What’s love got to do with it? : a study on the effect of relationship types on eWOM intensity

(1)

What’s love got to do with

it?

A study on the effect of relationship types on eWOM intensity.

Name Marilou Coutty

Student ID 10167625

Master MSc. In Business Administration - Marketing track

University University of Amsterdam

Version Final version

Word count 9266

Supervisor Bob Rietveld

(2)

Statement of originality

This document is written by Marilou Coutty who declares to take full responsibility for the

contents of this document.

I declare that the text and the work presented in this document is original and that no sources

other than those mentioned in the text and its references have been used in creating it.

The Faculty of Economics and Business is responsible solely for the supervision of

(3)

Abstract

This study tries to get a deeper understanding of eWOM intensity and it's relationship with

consumer brand relationships. By investigating the concept eWOM intensity and which

relationship types have the greatest effect on this intensity, this study is filling a gap in the

literature. The five relationship types that are used in this study are self-connection,

interdependence, love/commitment, brand partner quality and intimacy. Both a survey and a

content analysis were performed on the PlayStation community. The survey gathered

information about the relationship types and the content analysis about the eWOM intensity.

The results show that the CBR intimacy has a positive effect on both post and comment

activity, thus on eWOM intensity.. Love/commitment only has a positive relationship with

comment activity. Hence, it is important to stimulate those relationship types as marketers and

brand managers in order to enhance sales, firm performance and other marketing outcomes.

1. Introduction

“Just bought the latest game of PlayStation. It is amazing!” is one of the many ways people

nowadays share their thoughts and feelings about a brand or product. Without them even

realizing, they are participating in one of the most valuable marketing strategies of these days,

word of mouth.

Word of mouth (WOM) and electronic word of mouth (eWOM) are becoming more and more

important in marketing (Sweeney, Soutar, & Mazzarol, 2012) as it is proven to be more

effective than advertising (Trusov, Bucklin, Pauwels, 2009). It is a new way of

(6)

outcomes (Rosario, Sotgiu, De Valck & Bijmolt, 2016). In other words, praising or speaking

negatively about a product in a message has less impact on marketing outcomes (e.g. sales,

firm performance) than the volume of a message. Thus, the intensity of a message proofs to

be more important than the valence. With eWOM being one of the most important marketing

tools and the declining importance of traditional advertising it is important to disentangle this

intensity part of eWOM. One way to do this is by not only looking at the concept eWOM

intensity but also by examining it's drivers.

Showing your “deepest and most meaningful sentiment” (Rubin, 1970, p.266)to your loved

ones makes perfect sense, no one will judge you for that. But is it also possible to feel such

love and passion towards a brand? And create a relationship with a brand? Consumer brand

relationship theory assumes that consumers are able to form relationships with brands in a

way they also form relationships with persons (Fournier, 1998).

Maintaining and creating consumer brand relationships (CBR) is a topic that is getting more

and more attention in the literature (Batra, Ahuvia, Bagozzi, 2012). The topic has been in the

mind of the marketers for along time, but just now brand relationships are seen as a powerful

mean to create a sustainable competitive advantage (Fournier, Breazeale, & Fetscherin, 2012).

Previous research shows that consumers perceive and evaluate brands through brand equity,

(Keller, 1993; McQueen, Foley & Deighton 1993) brand personality (Aaker, 1997 ) and brand

extensions (Aaker & Keller, 1990). Today consumer brand relationship has joined the group.

These relationships are “the repeated interactions between a brand and a customer that start to

reflect similar characteristics of relationships between people, such as love, connection,

(7)

The creation and management of CBR is crucial to the practice of marketing as there are

many different benefits that can arise from these relationships such as financial benefits,

brand loyalty, repeated purchases and an increased willingness to pay (Fournier et al., 2012).

Other advantages of CBR are the more positive and forgiving reactions of consumers in case

of brand failure (Cheng, White & Chaplin, 2012) and the increase in participation of

(positive) word of mouth (Fehr & Russell, 1991; Fullerton, 2005; Hennig‐ Thurau, Gwinner,

Walsh & Gremler, 2004). It is the latter outcome where this research is building on.

Various studies have shown that brand love (dimension of consumer brand relationships)

leads to more eWOM and positive evaluations of a brand (Albert & Merunka, 2013; Alnawas

& Altarifi, 2015; Batra et al., 2012). Although this information is important, not all

relationships are regarded as “committed relationships”. Many different types of relationships

are defined such as intimacy, interdependence and self-connection (Fournier, 1989). The

relationship types find common ground in the fact that there are all based on a feeling of

connection towards a brand. One should have some connection or a sense of love towards a

brand before a relationship can be created. What are the consequences of these types of

relationships on eWOM intensity? Does having a committed relationship with a brand result

in a different eWOM intensity compared to having an intimate relationship with a brand? And

what is eWOM intensity exactly? Just the frequency of posts or is there more? In other words,

which brand lovers are also brand ambassadors (share their brand love with peers) and why?

This study will try to answer these questions and thereby make several contributions to the

field. First, it is the first research disentangling brand love with the five different relationship

types defined by Fournier (1989). This is important because not all types of brand love are

(8)

our understanding of the concept eWOM intensity to see whether there is more than just the

frequency of messages. This research has managerial implications for brand managers and

marketers. It emphasizes on the relationship between different consumer brand relationships

and eWOM intensity. This provides a clearer picture of how relationship types can influence

the intensity of eWOM of consumers, which can ultimately lead to positive marketing

outcomes. As marketers invest more in influencing strategies (e.g. finding people with high

love and facilitating eWOM) this research helps design those strategies.

Since the differentiation of brands from other brands has become more difficult for marketers

lately, it is important to disentangle brand love in order to create different strategies (for

specific relationship types) in order to enhance and motivate positive marketing outcomes. In

other words, both eWOM intensity and CBR can help brands and products to create a

competitive advantage and the results of this study helps designing such an advantage.m

In order to analyze the effect of different consumer brand relationships on eWOM intensity

the following research question has been drawn up:

RQ: "Do the relationship types self-connection, interdependence, intimacy, love/commitment and brand partner quality predict eWOM intensity?”

2. Theory

Relationships are “sequences of interactions between parties where the probable course of

future interactions between them is significantly different from that of strangers” (Aggarwal,

2004, p. 4). Hinde (1979) states that in order to have a relationship there must be a form of

(9)

context (Berscheid & Peplau, 1983), but they also help developing a personality (Kelley,

1986).

The psychological context of a relationship is “to specify the identity activity in which the

relationship is grounded” (Fournier, 1998, p. 346). In this context relationships may help

resolve life themes, deliver on important life projects or tasks and add meaning to lives. One

function of forming relationships lies in the sociocultural context and talks about the strength

of the relationship and how this is moderated by age, life, gender, social network and culture

(Dion & Dion, 1996). Another function of forming relationships is on a more

socio-emotional level; it helps with personality and identity creation. On this level relationships may

help to reassure self worth, social integration and the projection of an image (Fournier, 1998).

Relationships can either be “substantively grounded” such as obligations or “emotionally

based” like love or friendly affection (Fehr & Russel, 1991; Sternberg, 1986). Other important

relationship dimensions are equal versus unequal, friendly versus hostile, voluntary or

involuntary and formal versus informal (Wish, Deutsch & Kaplan, 1976).

Consumer Brand Relationships

Fournier (1998) introduced personal relationships as a metaphor for the associations between

brand and customers. Brand relationships are “the repeated interactions between a brand and

customer that start to reflect similar characteristics of relationships between people, such as

love, connection, interdependence, intimacy and commitment” (Mogliner & Aaker, 2009).

The nature of consumer brand relationships has shifted over time. At first, the goal of

customer brand relationships was giving meaning to peoples lives. Now, brand relationships

are more used as facilitators; they do not give meaning to peoples lives but they support

(10)

relationships (Aggarwal, 2004). According to (Fong & Markus, 1982) people judge objects

differently from persons because they use themselves as a frame of reference when judging

others. Something that is not possible when judging nonsocial objects (Fiske & Taylor, 1991).

Therefore they believe that it is impossible to form a relationship with a brand.

A way to legitimize brands as a relationship partner comes from Aaker (1997). He shows that

consumers are personalizing brands; thinking about them as if they are human. The

acceptance and creation of the humanized brand assumes the willingness of consumers to let

brands play a role in the relationship dyad. Fournier (1998) uses the theory of animism to

support the claim that consumers can form relationships with brands. The first animistic form

involves “instances in which the brand is somehow possessed by the spirit of a past or present

other” (p. 345). This relates to the associations either from the past or the present consumers

might have with brands. For example the perfume of your grandmother can hold strong

associations with the past but the gift-wrap of a present you just got has more associations

with the present. The anthropomorphization of the brand, with emphasize on the humanistic

part (emotion, thought) is another form of animism. These brand characters serve as an

example and help memorize the brand in the mind of consumers. George Clooney helping

memorizing Nespresso by being in their ad is an example of that.

In order to become a legitimate relationship partner, brands must go beyond personification

and associations and behave as an active partner of the dyad (Fournier, 1998). This can be

done through interaction: when consumers interact with brands, they can develop an active

relationship with them like people can do with loved ones or friends (Aaker, 1996). Despite

some criticism, the idea that brands exceed their functional value, thereby leading consumers

(11)

Brand relationship is a complex construct; there are a dozen types of brand relationships each

associated with different emotions and norms (MacInnis, Park & Priester, 2014). Fournier

(1998) has done extensive research on brand relationships and the different types in which

they occur. She discovered that a critical purpose of brand relationships is the communication

of an identity; who are you, who were you and who do you want to be? Using brands

communicates those things about you to the outside world because people have associations

with that brand. Thus, consumers benefit from those associations and add meaning to their

lives.

The typology of consumer brand relationship forms

There are fifteen different brand relationships forms according to Fournier (1998). They all

serve a different purpose and have a variation in level of attachment to a brand. If consumers

have an arranged marriage with a brand the relationship is not voluntary but “imposed by

preferences of a third party” (p. 362). The level of affective attachment is relatively low in

this relationship form. Another relationship form is committed partnership where intimacy,

trust and commitment play an important role. This relationship type is voluntary and aimed on

the long run. Here, the level of attachment is quite high. Secret affairs are highly held private,

exciting and risky. Emotion plays a huge role in this customer brand relationship form and

therefore has a high attachment level. A relationship form where there is an “obsessive, highly

emotional, selfish attractions cemented by feeling that the other is irreplaceable” (p. 362) is

called dependencies. This relationship is so strong that whenever a brand makes a mistake,

there is still a high tolerance for its transgression. Other relationship forms are such as best

(12)

are the five underlying dimensions that keep relationships alive (Fournier, 1998). All together

they form a construct that measures the brand relationship quality, depth, durability and

strength. “The multifaceted nature of the construct highlights that there is more to keeping a

relationship alive than the pull of positive feelings” (Fournier, 1998, p. 363). The five factors

can be divided into affective and socio-emotive attachment (love/commitment and

self-connection), behavioral ties (interdependence) and supportive cognitive beliefs (intimacy and

brand partner quality).

The following section will discuss the five dimensions of consumer brand relationships

defined by Fournier (1998).

Love and commitment

The brand love relationship is “deep and enduring (beyond simple affect), such that the loved

brand is considered irreplaceable” (Albert & Merunka, 2013). Here, a deeply rooted feeling of

affecting towards the brand lies within the consumer. Consumers, who feel love and

commitment towards a brand, feel as if something is missing when deprived of the brand for a

certain amount of time (Fournier, 1998). Brand love leads to biased, positive perceptions of

the brand (Albert & Merunka, 2013). Ahluwalia, Burnkrant and Rao (2000) even found that

committed consumers are less usceptible to negative information about a brand and weigh

positive information more strongly. A committed relationship with a brand leads to

willingness to maintain a valued relationship over time (Chaudhure & Holbrook, 2001). It

creates stability by encouraging derogation of alternatives in the environment (Fournier,

1998). Brand commitment consists of affective components (emotional, shared values,

attachment and trust) and continuance components, which is “rooted in the economic and

(13)

The self-connection dimension focuses on “the degree to which the brand delivers on

important identity concerns, tasks, or themes, thereby expressing a significant aspect of the

self” (Fournier, 1998, p. 364). Does the brand express the identity that I want to express and

how does this brand fit within my identity are some of the questions that might be asked here.

Strong self-connection to a brand helps maintaining a relationship through protective feelings

one might have towards the brand. This can be explained as follows, when a consumer feels

as if the brand is a reflection of themselves, an attack on a brand might feel as attack on their

own identity (Drigotas & Rusbult, 1992).

Interdependence

An interdependence relationship with a brand consist of the thought that consumer cannot

function without that brand. This type of relationship is based on “frequent brand interactions,

increased scope of brand-related activities and a heightened intensity of individual interaction

events” (Fournier, 1998, p. 365).

Intimacy

Another aspect of a strong brand relationship is intimacy; it describes the closeness and

harmonization of the consumer brand relationship. “It mirrors consumers’ perceptions of the

care, understanding and attention they receive from the brand” (Jun, Tat & Siqing, 2009). A

memory of personal associations and experiences is an example of an intimate relationship

between brand and consumer. This type of relationship develops a strong and deep feeling

towards the brand (Escalas, 1996).

Brand partner quality

This component focuses on the overall satisfaction of the relationship between brand and

consumer. There are five components of brand partner quality. First, the consumer needs to

(14)

respected). Secondly, there needs to be a judgment of the brand’s partnership role. Another

component is the brand’s ability to oblige to certain rules (implicit relationship contract).

Furthermore consumers need to have trust and faith in the brand delivering what is expected.

Lastly, there should be comfort in the brand’s accountability for its actions (Clark, Helgeson,

Mickelson & Pataki, 1994).

Word of mouth & electronic word of mouth

In a world where people get overloaded with product information from a lot of sources

(Plummer, 2007) consumers pay less attention to traditional advertising (Nielsen, 2007). In

the marketing literature it is well known that WOM and eWOM are powerful tools to shape

the public opinion on a variety of issues (Kareklas, Muehling & Weber, 2015). A reason for

this is that WOM is perceived as more credible and trusted by consumers than traditional

advertising (Brown, 2005). Word of mouth is defined as “the informal communication

directed at other consumers about ownership, usage or characteristics of particular goods and

services or their sellers” (Westbrook, 1987, p. 261). This concept has a great influence on

consumer behavior since people rely heavily on the evaluations of other people (Gruen,

Osmonbekov & Czaplewski, 2005).

Since the rise of the Internet the importance and influence of worth of mouth has drastically

increased (Park & Lee, 2009). The Internet has made it possible for ordinary people to reach a

mass audience and “grab a hold of the megaphone” (Bourdieu, 1990, p. 28). These people

have the opportunity to reach thousands of people and share their opinion about certain

products with the mass audience. Electronic word of mouth (eWOM) is defined as “any

(15)

Internet” (Henning-Thurau et al., 2004, p.39). It is something produced in consumer networks

(social media, forums) that help to create an opinion about a certain product or brand

(Kulmala, Mesiranta & Tuominen, 2013).

EWOM intensity

There is little consensus on the metrics of eWOM (King, Racherla & Bush, 2014). Although

there is an agreement that eWOM has significant monetary effects on sales beyond other

marketing mix effects (Chen, Wang & Xie, 2001), there is still disagreement on what drives

that effect. Some studies show that the number of online reviews predicts products sales

whereas others show that it is not the volume of eWOM but its valence (Chintagunta,

Gopinath & Venkataraman, 2010). To figure out which one (volume or valence) has a greater

effect on firm performance Rosario, Sotgiu, De Valck and Bijmolt (2016) performed a

meta-analysis. Their results show that volume plays a more important role on firm performance

than positive sentiment. In other words, praising or speaking negatively about a product in a

message has less impact on certain marketing outcomes (e.g. sales, firm performance) than

the volume of a message.

With results like these it is only logical to dive deeper into the meaning of volume and how

this is part of eWOM intensity. Intensity is defined as “the quality or state of being intense,

especially the extreme degree of strength, force, energy or feeling”, it is the “magnitude of a

quantity” (Merriam-Webster’s Collegiate Dictionary, 2016).

The concept of intensity can also be linked to electronic word of mouth since there can be

variation in the degree of strength and the magnitude of quantity in an online message.

Quantity (or frequency) of eWOM messages, which is part of the definition of intensity, is a

(16)

Jansen & Chowdhury, 2011). Frequency is an important part of measuring the strength and

therefore the intensity of eWOM. But is intensity only frequency? If someone posts three

messages on a forum does that measure intensity as a whole? Or is writing one in depth

(500+) message more “intense”? According to Goyette, Richard, Bergeron and Marticotte

(2010) eWOM intensity is not just the activity but also the volume of a message. Various

studies try to measure the strength of a message by looking at “the total amount of

interaction”(Liu, 2006, p.75), also known as activity or engagement.. This interaction

determines the effect of eWOM on things like firm performance and trustworthiness of the

message (e.g. Rosario, Sotgiu, De Valck &, 2016; Hu, Koh & Reddy, 2014).

Something that is closely related to interaction in the level of engagement. Kim et al., (2008)

researched engagement in online communities and identified engagement as a key connector

between online communities and brand outcomes. This engagement is defined as "the

consumer intrinsic motivation to interact and cooperate with community members" (Writz et

al., 2013, p. 229). Engagement can take place in different ways such as blogging, word of

mouth and writing reviews. Men and Tsai, (2013)split engagement up in two things; reactive

engagement and proactive engagement. Proactive engagement consist of creating and

contributing content, here the content is initiated by the consumer. An example of this is

writing posts, blogs or any other type of content. On the other hand, reactive engagement

focusses more on consuming content and commenting on that. Responding on a blog, review

or post are good examples of reactive engagement. In this research proactive engagement is

defined as post activity and reactive engagement is defined as comment activity.

By putting this information all together a clear picture can be drawn of eWOM intensity.

Building further on the definition given by Goyette et al., (2010), Liu (2006) and Men and

(17)

eWOM message (volume) and the type of engagement of these messages (proactive vs reactive).

Drivers of eWOM

Berger (2014) has done extensive literature research into the factors that drive word of mouth,

which can be translated to electronic word of mouth. Berger has identified five key functions

that predict the actual word of mouth behavior: impression management, emotion regulation,

information acquisition, social bonding and persuasion. These five functions are all

self-serving “and drive what people share even outside of their awareness” (Berger, 2014, p.588).

Meaning that people are not always conscious about if and how they engage in word of

mouth. Three of the five functions are described below and linked to three consumer brand

relationship types in order to predict the eWOM intensity. The other two (information

acquisition and persuading others) are not described here, as they can’t be linked to the CBR

types .The driver information acquisition influences WOM when people have a feeling of

uncertainty or a lack of trust (Berger, 2014). According to this research these feelings can’t be

linked to one of the five relationship types, as there is no type that concerns these feelings.

The other driver, persuading others, also has no link with one of the relationship types.

Therefore both information acquisition and persuading others are not elaborated on in this

(18)

Figure 1. Conceptual model

Impression management

One of the reasons consumers engage in word of mouth is to shape the impressions others

have of them (and they have of themselves). The social interaction occurring from word of

mouth can be seen as a performance (Goffman, 2002) where people present themselves in

particular ways. Sharing word of mouth may present who people are or who they want to be

in three ways: self-enhancement, identity signaling and filling conversational space. The

sharing of high status things, self-concept relevant things and unique and special things are a

few interesting effects of word of mouth in terms of impression management Berger (2014).

The dimension self-connection of the brand relationship quality scale is closely related to

impression management, especially identity signaling. The more people feel a sense of

self-connection (the brand = me), the more they will use the brand to signal their identity to the

world. In order words “the brand serves as a means of self-expression” (Huber, Meyer &

Schimd, 2015, p.569). Hence, the following hypothesis is drawn up in order to investigate the

relationship between self-connection and the post activity part of eWOM intensity. Love/Commitment Self-connection Interdependence Intimacy Partner quality eWOM intensity

(19)

H1a: The consumer brand relationship type self-connection has a positive relationship with post activity.

Next to expressing one's identity, self-connection also creates protective feelings towards a

brand. Consumers see the brand as a reflection of themselves and if it gets attacked, it might

feel as an attack on their own identity. On the other hand, if the brand gets praised, it could

give consumers a good feeling, because their own identity gets praised (Drigotas & Rusbult,

1992). Hence, a negative message about a brand could lead to comments in order to protect

the brand. Positive messages might also lead to comments, approving the content of the

message. Therefore the following hypothesis is drawn up:

H1b: The consumer brand relationship type self-connection has a positive relationship with comment activity.

Emotion regulation

Another function of word-of-mouth is emotion regulation. This function helps consumers to

regulate, manage and express their emotions (Gross, 2008). Research even shows that

communal aspects helps coping with emotions. This social aspect of sharing “provides an

important channel for sharers to regulate their emotion” (Berger, 2014, p.592). Emotion

regulation can be divided into generating social support, venting, facilitating sense making,

reducing dissonance, taking vengeance and encouraging rehearsal (Berger, 2014). This study

only focuses on encouraging rehearsal, used by consumers to regulate their emotion by

reliving a positive experience (Rimé, 2009). Encouraging rehearsal can be linked to the

consumer brand relationship type love and commitment. If one feels love and commitment

(20)

want to share with other people (communal aspect). To investigate the relationship between

love/commitment and the eWOM intensity a hypotheses are drawn up:

H2: The consumer brand relationship type love/commitment has a positive relationship with post activity.

Building further on Rimé (2009) reliving positive experiences are a mean to regulate

emotions. This moment of reliving is only experienced by the one who had the positive

experience with the brand. Therefore, commenting on a positive experience of someone else

makes other consumers not relive the experience since it was not theirs. Hence, the following

hypothesis has been drawn up:

H2b: The consumer brand relationship type love/commitment has a negative relationship with comment activity.

Social bonding

A third function of word-of-mouth is social bonding in which people connect with others

(Rimé, 2009). Social bonding should drive people to talk about things they have in common

as people prefer common ground topics since everyone can relate to them (Clark, 1996). A

community is a perfect place to share and bond with people on common ground. Social

bonding can be linked to interdependence component of the brand relationship scale and

eWOM intensity. According to Lee, Soo Kin and Ku Kim (2012) “active eWOM participation

can be evoked by how consumers see themselves in relation to other members in their online

brand communities” (p. 1054). One could state that when consumers see that people around

them also have a level of interdependency towards a certain brand, they feel they hit common

(21)

within an online community or when they want to respond to someone else’s post. Hence, the

following hypotheses are drawn up in order to investigate the relationship between

interdependence and eWOM intensity.

H3a: The consumer brand relationship type interdependence has a positive relationship with post activity.

H3b: The consumer brand relationship type interdependence has a positive relationship with

comment activity.

Intimacy & brand partner quality

Having an intimate relationship with a brand is about feelings of closeness, harmonization and

a sense of privateness (Jun, Tat & Siqing, 2009). Having an intimate and private relationship

with a human being would assume that one would not share everything with the entire world.

Building on the parallel of Fournier (1998) between interpersonal –and consumer brand

relationship the following hypothesis is drawn up:

H4a: The consumer brand relationship type intimacy has a negative relationship with post activity.

Even though a negative relationship is expected between intimacy and post activity, the social

bonding theory suggests a positive relationship between intimacy and comment activity.

Again this can be explained by the "common ground" factor; if consumer feel like other

consumers also have an intimate relationship with a brand they feel a sense of belonging and

might respond to that (Rimé, 2009; Clark, 1996). Hence, the following hypothesis is stated:

H4b: The consumer brand relationship type intimacy has a positive relationship with comment activity.

(22)

Brand partner quality is about the overall satisfaction of the relationship between brand and

consumer. The evaluation of the brand partner quality is based on the partnership role of the

brand, the implicit relationship contract, judgment of the brand’s partnership role, the comfort

in the brand’s accountability for its actions and trust in the brand (Clark, Helgeson, Mickelson

& Pataki, 1994). If all the expectations are met by the brand it is likely that consumer are

sharing these qualities about the brand with the outside world (Gruen, Osmonbekov &

Czaplewski, 2005). Therefore the following hypothesis is drawn up:

H5a: The consumer brand relationship type brand partner quality has positive relationship with post activity.

For the reactive part of eWOM intensity, social bonding is once again a good theory to build

on. People might feel a sense of connection when they see that other people share the qualities

about the brand to the rest of the world (Rimé, 2009). They hit common ground, since both

the proactive and the reactive consumer can relate to the brand partner quality (Clark, 1996).

One could state that when consumers see that people around them also have a great level of

satisfaction towards a brand (and post about it) they start responding to these messages.

Therefore the following hypothesis is drawn up:

H5b: The consumer brand relationship type brand partner quality has a positive relationship with comment activity.

Four of the five relationship types are expected to have positive relationship with post

activity. The more people have either a love/commitment, interdependence, brand partner

(23)

Self-connection, interdependence, intimacy and brand partner quality are expected to have a

positive with comment activity. Only love/commitment is expected to have a negative

relationship with comment activity.

Table 1. Hypothesis Love/ Commitment

Self-connection Interdependence Intimacy

Brand partner quality eWOM intensity post activity + + + - + eWOM intensity comment activity - + + + +

3. Method

The best way to give an answer to the research question is through a survey and a content

analysis. By using triangulation “a combination of methodologies in the study of the same

phenomenon” (Denzin, 1978, p.291) a higher validity can be assured. The main goal of

performing a survey is “to describe, compare or explain individual and societal knowledge,

feelings, values, preferences and behaviors” (Fink, 2015, p. 1). A content analysis on the

other hand is “a research technique for the objective, systematic and quantitative description

of the manifest content of communication” (Bryman, 2001, p. 178).

The setting of this research is within a brand community under the assumption that members

of these communities most likely already have a certain level of brand love (Muniz &

O’Guinn, 2001). This existing level of brand love will be measured as a zero-point in order to

make clear distinctions between the different types of brand love and their effect on eWOM

(24)

(Muniz & O’Guinn, 2001). In line with this thought, members of the PlayStation community

were selected to participate in this research.

Pre-test

A total of 32 respondents participated in a pre-test. The pre-test was performed to make sure

the questions of the survey were clear in terms of wording and format to avoid ambiguity. The

distribution of the respondents on the five relationship types was also measured to test

whether there was enough variation between them and no correlation. The results of the

reliability test can be found in Appendix 1.

Data collection procedure

By using the search engine Google, different websites came up with lists of the best online

brand communities. Ridings et al., (2002) listed a few requirements to identify the activity on

communities. First, the message board must have at least 10 postings per day for each of three

days chosen at random. Second, at least 15 different individuals should post on the message

board over a randomly selected three-day period. Lastly, at least 80% of the postings must

have one reply or more of the three days chosen at random (p. 282). The PlayStation

community met all three requirements. Permission was asked and given to post an online

survey and investigate the posts of their community members. The data about the type of

relationship the PlayStation community members had was gathered with a survey. A content

analysis on the community website gathered the data about eWOM intensity.

Sample

(25)

in the survey and their online profile on the community including their blog posts. To gather

the data the non-probability sampling technique “self selection” was used. The members of

the PlayStation forum were asked to fill out a 5-minute survey to get their username and to

research their relationship type with PlayStation. The survey was online for almost three

weeks between November and December 2016 and resulted in 133 responses. After errors

were deleted116 surveys remained valuable. Since there was no information on the total

number of community members a response rate could not be determined.

Participants

Of the 116 respondents, 91.4% is male and 8.6% female. Most of the respondents are in the

age group 26-35 (45.7%), have a college degree (44.0%) and are of Dutch nationality

(82.8%).

Table 2. Characteristics of the respondents

Characteristics Frequency Percent

Gender Male Female 106 10 91.4% 8.6% Age 18-25 26-35 36-45 46+ 50 53 10 3 43.1% 45.7% 8.6% 2.6% Education High school College Bachelor Master Other 17 51 46 2 0 14.7% 44.0% 39.7% 1.7% 0.0%

(26)

Nationalitya Dutch German British American Other 96 5 4 5 5 82.8% 4.3% 3.4% 4.3% 4.3% Notes: Total N = 116 Measurement

The questionnaire measuring the different types of brand relationships was based on an

existing questionnaire and adapted to the context of this study. Using a 5 item 7-point Likert

scale ranging from 1 (strongly disagree) to 7 (strongly agree), the five different relationship

types (interdependence, love and commitment, partner quality, self-connection and intimacy)

were measured (Fournier, 2000). The reliability of the scale is presented in table 3.

The content analysis was performed with a web crawler, gathering data such as post

frequency, badge count, type of badge and the start date of the membership.

All the gathered data was imported into the statistical program SPSS version 20 and

SmartPLS. The latter is a software application that is designed for path modeling and analysis

(Ringle, Winde & Will, 2005).

Table 3. Reliability of BRS Cronbach’s Alpha Self- connection .92 Love/commitment .91 Interdependence .97 Intimacy .76

(27)

PCA

A principal component analysis (PCA) was conducted on the variables badges and post

frequency. The Kaiser-Meyer-Olkin measure verified the sampling adequacy for the analysis,

KMO = .76. Bartlett’s test for sphericity was significant, p<0.005. An initial analysis was run

to obtain eigenvalues for each component in the data.

The results show six factors with an eigenvalue over Kaiser’s criterion of 1. Looking closely

at the results it was noted that some of the factors existed only of one component and also

components that were present in other factors. These items were deleted since a factor

consisting only of one component is not very strong (Dunteman, 1989). Another PCA was

performed and now only two factors had an eigenvalue above 1. Looking at the components

of the two factors, badges concerning comment activity and liking activity were put both

under the same factor. Therefore, the liking badges were deleted from the data in order to get

two unambiguous factors namely post activity and comment activity. Together the two factors

explained 53.91% of the variance. Factor one explained 41.46% and factor 2 12.45%. In

agreement with Kaiser’s criterion, examination of the scree plot revealed a leveling off after

the second factor. Thus, two factors were retained and rotated leading to loadings after

rotation. Appendix III shows these factor loadings. The items that cluster on the same factors

suggest that factor 1 represents post activity and factor 2 comment activity. The explanations

of the different variables of the factors are in appendix VI. These results show that eWOM

intensity is a complex and multidimensional construct.

To check whether if the results are not due to the PCA the model PLS model has been run

twice. The first run contained the scores for posts and comments and the second run contained

(28)

Analyses

Before the research question could be answered different test are performed. First, all the

errors such as missing values were deleted from the dataset. Second, a principal component

analysis (PCA) has been performed to check whether the relationship types each form a

uni-dimensional scale, the results are shown in appendix II. Lastly, the five relationship types

were computed into scale means and checked for their reliability. Logarithms of the count

data were taken before they were used in the analysis, resulting in log variables.

Multiple analytical tests were performed in order to answer the hypotheses and research

question. A second principal component analysis has been performed to examine how many

factors explain eWOM intensity. The two major factors were then used for further analysis.

Both eWOM intensity and the relationship types, as the control variable days a member, were

then put together in a PLS (partial least square) regression to test how much the relationship

types predict eWOM intensity. PLS regression is a method that combines both features of

PCA and multiple regression. “Its goal is to predict or analyze a set of dependent variables

from a set of independent variables or predictors.” (p.1). It is particularly useful when there

are multiple dependent variables and a (large) set of independent variables (Abdi, 2007).

Another advantage of PLS, and very applicable to this study, is that it can be used with a

small sample size cause a bootstrap method can be applied to the analysis (Hensler, Ringle &

Sinkovics, 2009). Thus, to research the relationship between the brand relationship types and

eWOM intensity, a PLS was conducted on the sample of 116.

Table 4. Mean, Standard deviation, correlations and reliability

M SD G A I LC PQ SC I

(29)

Interdependence (ID) 4.18 1.29 .02 .07 (.96) Love/Comm (LC) 4.31 .94 .11 .19* .32** (.94) Partner quality (PQ) 3.59 .87 -.14 .06 -.04 -.06 (.91) Self-connection (SC) 4.55 .80 -.00 .09 .32** .23* .02 (.92) Intimacy (I) 4.22 .78 .07 .01 .31** .15 -.16 .26** (.80)

* Correlation is significant at the 0.05 level (2-tailed). ** Correlation is significant at the 0.01 level (2-tailed).

Results

This study tries to disentangle brand love and eWOM intensity. To check whether eWOM

intensity consists of multiple constructs (post activity and comment activity) a PCA analysis

has been performed. After that the predictability of the identified constructs of the different

brand relationship types are tested with a PLS.

Partial Least Square

According to Henseler, Ringle and Sinkovics (2009) the statistical output of PLS needs to be

analyzed based on a couple of criteria. Chin (1998) states that there is a two-step process in

analyzing the output. First each indicator needs to be tested for reliability and validity

(assessment of the measurement model) the second step is to examine the variance explained

(R²) for the endogenous constructs and the path coefficients between the constructs

(30)

The measurement model

The measurement model tests the reliability and validity of the constructs that are being

analyzed. These criteria’s are tested with discriminant validity, convergent validity and

reliability.

Reliability and convergent validity

Reliability is defined as the “consistency of outcomes after repeated measurements”

Cronbach, 1947, p.1). The correlations between the construct and its indicators assess the

reliability of the indicator’s. The same loadings represent the extent to which the indicators

are able to measure their construct. Since the indicators variate in their reliability, they should

all be tested. The indicators should explain a “substantial part of each indicator’s variance” (p.

299). The outcomes can be interpreted as Cronbach’s alpha, making the indicators reliable if

they are above 0.7.

Four indicators of self-connection were below the recommended 0.7. But in line with

Henseler et al., (2009) only one item (Q1) is deleted for the study since the elimination of the

other items did not lead to a substantial increase of the composite reliability. All the other

factor loadings were above .07. An overview of the outer loadings is given in appendix V.

To assess the internal consistency reliability, Cronbach’s alpha is one of the most used tests.

Nevertheless, this measure underestimates the PLS model and therefore the composite

reliability test is more suitable. A criterion for composite reliability is “using the average

variance extracted” (AVE). This value should be at least 0.5 for a sufficient convergent

validity (Henseler et al., 2009, p. 299).

In table 5 the AVE and composite reliability are displayed. All the constructs have a sufficient

(31)

Table 4: AVE and composite reliability measures

AVE Composite Reliability

Self-Connection .56 .93 Brand Partner Q .72 .93 Intimacy .86 .95 Interdependence .92 .97 Love/Commitment .74 .96 Discriminant validity

Discriminant validity can be measured with Fornell-Larcker criterion and the cross loadings.

The first criterion states that the “AVE of each latent variable should be greater than the latent

variable’s highest squared correlation with any other latent variable” (p. 300). The latter

criterion measures the discriminant validity of the indicators. Therefore, the loading of each

indicator is expected to be higher than any other construct (Henseler et al., 2009). Table 6

shows that all latent variables are greater than the highest squared correlation of any other

variable. The cross-loadings are also tested and show that all the indicators have sufficient

discriminant validity. The results can be found in appendix VI.

Table 6: Fornell-Larcker Criterion

BPQ ID I- BC I - CB L/C SC BPQ 0.850 ID -0.035 0.960 I -0.170 0.203 0.928 L/C -0.039 0.344 0.156 0.066 0.861 SC -0.021 0.312 0.170 0.229 0.217 0.748 Conclusion

(32)

that the data is appropriate for further testing. All the requirements were met with the

exclusion of the first indicator of self-connection. The reliability is all above 0.7, the AVE

above 0.5, the cross-loadings are sufficient and the Fornell-Larcker criterion is met.

The structural model

As the measurement model determines estimations of the evaluation of the inner path model,

the structural model is concerned with the structural model. The far most important criterion

for this examination is the coefficient of determination (R²) of the endogenous constructs. R²

values can either be substantial (0.67), moderate (0.33) or weak (0.19) according to Chin

(1998).

The path coefficients of the structural model are the standardized regression coefficients (ß)

(Henseler et al., 2009). The bootstrapping technique indicates the significance of the path

model by providing a t-value. With a one-tailed significant level of 0.05 and 498 degrees of

freedom the significant value of t is 1.65. Implying that when t-values are 1.65 or above the

hypothesis is accepted, otherwise the hypothesis should be rejected.

Table 7: Bootstrap statistical output

Hypotheses Path coefficient Standard Deviation T Statistics P Values

H5b -.11 .11 .82 .413 H5a -.04 .12 .36 .047 H3b .14 .10 1.22 .223 H3a -.09 .12 .95 .341 H4b .19 .09 2.00 .047 H4a .25 .10 2.90 .004 H2b .24 .10 2.03 .043 H2a -.03 .13 .30 .762 H1b .08 .18 .68 .500 H1a .01 .16 .13 .895

(33)

Table 8: Path coefficients and p-values

According to Chin (1989) the predictive power of the model for the dependent variables is

weak. The variance in both post activity and comment activity can for a small part be

explained by the relationship types. The variance in post activity can be explained for 19%

and the variance in comment activity for 12%. It could be stated that the model is capable of

explaining the endogenous constructs..

(34)

Hypotheses

The first hypothesis “The consumer brand relationship type self-connection has a positive

relationship with eWOM intensity (post activity)” is rejected. The relationship between self-connection and post activity was found non-significant t = 0.13, p = 0.90.

Hypothesis 1b is also found not significant t = 0.68, p = 0.50. There is no positive relationship

between self-connection and comment activity and therefore the hypothesis is rejected.

Hypothesis 2a is also found not significant. The consumer brand relationship type

love/commitment has no positive relationship with post activity t = 0.30, p = 0.76. On the

other hand, love/commitment does have a significant positive relationship with comment

activity t = 2.03, p = 0.04. Thus, the more consumers have a love and committed relationship

with a brand, the more they will comment on posts in an online community.

Hypothesis 3a and 3b tested whether there was a positive relationship between

interdependence and post activity and comment activity. The results show neither a

significant relation t = 0.95, p = 0.22 with post activity nor with comment activity t = 1.22, p

= 0.34.

The first part of the fourth hypothesis states a negative relationship between intimacy and post Construct R Square

Post activity .19

(35)

p< 0.47. Therefore the hypothesis is partially confirmed. Hypothesis 4b is confirmed, as there

is a significant positive relationship between intimacy and comment activity, t = 2.0, p = 0.04.

The more one has an intimate relationship the more one will engage in comment activity and

post activity.

Brand partner quality has neither a significant relationship with comment activity t = 0.36, p =

0.04 nor with post activity t = 0.82, p = .41 Therefore H5a and H5b are both rejected.

Conclusion & discussion

This research sheds some light on the construct eWOM intensity and on the ability to predict

the variance in eWOM intensity with the five different relationship types defined by Fournier

(1998). In order to answer the research question a survey was conducted on the PlayStation

forum to examine the relationship types and a content analysis was performed to gather data

about eWOM intensity. This study tries to fill a gap in the literature by examining eWOM

intensity more closely than previous research and by linking this construct to the different

relationship types. Both eWOM intensity and the relationship types show some interesting

results that will be further discussed in order to answer the research question "Do the

relationship types self-connection, interdependence, intimacy, love/commitment and brand partner quality predict eWOM intensity?”

The PCA shows that eWOM intensity is a multidimensional construct. It consists of actively

posting content online (post activity) and commenting on existing posts (comment activity).

This confirms that eWOM intensity is a complex, multidimensional construct that needs to be

(36)

of posts nor does it simply consist of “the total amount of interaction”(Liu, 2006, p.75; Zhang,

Jansen, & Chowdhury, 2011). In other words, eWOM intensity should not only be determined

by volume but also by the level of engagement.

The second part of this research examined what effect the different relationship types had on

eWOM intensity. What has brand love got to do with brand ambassadorship? More

importantly, when does brand love lead to eWOM intensity and thus sharing information with

peers?

The results of this research are somewhat shocking, as they show a significant, but limited

relationship between CBR and eWOM intensity. First of all, answering the research question,

the findings of this study suggest that the variance in eWOM intensity can partially be

explained by by all CBR types (self-connection, intimacy, interdependence, love/commitment

and brand partner quality) together. The variance in post activity is explained for 19% and the

variance in comment activity for 12%. In the prediction of post activity, a significant unique

contribution is found for intimacy. In the prediction of comment activity, a significant unique

contribution is found for love/commitment and intimacy.These results cannot be compared to

previous studies since this is the first study looking at the predictability of eWOM intensity

through consumer brand relationship types. The contributions of each CBR type to the

prediction of eWOM intensity will be elaborated on in the next paragraphs.

The fact that the CBR type self-connection does not make a significant unique contribution

prediction of neither post nor comment activity and thus eWOM intensity is not in line with

the expectation of Berger (2014) and the results of Huber et al., (2015). They state that the

more people feel a sense of self-connection (the brand = me), the more they will use the brand

(37)

constructs and might have different drivers.

Secondly, a relationship between love/commitment and post activity was also not confirmed.

These findings are not in line with a lot of previous research showing a positive relationship

between brand love and eWOM (Albert & Merunka, 2013; Alnawas & Altarifi, 2015). Then

again, this research looked at eWOM intensity, which apparently has different drivers than

regular eWOM.

Contrary to expectations, a positive relationship between love/commitment and comment

activity was found. The theory of Rimé (2009) about reliving experience could be an

explanation for this discrepancy. People might relive a positive experience with a brand when

they read about positive experiences of others. Therefore they might share there own positive

experience with the brand by commenting on them. Thus, having a love/commitment

relationship predicts significantly more comment activity but has no effect on post activity.

The CBR type interdependence has no significant relationship with neither post activity nor

comment activity. This is not in line with the literature, which explains that “active eWOM

participation can be evoked by how consumers see themselves in relation to other members in

their online brand communities” (Lee, Soo Kin & Ku Kim, p.1054, 2012). Both results might

be explained by the social bonding theory, in which people connect with others (Rimé, 2009)

which drives them to talk about things they have in common (Clark, 1996). Therefore, an

explanation for the absent relationship could be that people first want to have proof of the fact

that other people are also interdependent of a brand before they start sharing.

In line with the expectations a positive relationship was found between intimacy and

(38)

to comment activity, which is only evoked by an intimate relationship (Rimé, 2009; Clark,

1996).

The fact that no relationship was found between intimacy and post activity indicates that no

parallel can be drawn between interpersonal and consumer brand relationships, which is not in

line with the thought of Fournier (1998). Apparently, if people have an intimate relationship

with a brand they are willing to show that to the world, only when someone else has started

the conversation already. It could be that only then people feel safe enough to share.

The results of this research show that there is no relationship between brand partner quality

and eWOM intensity, neither post nor comment activity. This outcome is in contrast to the

literature. Gruen, Osmonbekov and Czaplewski (2005) state that consumers are likely to share

qualities about the brand only if certain expectations are met.

Managerial implications & contribution

The results illustrate that forming the brand relationship intimacy and love/commitment have

a positive effect on eWOM intensity. The more people have an intimate relationship with a

brand the more they are participating in both comment and post activity. The CBR

love/commitment leads to more comment activity. Thus, it is important for brand managers

and marketeers to stimulate these relationship types with consumers which enhance eWOM

intensity. That way, they can enhance firm performance and create a sustainable competitive

advantage (Fournier, Breazeale, Fetscherin, 2012; Trusov et al., 2012).

This study contributes to the existing literature on eWOM. While a lot of research has been

(39)

of eWOM intensity. While previous research considered intensity to be only frequency, this

study extended this concept by adding reactive engagement (comment activity) and proactive

engagement (post activity) to the mix.

Limitations

This study knows several limitations. First of all, the non-probability sampling technique “self

selection” might have an influence on the generalizability of the results. Another limitation

considering generalizability is that this research was situated in only one community. It is

recommended that future research studies more communities to increase the generalizability.

A third limitation is that there were only 116 respondents participating in this research. Even

though bootstraps were performed to be able to test the small sample size, it might still affect

the outcomes of this study. Furthermore, the PLS path modeling technique “does not have less

stringent assumptions about the representativeness of the sample” (p. 283). This could also

affect the generalizability of the outcomes of this research. Another limitation of PLS is that it

does not resist multicollinearity, which could lead to wrong calculations regarding individual

predictors (Henseler et al., 2009). Lastly, the principal component analysis showed six factors

with an eigenvalue above 1. Even though there were valid reasons for reducing the six factors

to two it might have an impact on the results. With these limitations in mind the results of this

study should be interpreted with caution.

Future research

One of the main results of this study is that eWOM intensity is a multidimensional construct.

Since the relationship types are only partial predictors for participation in eWOM intensity

(40)

interesting suggestion for future research will be to study if the relationship types do have an

effect on WOM and eWOM. It is likely that the relationship types do predict the participation

in WOM and eWOM. As stated before, it would also be interesting to investigate the effect of

(41)

References

Aaker, D. A. (1996). Measuring brand equity across products and markets. California

Management Review, 38(3), 102-120.

Aaker, D. A., & Equity, M. B. (1991). The free press. New York, 206

Aaker, D. A., & Keller, K. L. (1990). Consumer evaluations of brand extensions. The Journal

of Marketing, , 27-41.

Aaker, J. L. (1997). Dimensions of brand personality. Journal of Marketing Research, ,

347-356.

Abdi, H. (2003). Partial least square regression (PLS regression). Encyclopedia for Research

Methods for the Social Sciences, , 792-795.

Aggarwal, P. (2004). The effects of brand relationship norms on consumer attitudes and

behavior. Journal of Consumer Research, 31(1), 87-101.

Ahluwalia, R., Burnkrant, R. E., & Unnava, H. R. (2000). Consumer response to negative

publicity: The moderating role of commitment. Journal of Marketing Research, 37(2),

203-214.

Albert, N., & Merunka, D. (2013). The role of brand love in consumer-brand relationships.

Journal of Consumer Marketing, 30(3), 258-266.

Alnawas, I., & Altarifi, S. (2015). Exploring the role of brand identification and brand love in

generating higher levels of brand loyalty. Journal of Vacation Marketing, ,

1356766715604663.

Babić Rosario, A., Sotgiu, F., De Valck, K., & Bijmolt, T. H. (2016). The effect of electronic word of mouth on sales: A meta-analytic review of platform, product, and metric factors.

(42)

Batra, R., Ahuvia, A., & Bagozzi, R. P. (2012). Brand love. Journal of Marketing, 76(2),

1-16.

Berger, J. (2014). Word of mouth and interpersonal communication: A review and directions

for future research. Journal of Consumer Psychology, 24(4), 586-607.

Bourdieu, P. (1990). The logic of practice Stanford University Press.

Bowman, D., & Narayandas, D. (2001). Managing customer-initiated contacts with

manufacturers: The impact on share of category requirements and word-of-mouth

behavior. Journal of Marketing Research, 38(3), 281-297.

Brown, T. J., Barry, T. E., Dacin, P. A., & Gunst, R. F. (2005). Spreading the word:

Investigating antecedents of consumers’ positive word-of-mouth intentions and behaviors

in a retailing context. Journal of the Academy of Marketing Science, 33(2), 123-138.

Bryman, A., & Cramer, D. (2002). Quantitative data analysis with SPSS release 10 for

windows: A guide for social scientists Routledge.

Bughin, J., Doogan, J., & Vetvik, O. J. (2010). A new way to measure word-of-mouth

marketing. McKinsey Quarterly, 2, 113-116.

Carroll, B. A., & Ahuvia, A. C. (2006). Some antecedents and outcomes of brand love.

Marketing Letters, 17(2), 79-89.

Chang, A., Hsieh, S. H., & Tseng, T. H. (2013). Online brand community response to

negative brand events: The role of group eWOM. Internet Research, 23(4), 486-506.

Chardonnay. (2016). In Merriam-Webster’s online dictornary (13th edition). Retrieved from

http://www.merriam-webster.com/dictionary/chardonnay.

Chaudhuri, A., & Holbrook, M. B. (2001). The chain of effects from brand trust and brand

affect to brand performance: The role of brand loyalty. Journal of Marketing, 65(2),

(43)

Chen, Y., Wang, Q., & Xie, J. (2011). Online social interactions: A natural experiment on

word of mouth versus observational learning. Journal of Marketing Research, 48(2),

238-254.

Cheng, S. Y., White, T. B., & Chaplin, L. N. (2012). The effects of self-brand connections on

responses to brand failure: A new look at the consumer–brand relationship. Journal of

Consumer Psychology, 22(2), 280-288.

Chernev, A., Hamilton, R., & Gal, D. (2011). Competing for consumer identity: Limits to

self-expression and the perils of lifestyle branding. Journal of Marketing, 75(3), 66-82.

Cheung, C. M., & Thadani, D. R. (2012). The impact of electronic word-of-mouth

communication: A literature analysis and integrative model. Decision Support Systems,

54(1), 461-470.

Chin, W. W. (1998). The partial least squares approach to structural equation modeling.

Modern Methods for Business Research, 295(2), 295-336.

Chintagunta, P. K., Gopinath, S., & Venkataraman, S. (2010). The effects of online user

reviews on movie box office performance: Accounting for sequential rollout and

aggregation across local markets. Marketing Science, 29(5), 944-957.

Clark, H. H. (1996). Using language. 1996. Cambridge University Press: Cambridge), 952,

274-296.

Clark, M. S., Helgeson, V. S., Mickelson, K., & Pataki, S. P. (1994). Some cognitive

structures and processes relevant to relationship functioning. Handbook of Social

Cognition, 2, 189-237.

Cronbach, L. J. (1947). Test “reliability”: Its meaning and determination. Psychometrika,

12(1), 1-16.

(44)

Dion, K. K., & Dion, K. L. (1996). Cultural perspectives on romantic love. Personal

Relationships, 3(1), 5-17.

Drigotas, S. M., & Rusbult, C. E. (1992). Should I stay or should I go? A dependence model

of breakups. Journal of Personality and Social Psychology, 62(1), 62.

Dunteman, G. H. (1989). Principal components analysis Sage.

Dwyer, P. (2007). Measuring the value of electronic word of mouth and its impact in

consumer communities. Journal of Interactive Marketing, 21(2), 63-79.

Escalas, J. A. E. (1998). Narrative processing: Building connections between brands and the

self UMI.

Escalas, J. E. (1996). Narrative processing: Building connections between brands and the self,

unpublished ph. D Dissertation, Department of Marketing, Duke University, Durham,

NC, 27708

Fehr, B., & Russell, J. A. (1991). The concept of love viewed from a prototype perspective.

Journal of Personality and Social Psychology, 60(3), 425.

Fink, A. (2015). How to conduct surveys: A step-by-step guide Sage Publications.

Fiske, S. T., & Taylor, S. E. (2013). Social cognition: From brains to culture Sage.

Fong, G. T., & Markus, H. (1982). Self-schemas and judgments about others. Social

Cognition, 1(3), 191.

Fournier, S. (1998). Consumers and their brands: Developing relationship theory in consumer

research. Journal of Consumer Research, 24(4), 343-373.

Fournier, S. (2009). Lessons learned about consumers’ relationships with their brands.

Handbook of Brand Relationships, 5, 23.

Fournier, S., Breazeale, M., & Fetscherin, M. (2012). Consumer-brand relationships: Theory

(45)

Fullerton, G. (2005). The impact of brand commitment on loyalty to retail service brands.

Canadian Journal of Administrative Sciences/Revue Canadienne Des Sciences De l'Administration, 22(2), 97-110.

Godes, D., & Mayzlin, D. (2004). Using online conversations to study word-of-mouth

communication. Marketing Science, 23(4), 545-560.

Goffman, E. (2002). The presentation of self in everyday life. 1959. Garden City, NY,

Goyette, I., Ricard, L., Bergeron, J., & Marticotte, F. (2010). e‐ WOM scale: Word‐ of‐

mouth measurement scale for e‐ services context. Canadian Journal of Administrative

Sciences/Revue Canadienne Des Sciences De l'Administration, 27(1), 5-23.

Graeff, T. R. (1996). Using promotional messages to manage the effects of brand and

self-image on brand evaluations. Journal of Consumer Marketing, 13(3), 4-18.

Gruen, T. W., Osmonbekov, T., & Czaplewski, A. J. (2006). eWOM: The impact of

customer-to-customer online know-how exchange on customer value and loyalty.

Journal of Business Research, 59(4), 449-456.

Hennig‐ Thurau, T., Gwinner, K. P., Walsh, G., & Gremler, D. 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.

Henseler, J., Ringle, C. M., & Sinkovics, R. R. (2009). The use of partial least squares path

modeling in international marketing. Advances in International Marketing, 20(1),

277-319.

Hess, J., & Story, J. (2005). Trust-based commitment: Multidimensional consumer-brand

relationships. Journal of Consumer Marketing, 22(6), 313-322.

What’s love got to do with it? : a study on the effect of relationship types on eWOM intensity