recommendation strength. However, we ensured that the final sample included at least 500 respondents who had given PWOM in the past 6 months about the service they had selected, as well as a comparative sample of at least 100 respondents who had not.

We then applied criteria to exclude questionable responses and ensure data quality. We paid particular attention to excluding respondents who took insufficient time to complete the survey and those who flat-lined, which removed approximately 20% of the fully completed questionnaires. The final sample of 621 participants represented a response rate of close to 20%. The respondent profile was very similar to that of Study 2.² Participants who had given some WOM (n = 504) reported on their general PWOM and advocacy behaviors with respect to a selected service provider, using the measures from Study 2.

They also answered questions related to the predicted antecedents and outcomes, using established scales (see Online Appendix Tables A2–A4). The outcome measure items were clearly specified as outcomes of giving WOM, with wording such as “Giving word of mouth about [name of service] makes me more or less likely to….”³ To measure outcomes, we used a 7-point scale, where 1 represents less likely than before and 7 represents more likely than before, unless indicated otherwise in Online Appendix Table A4, where we provide the full details of each construct measurement.

For the outcomes, we tested Hypothesis 1 using the mean or median difference across outcomes; with regard to the antecedents, as predicted in Hypotheses 2–9, we used structural equation models for two different frameworks. To compare the different paths from CBI, service quality, satisfaction, or delight to advocacy, relative to the parallel paths from these antecedents to PWOM (Hypotheses 2b, 6b, 7b, 8b, and 9b), we used Satterthwaite’s path comparison test (Kock et al. 2006). We repeated the Rasch scaling procedure from Study 2 for the Study 3 advocacy items and obtained nearly identical results in terms of model fit and the order of the scale items, as we detail in the Online Appendix.

Results: Outcomes

Because we use two models to test Hypotheses 2–9, and the outcomes would be common to both models of advocacy, we decided to examine the outcomes according to a common analysis, across different levels of advocacy/PWOM. Testing the effects of different levels of advocacy/PWOM on outcomes is especially insightful for practice, and the outcomes predicted in Hypothesis 1, related to commitment and loyalty, also are of interest for examining the implications of advocacy. For these reasons, as well as for parsimony, we tested each outcome separately using simple analyses of variance. To start, for the test of Hypothesis 1, we divided the Study 3 sample into four groups:

Strong advocates (n = 204) who exhibit high individual Rasch scores on items that represent proactivity (Items 10–12 shown in Table 2), as well as Items 1–9, by implication of the Rasch analysis. These respondents scored at least 1.09 on the Rasch Scale, equivalent to the lowest location of the three most challenging items in the Study 3 data set. These strong, passionate advocates consistently strive to go out of their way to promote the service voluntarily and over market alternatives.

Moderate advocates (n = 208) whose scores fall below 1.09 but above −1.45 (which pertains to Item 1, the lowest advocacy item location in the Study 3 data set). That is, they score high on the first 9 items but not Items 10–12. They offer consistently strong descriptions of and support for the provider, but unlike strong advocates, they do not go the extra mile.

PWOMers are nonadvocates who give PWOM (n = 92), with Rasch scores less than −1.45. They have offered PWOM about the service provider in the past 6 months, but they are unlikely to agree with any of the 12 advocacy items, including giving strong recommendations. Their WOM is not powerful.

Non-PWOMers, neither advocates nor PWOM givers (n = 117), they have not given WOM about the selected provider in the past 6 months.

The findings suggest that advocates are significantly more valuable than those who give general PWOM (PWOMers) or do not give PWOM. Moreover, strong advocates are significantly more valuable than moderate advocates, and the latter are significantly more valuable than PWOMers or non-PWOMers. Critically, strong advocates are significantly more committed to the service provider (5.08), loyal (5.46), and willing to pay a higher price (4.35) and less likely to complain or switch (2.97) than moderate advocates (scores of 4.47, 4.76, 3.79, and 3.35, respectively) and PWOMers (scores of 4.27, 4.45, 3.57, and 3.45, respectively), in support of Hypothesis 1 (Table 3). The data from our sample also suggest that strong advocates spend significantly more with the service provider per year (AUD500 vs. AUD385 moderate advocates, AUD325 PWOMers, and AUD245 non-PWOMers; see Online Appendix Table A1). Moderate advocates are more valuable than PWOMers, such that they spend more with the service provider annually, offer more PWOM to others, and exhibit more positive changes in loyalty and commitment. These findings highlight the power of advocates, particularly strong advocates, compared with those giving general PWOM or no PWOM.

Table 3.

Differences in Relationships of Advocacy and WOM With Outcomes (Hypothesis 1).

Likely change in behavioral intentions, as a result of giving WOM^b	Group 1	Group 2	Group 3	Group 4
	Strong Advocates^a	Moderate Advocates	PWOM	No PWOM	Overall
	(n = 204)	(n = 208)	(n = 92)	(n = 117)	(n = 621)	Significance	Significance Across Pairs
Likely change in intended loyalty	5.46	4.76	4.45	NA	4.99	F = 31.30 (p < .01)	All pairs p < .01, except 2_3 (p < .05)
Likely change in intentions to pay more	4.35	3.79	3.57	NA	3.97	F = 14.35 (p < .01)	All pairs p < .01, except 2_3 n.s.
Likely change in intended external response (e.g., complain)	2.97	3.35	3.45	NA	3.21	F = 6.56 (p < .01)	All pairs p < .01, except 2_ 3 n.s.
Likely change in commitment, as a result of giving WOM	5.08	4.47	4.27	NA	4.68	F = 27.93 (p < .01)	All pairs p < .01, except 2_3 n.s.

Note. Measures were taken from Zeithaml, Berry, and Parasuraman (1996), with the exception of commitment which was a single item generated for the study. n.s. = not significant; WOM = word of mouth; PWOM = positive word of mouth; NA = not applicable.

^a Strong (above or equal to 1.09 on Advocacy Scale); moderate (less than 1.09 on Advocacy Scale but equal to or above lowest point of −1.45); PWOM (less than −1.45 on Advocacy Scale, unlikely to agree with any advocacy items in Table 2); and “no PWOM (have not given PWOM about <NAME> either verbally or online to others in the past 6 months?” ^bScales: 1 = less likely, 7 = more likely.

Results: Antecedents

As noted, our tests of Hypotheses 2–9 rely on both a brand-focused model (e.g., Stokburger-Sauer, Ratneshwar, and Sen 2012) and a model that incorporates psychological motivations for WOM (Berger 2014). In each case, we first check the reliability and discriminant validity of the constructs, before determining the appropriate structural model. Some constructs did not achieve discriminant validity and thus were combined, such as brand prestige and distinctiveness in the case of the CBI model. By combining strongly related constructs, we ultimately modeled brand prestige/distinctiveness, brand social benefits, and memorable brand experiences as antecedents of CBI (Stokburger-Sauer, Ratneshwar, and Sen 2012). In this model, which includes advocacy, the maximum correlation (.80) arises between service quality and satisfaction. As we show in the Online Appendix (Table A2), the squared correlations are always less than the average variance extracted for each construct, so for the CBI model, we obtain evidence of discriminant validity (Fornell and Larcker 1981).

In the motivations model (Berger 2014), we found close interrelations of opinion leadership, extroversion, altruism, information acquisition, and disposition to persuade, so we included them as first-order constructs within the second-order reflective construct, “desire to influence.” Desire to influence, social bonding, social risk, and subjective norms, together with advocacy, all demonstrate discriminant validity (Fornell and Larcker 1981; see Table A3 in the Online Appendix). The highest correlation of pairs of constructs in this model was .78, between desire to influence and advocacy. The constructs in the two modeling approaches also indicated strong construct reliabilities, ranging from .75 to .95.

As explained in the Online Appendix, we undertook common method bias tests by introducing a common method factor into each structural equation model (Podsakoff et al. 2003; Stokburger-Sauer, Ratneshwar, and Sen 2012). We found no significant differences in the model paths when we added the common method factor, so common method bias does not appear to be an issue. We also tested the reliability of the Advocacy Scale in Study 3, by comparing it with the Study 2 scale.

Among the brand and managerial factors as antecedents, we find that CBI relates significantly and positively to advocacy (.19, p < .01), in a stronger relationship than with PWOM (.13, p < .01; t _diff = 1.83, p < .05), in support of Hypotheses 2a and 2b (Table 4). Furthermore, CBI relates significantly and positively to delight (.61, p < .01), as we predicted in Hypothesis 3. The paths from the various brand and social benefit antecedents to CBI do not vary significantly across the advocacy and PWOM models, suggesting the equivalent structure of CBI across both two models. Service quality relates significantly to satisfaction (.84, p < .01), delight (.26, p < .01), and advocacy (.38, p < .01), in support of Hypotheses 4, 5, and 6a, respectively. Although service quality exhibits a stronger relationship with PWOM (.51, p < .01) than with advocacy (.38, p < .01), the difference is not significant (t _diff = −0.85, n.s.), so we cannot confirm Hypothesis 6b. These results suggest that advocacy behavior is more strongly linked to internalization, in terms of identification with the brand, than general PWOM is. This latter behavior instead appears more associated with cognitive, readily evaluated, and perceived service quality. In this sense, advocacy emerges as a long-term, persistent behavior that reinforces publicly articulated views, whereas PWOM may be more short term and spontaneous.

Table 4.

Brand- and Product-Related Factors, Including CBI and Advocacy.

Path	Hypothesis	Advocacy (Standardized Path Coefficients, Standard Error)	WOM (Standardized Path Coefficients, Standard Error)	Path Comparison (Hypotheses 2b, 6b, 7b, and 8b Comparing Paths)
Brand prestige and distinctiveness to CBI		.33	.34	n.s.
Brand social benefits to CBI		.68	.68	n.s.
Memorable brand experiences to CBI		.04, n.s.	.04 n.s.	n.s.
CBI to advocacy	H2a	.19	.13	H2b (1.83, p < .05)^a
CBI to delight	H3	.61	.61	n.s.
Service quality to satisfaction	H4	.84	.84	n.s.
Service quality to delight	H5	.26	.26	n.s.
Service quality to advocacy	H6a	.38	.51	H6b (−0.85, n.s.)^a
Satisfaction to advocacy	H7a	−.04, n.s.	.13 (p < .05)	H7b (−1.11, n.s)^a
Delight to advocacy	H8a	.27	.07, n.s.	H8b (1.85, p < .05)^a
SMC (advocacy)		.49	.57
Total effect of CBI on advocacy^b		.38	.13

Note. All paths significant at p < .01 unless otherwise indicated. With measures taken from Stokburger-Sauer, Ratneshwar, and Sen (2012). Satisfaction and delight measures were taken from Oliver (2010). n.s. = not significant =; SMC = squared multiple correlation; WOM = word of mouth; CBI = consumer brand identification.

^a Path comparison uses Satterthwaite’s t statistic (Kock et al. 2006). ^bIncludes direct and also indirect effect via delight.

Finally, we tested the relationship of satisfaction and delight with advocacy finding that delight had a significant positive relationship with advocacy (.27, p < .01), while satisfaction did not (−.04, n.s.) confirming Hypothesis 8a, but rejecting Hypothesis 7a. In contrast, satisfaction is more strongly related to PWOM (.13, p < .05) than to advocacy (−.04, n.s.), though this difference is not significant (t _diff = −1.11, n.s.). We thus reject Hypothesis 7b. The results are reversed for delight, which indicates a significantly stronger link to advocacy (.27, p < .01) than to PWOM (.07, n.s.; t _diff = 1.85, p < .05), in support of Hypothesis 8b.

Next, when we consider interpersonal behavioral motivations as antecedents, using Berger’s (2014) conceptual model, we obtain the results contained in Table 5, in relation to Hypothesis 9. The paths and weights of the reflective first-order factors do not vary across the advocacy and PWOM models, in support of the equivalence of the second-order desire to influence construct (Table 5). The desire to influence (.77, p < .01), social bonding (.58, p < .01), and subjective norms (.45, p < .01) all relate positively to advocacy; social risk (−.05, n.s.) does not. The empirically determined dimensionality of Berger’s motivational constructs differs in structure from the conceptual list of motivations, yet the final constructs collectively represent all five of the original conceptual dimensions. This evidence offers partial support for Hypothesis 9. Advocacy also is significantly more strongly related, compared with general PWOM, to the desire to influence (.77, p < .01; .67, p < .01; t _diff = 4.23, p < .01) and social bonding (.58, p < .01; .47, p < .01; t _diff = 4.81, p < .01), but we find no difference in the relationships of advocacy versus PWOM with subjective norms (.45, p < .01; .47, p < .01; t _diff = −0.13, n.s.). Finally, social risk is significantly more negatively related to PWOM (−.26, p < .01) than to advocacy (−.05, n.s.; t _diff = 2.07, p < .05).

Table 5.

Motivational Antecedents of Advocacy.

Path	Model
Path	Advocacy (Standardized Path Coefficients; Hypothesis 9a)	WOM (Standardized Path Coefficients)	Path Comparison^a (Hypothesis 9b)
Desire to influence to disposition to persuade^b	.87	.89	n.s.
Desire to influence to opinion leadership^b	.91	.87	n.s.
Desire to influence to extroversion–altruism^b	.93	.95	n.s.
Desire to influence to information acquisition^b	.74	.70	n.s.
Desire to influence to advocacy	.77	.67	4.23^a
SMC (advocacy)	.59	.46
Social bonding to advocacy	.58	.47	4.81^a
SMC (advocacy)	.34	.21
Social risk to advocacy	−.05, n.s.	−.26	2.07^a
SMC (advocacy)	.01	.07
Subjective norms to advocacy	.45	.47	−.03^a, n.s.
SMC (sdvocacy)	.20	.22

Note. All values are significant at p < .01 unless otherwise stated. Measures were taken from Flynn, Goldsmith, and Eastman (1996); Eisingerich et al. (2015); Hennig-Thurau et al. (2004); Perugini and Bagozzi (2001); Berger (2014); Bennett (1988); and Sweeney et al. (2014). n.s. = not significant; SMC = squared multiple correlation; WOM = word of mouth.

^a Satterthwaite’s t statistic (Kock et al. 2006). ^bDesire to influence is a second-order construct comprising the following first-order constructs: disposition to persuade, opinion leadership, extroversion–altruism, and information acquisition.

Discussion

Despite widespread acknowledgment of the benefits of advocacy for service providers, a clear conceptualization of advocacy and its distinguishing components has remained lacking. Scholars suggest that advocacy invokes beneficial performance outcomes, but surprisingly little investigation focuses on this important concept. We distinguish advocacy from general PWOM and determine specific behavioral components of advocacy. Poor prior conceptualizations have led to confusion and a lack of scholarly studies; with this empirical examination of advocacy, we seek to establish a clear definition of the concept and develop a Robust Scale to measure and identify its strengths. This scale sets out a hierarchy of behaviors that represent increasing levels of advocacy; it is not a simple, dichotomous concept. Our work in turn reveals that the strength of advocacy relates to important outcomes, such as increased commitment and willingness to pay, highlighting the significance of strong advocates. Developing strong advocates has especially important benefits for firms. Finally, we identify antecedents of advocacy and how they differ in their relationships with advocacy and PWOM.

Theoretical Contributions

In three related studies, we contribute to broader WOM literature. First, we examine the concept of advocacy, identifying it as a distinctive form of PWOM, and provide a concise definition of this important construct. It is more forceful, enduring, and relational than PWOM, motivated by a strong sense of internalized identity with the brand, such that it results in persuasive recommendations.

Second, we conceptualize advocacy across multiple levels. The hierarchy of advocacy behaviors that we propose reflects these various levels of intensity and effort. Third, we develop a robust and reliable advocacy scale. A Rasch modeling approach acknowledges the different degrees of advocacy behaviors, rather than scoring different levels of agreement with a single item, as is the case for the criticized recommendation intentions item used in the NPS. As we discuss further in the next section, our study is the first to recognize that distinct types of behavior reflect different levels of advocacy. A strong recommendation is a minimum behavior of an advocate, but stronger advocates also defend the service against market alternatives, and the strongest ones go to considerable lengths to promote the service proactively, such as taking the initiative to recommend or provide positive written feedback online.

Fourth, we use two approaches to identify important antecedents, including both brand and interpersonal measures such as CBI, service quality, impression management, emotion regulation, desire to influence others, information acquisition, and social bonding. Satisfaction and delight appear as mediators between brand-related variables and advocacy. In the brand-focused model, service quality as a traditional antecedent of WOM is strongly associated with advocacy, but advocacy also appears far more closely linked to the richer constructs of CBI and delight than is PWOM (Table 4), in support of our proposal of a richer, deeper conceptualization of advocacy. With our second model, we find that two of the four motivational antecedents of advocacy proposed by Berger (2014), desire to influence and social bonding, have stronger relationships with advocacy than with PWOM. Subjective norms also are equally linked to advocacy and PWOM. In contrast, social risk appears to reduce PWOM behavior but has no evident link to advocacy behavior (Table 5). Moreover, within the CBI model, the interplays of satisfaction, delight, PWOM, and advocacy suggest that advocacy and delight are more strongly linked, whereas PWOM is more strongly related to satisfaction. These findings match prior distinctions of delight versus satisfaction (Finn 2012) and suggest that advocacy results from deeply embedded emotions. We therefore conclude that advocacy, as conceptually and operationally defined herein, is embedded within a network of well-established psychological and consumption constructs that explain outcomes beyond satisfaction and PWOM.

Managerial Contributions

We reveal some critical managerial outcomes of advocacy that emphasize the benefits of identifying advocates. Empirically, we confirm differences among strong advocates, moderate advocates, PWOMers, and non-PWOMers and demonstrate the power of advocacy and the need to differentiate groups of advocates and PWOMers. First, strong advocates are significantly more valuable than moderate advocates who are in turn more valuable than those who give general PWOM or no PWOM. Using our 12-item Advocacy Scale, managers might identify which customers represent the best opportunities and prioritize those who require special investments, due to their advocacy potential. Incorporating this scale into customer profiling efforts could lead to new insights for tailored relationship marketing strategies and help determine where to focus resources to leverage potentially lucrative relationships. A shorter, managerially oriented scale also might be developed, for example, a 4-item scale that spans the overall construct’s conceptual and operational breadth might suffice, if the items represented the full range of the scale. We suggest Items 1, 5, 9, and 12 from Table 2; by applying a scale with these items, managers would lose little summary information. The correlation of this 4-item scale with the 12-item scale reaches .87 with our data. Although the full scale is useful to conceptualize the construct, we encourage efforts to test a short form, especially given criticisms of the psychometric properties of the NPS (Keiningham et al. 2007) and the robustness, depth, and breadth of our advocacy measure, developed through Rasch modeling. A short form might offer a good alternative to the NPS in terms of predicting organizational outcomes such as churn and changes in revenue (e.g., de Haan, Verhoef, and Wiesel 2015).

Second, by acknowledging the distinction between strong advocacy and PWOM, managers can avoid the risk of relaxing their efforts once they evoke PWOM. Passionate, strong promotion by advocates is a better representation of customers’ true support for a firm. Firms should not be satisfied with moderate levels; they should target moderate advocates to move them up the scale by appealing to their intrinsically held desires.

Third, strong advocates are more likely to be more loyal and committed after they issue their advocacy, so firms must identify them and develop these beneficial relationships as a priority. Moreover, in this study, strong advocates spend more than moderate advocates and PWOMers, while also sharing more PWOM with others than moderate advocates do. This finding is consistent with the idea of direct (e.g., purchasing) and indirect (e.g., referrals) behavioral forms of engagement that result from a service experience (e.g., Kumar et al. 2019). The implications that would arise regarding how spending relates to different levels of advocacy suggest the need for further work to test the validity of this finding.

Fourth, managers should work to motivate the antecedents of customer advocacy. In particular, our findings reveal that advocacy is associated with brand identification, derived from consumers’ bonds with brand values. We also find that advocacy is enduring and motivated by a desire to connect with others, as well as a desire to exert influence, consistent with our definition. Advocates tend to identify and then adopt brand values so strongly that they express their endorsement passionately and consistently. Managers should develop brand and advocacy programs that nurture connectedness with the offering, communicating how a brand relates to an individual and is responsive to her or his particular needs. Programs that inspire emotional connectedness can nurture such social bonding and identification. Because advocates are motivated by their desire to influence others, managers could provide a ready channel for passionate endorsements, engaging them empathetically while encouraging passionate dialogue. Advertising messages also should present brands as personal, authentic, and relevant, to prompt customers to internalize their emotional bonding. For example, Apple provides advocacy programs that encourage customers to connect and internalize brand values (Cant, Machado, and Seaborne 2014), consistent with our empirical recommendations.

Fifth, managers need to understand the potential of advocacy-generating tools. Limited research addresses this need, though one study (Marsden, Samson, and Upton 2005) identifies eight tools: advocacy tracking, influencer outreach, causal campaigns, brand ambassador campaigns, empowered involvement, tryvertising, innovation, and referral programs. We recommend caution with regard to referral programs though. A recent review indicates that they may appear partial or suggest ulterior motives (Wirtz, Tang, and Georgi 2019). Despite such concerns, Dropbox’s referral program offers a successful example; it increased registered users from 100,000 to 4 million within 4 months, suggesting potential opportunities of carefully designed and implemented referral programs (Williams 2017).

Sixth and finally, managers should actively seek to benchmark best practices. The benefits of developing programs that motivate brand and social connectedness are likely to relate positively to firm’s profitability. Advocacy program designers should acknowledge the strong desire of advocates to relate to others, through both face-to-face and online interactions, and recognize the value of carefully constructed initiatives that nurture customer communities and user groups. Our study also suggests the need to develop initiatives to address relational value, which has implications for the allocation of marketing resources, including focused programs aimed at developing strong advocacy relationships.

Limitations and Further Research

With this initial study, we uncover some significant research opportunities that also might address some limitations of our work. In particular, we purposefully limited our study context to consumer services, for which communication between frontline employees and customers is especially important. Further research could explore other contexts, such as those in which face-to-face contacts are less frequent and when the object is a service employee versus the overall service firm. Personal loyalty to a salesperson can enhance customer loyalty to the business (Bove and Johnson 2006), but not necessarily (Chow and Holden 1997). Additional work is needed to clarify customer advocacy behaviors toward a service employee relative to a service business.

We also found that measurement and structural models related to the antecedents of advocacy do not differ across participants who primarily gave online PWOM and those who gave mainly off-line PWOM (see the Online Appendix). The antecedents may be relevant across both channels, though Study 3 indicates that most PWOM occurs off-line. Engagementlabs (2018) reports that 70% of WOM conversations, in terms of number of WOM messages, occur face-to-face, and the rest are online. However, the mean scores of participants who share information offline and online differ in the level of advocacy (.55 vs. 1.01 mean Rasch score, p < .05). The relative size of the group of online recommenders may be smaller (87% vs. 13% in this study), but they are proportionally more likely to be advocates, suggesting the importance of investigating advocacy in online contexts further. Content analyses of online customer reviews could offer useful insights. We also note that our proposed Advocacy Scale is relevant and robust for both online and off-line PWOMers.

To extend our work further, researchers also might investigate the outcomes of advocacy more deeply to reveal additional insight about which behaviors are most beneficial. For example, commitment arguably consists of five components, each with distinct outcomes (Keiningham et al. 2015). Commitment is an outcome in our model; identifying how advocacy relates to each form of commitment and then to subsequent outcomes is beyond the scope of our study but provides an interesting topic for further investigation. We also suggest examining how advocacy and its outcomes, such as increased commitment, iteratively enhance brand perceptions and transmitter functions (e.g., desire to influence, social bonding) in future periods. These functions in turn should affect future advocacy. Such investigations will require longitudinal studies. Another option would be to extend our research to consider effects of advocacy on receivers. Customers recruited through WOM may be more likely to offer WOM themselves (Wangenheim and Bayón 2004); continued research might investigate how this tendency plays out among advocates. Do customers recruited through advocacy tend to become advocates themselves? Examining the extent to which strong advocates for one service organization also might be strong advocates for other service organizations would be insightful. Finally, we view advocacy as a distinct and powerful form of PWOM; it also seems natural to consider the effect of an even more extreme form, namely customer evangelism. Evangelism may be too “over the top,” such that it could evoke negative responses or outcomes (e.g., Sweeney, Soutar, and Mazzarol 2008). However, anecdotal evidence indicates that firms seek evangelists (e.g., Kawasaki 1992), so investigating this consumer group could represent a useful extension to the current study.

Supplemental Material

Supplemental Material, JSR-17-028_R3_EXECUTIVE_SUMMARY - Customer Advocacy: A Distinctive Form of Word of Mouth

Supplemental Material, JSR-17-028_R3_EXECUTIVE_SUMMARY for Customer Advocacy: A Distinctive Form of Word of Mouth by Jillian Sweeney, Adrian Payne, Pennie Frow and Dan Liu in Journal of Service Research

Supplemental Material

Supplemental Material, JSR-17-028_R3_Online_Appendix_Final_9-12-2019 - Customer Advocacy: A Distinctive Form of Word of Mouth

Supplemental Material, JSR-17-028_R3_Online_Appendix_Final_9-12-2019 for Customer Advocacy: A Distinctive Form of Word of Mouth by Jillian Sweeney, Adrian Payne, Pennie Frow and Dan Liu in Journal of Service Research

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Pennie Frow

Supplemental Material

Supplemental material for this article is available online.

Notes

References

Andrich

David

(2011), “Rating Scales and Rasch Measurement,” Expert Review of Pharmacoeconomics and Outcomes Research, 11 (5), 571–585.

Andrich

David

Sheridan

Barry

Guanzhong

Luo

(2011), RUMM2030 Software and Manuals. University of Western Australia. http://www.rummlab.com.au (accessed April 13, 2015).

Balter

Dave

(2008), The Word of Mouth Manual, Vol. 2 . Boston, MA: Bzz Pubs.

Bansal

Harvir

Voyer

Peter

(2000), “Word of Mouth Processes within a Services Purchase Decision Context,” Journal of Service Research, 3 (2), 166–177.

Bem

Daryl

(1965), “An Experimental Analysis of Self-Persuasion,” Journal of Experimental Social Psychology, 1 (3), 199–218.

Bendapudi

Neeli

Berry

Leonard L.

(1997), “Customers’ Motivations for Maintaining Relationships with Service Providers,” Journal of Retailing, 73 (1), 15–37.

Bennett

Joel B.

(1988), “Power and Influence as Distinct Personality Traits: Development and Validation of a Psychometric Measure,” Journal of Research in Personality, 22 (3), 361–394.

Berger

Jonah

(2014), “Word of Mouth and Interpersonal Communication: A Review and Directions for Future Research,” Journal of Consumer Psychology, 24 (4), 586–607.

Bettencourt

Lance A.

(1997), “Customer Voluntary Performance: Customers as Partners in Service Delivery,” Journal of Retailing, 73 (3), 383–406.

10.

Bhati

Ragini

Verma

Harsh V.

(2017), “Advocacy, Customer Advocacy and Marketing Implications,” Journal of Management and Research, 2 (1), 29–43.

11.

Bloch

Peter H.

Richins

Marsha L.

(1983), “A Theoretical Model for the Study of Product Importance Perceptions,” Journal of Marketing, 47 (3), 69–81.

12.

Bove

Liliana L.

Johnson

Lester W.

(2006), “Customer Loyalty to One Service Worker: Should It Be Discouraged?” International Journal of Research in Marketing, 23 (1), 79–91.

13.

Bowen

John

(1990), “Development of a Taxonomy of Services to Gain Strategic Marketing Insights,” Journal of the Academy of Marketing Science, 18 (1), 43–49.

14.

Brown

Tom J.

Barry

Thomas E.

Dacin

Peter A.

Gunst

Richard F.

(2005), “Spreading the Word,” Journal of the Academy of Marketing Science, 33 (2), 123–138.

15.

Cant

Michael C.

Machado

Ricardo

Seaborne

Helen C.

(2014), “Brand Advocates—An Apple Phenomenon?” Corporate Ownership and Control, 11 (2), 535–541.

16.

Chow

Simeon

Holden

Reed

(1997), “Toward an Understanding of Loyalty: The Moderating Role of Trust,” Journal of Managerial Issues, 9 (3), 275–298.

17.

Christopher

Martin

Payne

Adrian

Ballantyne

David

(1991), Relationship Marketing: Together. Oxford, England: Butterworth Heinemann.

18.

Cialdini

Robert B.

(2009), Influence: Science and Practice. Boston, MA: Pearson.

19.

Cialdini

Robert B.

Goldstein

Noah J.

(2004), “Social Influence: Compliance and Conformity,” Annual Review of Psychology, 55, 591–621.

20.

Collins

Nathalie

Murphy

Jamie

(2009), “Customer Evangelists: A Conceptual Model,” Paper presented at ANZMAC Conference 2009, November–December, Melbourne, Australia.

21.

Cronin

J. Joseph

Jr Taylor

Steven A.

(1992). “Measuring Service Quality: A Reexamination and Extension,” Journal of Marketing, 56 (3), 55–68.

22.

Cross

Richard

Smith

Janet

(1996), Customer Bonding. Lincolnwood, IL: NTC Books.

23.

Davvetas

Vasileios

Diamantopoulos

Adamantios

(2017), “Regretting your Brand-Self? The Moderating Role of Consumer-Brand Identification on Consumer Responses to Purchase Regret,” Journal of Business Research, 80, 218–227.

24.

de Haan

Evert

Verhoef

Peter C.

Wiesel

Thorsten

(2015), “The Predictive Ability of Different Customer Feedback Metrics for Retention,” International Journal of Research in Marketing, 32 (2), 195–206.

25.

Doss

Samuel K.

Carstens

Deborah S.

(2014), “Big Five Personality Traits and Brand Evangelism,” International Journal of Marketing Studies, 6 (3), 13–22.

26.

Dye

Renee

(2000), “The Buzz on Buzz,” Harvard Business Review, 78 (6), 139–146.

27.

East

Robert

Hammond

Kathy

Wright

Malcolm

(2007), “The Relative Incidence of Positive and Negative Word of Mouth: A Multi-Category Study,” International Journal of Research in Marketing, 24 (2), 175–184.

28.

Eisingerich

Andreas

Eun

Helen Hae

Liu

Yeyi

Jia

Bell

Simon

(2015), “Why Recommend a Brand Face-to-Face but Not on Facebook?” Journal of Consumer Psychology, 25 (1), 120–128.

29.

Engagementlabs (2018), “Total Social: Make Every Conversation Count,” Engagementlabs Report, 16 March, https://blog.engagementlabs.com/totalsocial-make-every-conversation-count. (accessed 27 August 2019)

30.

Esquivias

Pedro

Knox

Steve

Sánchez-Rodríguez

Victor

Visser

Jody

(2013), Fueling Growth through Word of Mouth. Boston, MA: Boston Consulting Group.

31.

Ewing

Michael T.

Salzberger

Thomas

Sinkovics

Rudolf R.

(2005), “An Alternate Approach to Assessing Cross-Cultural Measurement Equivalence in Advertising Research,” Journal of Advertising, 34 (1), 17–36.

32.

Feick

Lawrence F.

Price

Linda L.

(1987), “The Market Maven: A Diffuser of Marketplace Information,” Journal of Marketing, 51 (1), 83–97.

33.

Finn

Adam

(2012), “Customer Delight: Distinct Construct or Zone of Nonlinear Response to Customer Satisfaction?” Journal of Service Research, 15 (1), 99–110.

34.

Fitzsimons

Gavan J.

Lehmann

Donald R.

(2004), “Reactance to Recommendations: When Unsolicited Advice Yields Contrary Responses,” Marketing Science, 23 (1), 82–94.

35.

Flynn

Leisa Reinecke

Goldsmith

Ronald

Eastman

Jacqueline

(1996), “Opinion Leaders and Opinion Seekers,” Journal of the Academy of Marketing Science, 24 (2), 137–147.

36.

Fornell

Claes

Larcker

David F.

(1981), “Evaluating Structural Equation Models with Unobservable Variables and Measurement Error,” Journal of Marketing Research, 18 (1), 39–50.

37.

Fullerton

Gordon

(2005). “The Impact of Brand Commitment on Loyalty to Retail Service Brands,” Canadian Journal of Administrative Sciences, 22 (2), 97–110.

38.

Fullerton

Gordon

(2011), “Creating Advocates: The Roles of Satisfaction, Trust and Commitment,” Journal of Retailing and Consumer Services, 18 (1), 92–100.

39.

Garnefeld

Ina

Eggert

Andreas

Helm

Sabrina V.

Tax

Stephen S.

(2013), “Growing Existing Customers’ Revenue Streams through Customer Referral Programs,” Journal of Marketing, 77 (4), 17–32.

40.

Garnefeld

Ina

Helm

Sabrina

Eggert

Andreas

(2011). “Walk Your Talk: An Experimental Investigation of the Relationship between Word of Mouth and Communicators’ Loyalty,” Journal of Service Research, 14 (1), 93–107.

41.

Harmeling

Colleen M.

Moffett

Jordan W.

Arnold

Mark J.

Carlson

Brad D.

(2017), “Toward a Theory of Customer Engagement Marketing.” Journal of the Academy of Marketing Science, 45 (3), 312–335.

42.

Harrison-Walker

L. Jean

(2001), “The Measurement of Word-of-Mouth Communication and an Investigation of Service Quality and Customer Commitment as Potential Antecedents,” Journal of Service Research, 4 (1), 60–75.

43.

Hennig-Thurau

Thorsten

Gwinner

Kevin P.

Walsh

Gianfranco

Gremler

Dwayne D.

(2004), “Electronic Word-of-Mouth via Consumer-Opinion Platforms: What Motivates Consumers to Articulate Themselves on the Internet?” Journal of Interactive Marketing, 18 (1), 38–52.

44.

Hill

Shawndra

Provost

Foster

Volinsky

Chris

(2006), “Network-Based Marketing: Identifying Likely Adopters via Consumer Networks,” Statistical Science, 21 (2), 256–276.

45.

Huffaker

David

(2010), “Dimensions of Leadership and Social Influence in Online Communities,” Human Communication Research, 36 (4), 593–617.

46.

Kawasaki

Guy

(1992), Selling the Dream. New York: Harper Business.

47.

Keh

Hean Tat

Pang

Jun

(2010), “Customer Reactions to Service Separation,” Journal of Marketing, 74 (2), 55–70.

48.

Keiningham

Timothy L

Rust

Roland T.

Lariviere

Bart

Aksoy

Lerzan

Williams

Luke

(2018), “A Roadmap for Driving Customer Word-of-Mouth,” Journal of Service Management, 29 (1), 2–38.

49.

Keiningham

Timothy L.

Frennea

Carly M.

Aksoy

Lerzan

Buoye

Alexander

Mittal

Vikas

(2015), “A Five-Component Customer Commitment Model: Implications for Repurchase Intentions in Goods and Services Industries,” Journal of Service Research, 18 (4), 433–450.

50.

Keiningham

Timothy L.

Cooil

Bruce

Andreassen

Tor Wallin

Aksoy

Lerzan

(2007), “A Longitudinal Examination of Net Promoter and Firm Revenue Growth,” Journal of Marketing, 71 (3), 9–51.

51.

Keller

(2007), “Unleashing the Power of Word of Mouth: Creating Brand Advocacy to Drive Growth,” Journal of Advertising Research, 47 (4), 448–452.

52.

Kilby

Nathalie

(2007), “Market Shift Opens Door to Word of Mouth,” Marketing Week, 30 (38), 12.

53.

Kim

Miran

Vogt

Christine A.

Knutson

Bonnie J.

(2015), “Relationships among Customer Satisfaction, Delight, and Loyalty in the Hospitality Industry,” Journal of Hospitality & Tourism Research, 39 (2), 170–197.

54.

Kock

Ned

Lynn

Gary S.

Dow

Kevin E.

Akgün

Ali E.

(2006), “Team Adaptation to Electronic Communication Media: Evidence of Compensatory Adaptation in New Product Development Teams,” European Journal of Information Systems, 15 (3), 331–341.

55.

Krapfel

Robert E.

Jr . (1985), “An Advocacy Behavior Model of Organizational Buyers’ Vendor Choice,” Journal of Marketing, 49 (4), 51–59.

56.

Kumar

Rajan

Bharath

Gupta

Shaphali

Pozza

Ilaria Dalla

(2019), “Customer Engagement in Service,” Journal of the Academy of Marketing Science, 47 (1), 138–160.

57.

Kunda

Ziva

(1999), Social Cognition: Making Sense of People. Boston: MIT Press.

58.

Lam

Son K.

Ahearne

Michael

Mullins

Ryan

Hayati

Babak

Schillewaert

Niels

(2013), “Exploring the Dynamics of Antecedents to Consumer–Brand Identification with a New Brand,” Journal of the Academy of Marketing Science, 41 (2), 234–252.

59.

Lam

Son K.

Ahearne

Michael

Schillewaert

Niels

(2010), “Resistance to Brand Switching When a Radically New Brand is Introduced: A Social Identity Theory Perspective,” Journal of Marketing, 74 (6), 128–146.

60.

Laurent

Gilles

Kapferer

Jean-Noel

Roussel

Francoise

(1995), “The Underlying Structure of Brand Awareness Scores,” Marketing Science, 14 (3), G170–G179.

61.

Lowenstein

Michael W.

(2011), The Customer Advocate and the Customer Saboteur, Milwaukee, WI: ASQ Quality Press.

62.

Magnini

Vincent P.

Crotts

John C.

Zehrer

Anita

(2011), “Understanding Customer Delight: An Application of Travel Blog Analysis,” Journal of Travel Research, 50 (5), 535–545.

63.

Mangold

W. Glynn

Miller

Fred

Brockway

Gary R.

(1999), “Word-of-Mouth Communication in the Service Marketplace,” Journal of Services Marketing, 13 (1), 73–89.

64.

Marsden

Paul

Samson

Alain

Upton

Neville

(2005), “Advocacy Drives Growth,” Brand Strategy, 198, 45–47.

65.

Matzler

Kurt

Pichler

Elisabeth

Hemetsberger

Andrea

(2007), “Who Is Spreading the Word? The Positive Influence of Extraversion on Consumer Passion and Brand Evangelism,” AMA Winter Educators’ Conference, San Diego, 25–33.

66.

Mazzarol

Tim

Sweeney

Jillian C.

Soutar

Geoffrey N.

(2007), “Conceptualizing Word-of-Mouth Activity, Triggers and Conditions: An Exploratory Study,” European Journal of Marketing, 41 (11/12), 1475–1494.

67.

Miles

Matthew B.

Michael Huberman

(1994), Qualitative Data Analysis: An Expanded Sourcebook. Thousand Oaks, CA: Sage.

68.

Murray

Keith B.

(1991), “A Test of Services Marketing Theory: Consumer Information Acquisition Activities,” Journal of Marketing, 55 (1), 10–25.

69.

Nielsen (2007), Trust in Advertising, a Global Nielsen Consumer Report, New York: The Nielsen Corporation, October.

70.

Oliver

Richard L.

(2010), Satisfaction: A Behavioral Perspective on the Consumer, 2d ed. Armonk, NY: M. E. Sharpe.

71.

Park

C. Whan

MacInnis

Deborah J.

(2006). “What’s in and What’s Out: Questions on the Boundaries of the Attitude Construct,” Journal of Consumer Research, 33 (1), 16–18.

72.

Payne

Adrian

Frow

Pennie

(2013), Strategic Customer Management: Integrating Relationship Marketing and CRM. Cambridge, England: Cambridge University Press.

73.

Perugini

Marco

Bagozzi

Richard P.

(2001), “The Role of Desires and Anticipated Emotions in Goal-Directed Behaviours: Broadening and Deepening the Theory of Planned Behaviour,” British Journal of Social Psychology, 40 (1), 79–98.

74.

Petty

Richard E.

Christian Wheeler

Tormala

Zakary L.

(2003), “Persuasion and Attitude Change,” in Comprehensive Handbook of Psychology, Personality and Social Psychology, Weiner

I. B.

Lerner

M. J.

, eds. New York: John Wiley, 353–382.

75.

Podsakoff

Philip M.

MacKenzie

Scott B.

Lee

Jeong-Yeon

Podsakoff

Nathan P.

(2003), “Common Method Biases in Behavioral Research: A Critical Review of the Literature and Recommended Remedies,” Journal of Applied Psychology, 88 (5), 879–903.

76.

Raassens

Néomie

Haans

Hans

(2017), “NPS and Online WOM: Investigating the Relationship between Customers’ Promoter Scores and eWOM Behavior,” Journal of Service Research, 20 (3), 322–334.

77.

Reichheld

Frederick E.

(2003), “The One Number You Need to Grow,” Harvard Business Review, 81 (12), 46–54.

78.

Rimé

Bernard

(2009), “Emotion Elicits the Social Sharing of Emotion: Theory and Empirical Review,” Emotion Review, 1 (1), 60–85.

79.

Rimé

Bernard

(2007), “Interpersonal Emotion Regulation,” in Handbook of Emotion Regulation, Gross

J. J.

ed. New York: Guilford, 466–485.

80.

Roberts

Kevin

(2005), Lovemarks: The Future beyond Brands. New York: powerHouse Books.

81.

Rogers

Everett M.

(1995), The Diffusion of Innovations. New York: The Free Press.

82.

Roy

Sanjit Kumar

(2015), “Modeling Customer Advocacy: A PLS Path Modeling Approach,” Journal of Strategic Marketing, 23 (5), 380–398.

83.

Rust

Roland T.

Oliver

Richard L.

(2000), “Should We Delight the Customer?” Journal of the Academy of Marketing Science, 28 (1), 86–94.

84.

Rust

Roland T.

Zahorik

Anthony J.

Keiningham

Timothy L.

(1995), “Return on Quality: Making Service Quality Financially Accountable,” Journal of Marketing, 59 (2), 58–70.

85.

Salzberger

Thomas

Koller

Monika

(2013), “Towards a New Paradigm of Measurement in Marketing,” Journal of Business Research, 66 (9), 1307–1317.

86.

Salzberger

Thomas

(2009), Measurement in Marketing Research: An Alternative Framework. Cheltenham, England: Edward Elgar.

87.

Stokburger-Sauer

Nicola

Ratneshwar

Sen

Sankar

(2012). “Drivers of Consumer-Brand Identification,” International Journal of Research in Marketing, 29 (4), 406–418.

88.

Summers

John O.

(1970), “The Identity of Women’s Clothing Fashion Opinion Leaders,” Journal of Marketing Research, 7 (2), 178–185.

89.

Swan

John E.

Oliver

Richard L.

(1989), “Postpurchase Communications by Consumers,” Journal of Retailing, 65 (4), 516–533.

90.

Sweeney

Jillian C.

Danaher

Tracey S.

McColl-Kennedy

Janet R.

(2015), “Customer Effort in Value Cocreation Activities Improving Quality of Life and Behavioral Intentions of Health Care Customers,” Journal of Service Research, 18 (3), 318–335.

91.

Sweeney

Jillian C.

Webb

Dave

Mazzarol

Tim

Soutar

Geoffrey N.

(2014), “Self-Determination Theory and Word of Mouth about Energy Saving Behaviors: An Online Experiment,” Psychology & Marketing, 31 (9), 698–716.

92.

Sweeney

Jillian C.

Soutar

Geoffrey N.

Mazzarol

Tim

(2012), “Word of Mouth: Measuring the Power of Individual Messages,” European Journal of Marketing, 46 (1/2), 237–257.

93.

Sweeney

Jillian C.

Soutar

Geoffrey N.

Mazzarol

Tim

(2008), “Factors Influencing Word of Mouth Effectiveness: Receiver Perspectives,” European Journal of Marketing, 42 (3/4), 344–364.

94.

Urban

Glen L.

(2004), “The Emerging Era of Customer Advocacy,” MIT Sloan Management Review, 45 (2), 77–82.

95.

Van Doorn

Jenny

Lemon

Katherine N.

Mittal

Vikas

Nass

Stephan

Pick

Doreén

Pirner

Peter

Verhoef

Peter C.

(2010), “Customer Engagement Behavior: Theoretical Foundations and Research Directions,” Journal of Service Research 13 (3), 253–266.

96.

Walz

Anna M.

Celuch

Kevin G.

(2010), “The Effect of Retailer Communication on Customer Advocacy: The Moderating Role of Trust,” Journal of Consumer Satisfaction, Dissatisfaction and Complaining Behavior, 23, 95–110.

97.

Wang

Xuehua

(2011), “The Effect of Unrelated Supporting Service Quality on Consumer Delight, Satisfaction, and Repurchase Intentions,” Journal of Service Research, 14 (2), 149–163.

98.

Wangenheim

Florian V.

Bayón

Tomás

(2004), “The Effect of Word of Mouth on Services Switching,” European Journal of Marketing, 38 (9/10), 1173–1185.

99.

Wilder

Kelly M.

(2015), Brand Advocacy: Conceptualization and Measurement. PhD Dissertation , Mississippi State University, Starkville.

100.

Williams

Azedah

(2017), “Food for Thought: Driving Customer Advocacy,” CMO, https://www.cmo.com.au/article/620836/food-thought-customer-advocacy/ (accessed November 20, 2018).

101.

Wilson

Jerry R.

(1994), Word-of-Mouth Marketing. New York: John Wiley.

102.

Wirtz

Jochen

Tang

Chris

Georgi

Dominik

(2019), “Successful Referral Behavior in Referral Reward Programs,” Journal of Service Management, 30 (1), 48–74. https://doi.org/10.1108/JOSM-04-2018-0111.

103.

Zaichowsky

Judith Lynne

(1985), “Measuring the Involvement Construct,” Journal of Consumer Research, 12 (3), 341–352.

104.

Zaki

Mohamed

Kandeil

Dalia

Neely

Andy

McColl-Kennedy

Janet

(2016), The Fallacy of the Net Promoter Score: Customer Loyalty Predictive Model. Cambridge, England: Cambridge Service Alliance University of Cambridge.

105.

Zeithaml

Valarie A.

(1988), “Consumer Perceptions of Price, Quality, and Value: A Means-End Model and Synthesis of Evidence,” Journal of Marketing, 52 (3), 2–22.

106.

Zeithaml

Valarie A.

Berry

Leonard L.

Parasuraman

(1996), “The Behavioral Consequences of Service Quality,” Journal of Marketing, 60 (2), 31–46.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.01 MB

0.08 MB