Relationship marketing on social media : the effects of organizational response messages on consumer initiated messages via social media on consumers’ evaluations

(1)

Relationship Marketing on Social Media

The effects of organizational response messages on consumer initiated messages via social media on consumers’ evaluations

Master Thesis

Author: Marcella van den Berg (10083251)

University of Amsterdam, Faculty Economics and Business

Date: May 30, 2014

Supervisor: J. Demmers MSc

(2)

(3)

Relationship Marketing on Social Media – Master Thesis Business Studies 3

Abstract

Social media are integrated in consumer’s daily lives. This makes these media important for

organizations which need to know how they could benefit from these media. In this study the effects of organizational response on a consumer initiated message on that consumer’s attitude towards the organization and the consumer’s purchase intention at that organization are investigated by a quasi-experiment and field study. The context of this study was the online microblogging service Twitter. The results show that the presence of an organizational response on a positive consumer initiated message has a positive effect on the consumer’s attitude towards the organization. This effect is not found for purchase intention and for the negative and neutral consumer initiated messages. Results show that the response of the organization should be appropriate. No significant effects were found for the timing of the organizational response and the consumer’s expectation of the response. It was also found that the valence of the messages sent by consumers is positively correlated with the consumers’ evaluations attitude and purchase intentions. Implications of the results are discussed.

(4)

Inclusion and exclusion criteria Respondent approach Survey instrument Measurements Interrater reliability Results 33 Data preparation Testing hypotheses Conclusions 40 Discussion 42 Theoretical contributions Managerial implications Limitations References 46 Appendix 1 52 Appendix 2 55

(5)

Introduction

‘These Olympics will be the Socialympics – the first Olympics that will not only be massively followed on television and radio, but will be followed through Twitter and Facebook as well. Four years ago social media were still in their infancy: Twitter had six million users, now there are over 140 million. Facebook had 100 million users, and these days there are 900 million. Just a few people were having a smart phone, let alone a tablet computer.’

This is the introduction of an article entitled ‘The first socialympics’ from the website of a Dutch newspaper, nrc.nl (Ketelaar, 2012). And it is a brief summary of the enormous evolution of the social media Twitter and Facebook in the last four years. The Olympics are organized once every four years, the way they are described in the media reflects the era we live in. It also shows how some aspects, like the main communication channels, have been evolved. These Olympics can be marked as the social media Olympics, a new period in history.

What do social media mean for society? Ketelaar (2012) mentioned in her article that the user amount of Twitter has multiplied more than 23 times and Facebook has 900 million accounts, at the moment of writing the article. This means that social media are more and more integrated with people’s daily lives. When focusing on Twitter, worldwide 175 million tweets are sent every day (infographicslab.com, 2012). In 2009 Nielsen reported ‘two-thirds of the world’s internet population visits social networking or blogging sites, accounting for almost 10% of all Internet time’ (Nielsen, 2009). In the final three months of 2011, the Netherlands had the most active Twitter accounts in the world (Bowling, 2012).The underlying motives for contributing in online consumer platforms were investigated by Hennig-Thurau et al. (2004). They started collecting the reasons for joining the traditional word of mouth (WOM) process and noticed that the motivation for positive WOM behavior can differ from the ones of negative WOM behavior. Some of the identified reasons for consumers to contribute to the WOM process in literature are product-involvement,

(6)

Relationship Marketing on Social Media – Master Thesis Business Studies 6 of their study concerning online WOM (eWOM) cleared up the primary motivations of consumers for composing eWOM: social benefits, economic incentives, concern for others, and extraversion. Based on all these motivations, consumers get more and more used to share their lives with others online.

From a business point of view, the social media evolution is an interesting subject of research. The importance of social media as companies is not only expressed by the stock-market flotation of Facebook and Twitter, which are in that way companies like any others.

Besides companies, social media are relatively new communication channels for organizations in reaching their target group.

At a rough estimate consumers in the Netherlands see 377 advertising messages a day. When confrontation to packages and other brand images are counted, this number is around 1500

(Mirande, 2012). Consumers talk about brands both in an offline and online setting. In online brand communities consumers share their opinions and evaluations about brands and products. Besides the ability to communicate with each other this brand communities give to consumers, the content of these communities is visible for organizations and give them the opportunity to monitor the opinions of their consumers. Monitoring eWOM enables organizations to use this information in their continuous fight to get and keep the attention of the consumer over the enormous amount of brand impressions consumers have each day (Adjei, Noble, and Noble, 2010).

Marketing managers cannot control what is said about their brand on the internet. The internet (Pitt, Berthon, Watson, and Zinkhan, 2002) and social media give the consumer more power and the ability to communicate his opinion to a much bigger audience, than he could reach with the traditional media or without any device at all. For marketing managers this evolution means the implication of new communication strategies to reach their target group (Mangold and Faulds, 2009). These managers recognize a knowledge gap between how consumers react on traditional media and new media and to what extent investing in this media is turning a profit (Winer, 2009). One of the ways organizations try to benefit from social media is by monitoring what is said by consumers active on these media. In their cross-industry study Barnes and Jacobsen (2014) reveal that only the

(7)

non-Relationship Marketing on Social Media – Master Thesis Business Studies 7 profit sector shows widespread activity in monitoring eWOM. The next step after monitoring what is said by consumers is not clear at all.

Example 1

In this example the customer asks a question showing his dissatisfaction with the organization, which reacts on the same day in an

informative way and apologizes for the

inconvenience. Furthermore, the organization shows its empathy by understanding the disappointment of the customer, but they cannot change the procedures.

(8)

Relationship Marketing on Social Media – Master Thesis Business Studies 8 From the reactions of the customer it becomes clear that he does not accept the apologies and that he does not understand the reasons

the organization has given him as excuses for the problems.

As a reaction on the conversation between customer and organization other consumers give their opinion as well.

One day later the consumer is still angry with the

organization and asks if they will change the procedures today because this is the right time for customer care.

(9)

Relationship Marketing on Social Media – Master Thesis Business Studies 9 Example 2

In this example the customer asks the organization a question about the availability of products. Again, the

organization reacts in an informative way and unfortunately, the products are not available anymore. After this reaction the customer thanks the organization for the prompt response .

These examples illustrate the direct way organizations can reach their consumers. More than in case of the traditional media like television commercials and bill board advertising the organization is able to communicate with individual consumers directly. In the situations in the examples social media are an extension of the customer care department. The messages consumers share on social media are public and visible for other consumers. In the first example the consumer is negative about the company Blokker and other consumers react on his emotions expressed. This attribute of social media of being an open communication channel that is visible for all other consumers who are interested in it, is an example of the shift of power from the organizations to the consumers.

Consumers are not only the receivers of messages sent by organizations via mass media, but the communication power is more balanced these days (Sinclare and Vogus, 2011; Labrecque et al., 2013). More than ever communication between organizations and consumers is bi-directional. Before there were social media, the power of the consumer existed in the willingness to pay for the products or services of organizations. In their consumer empowerment research Labrecque et al. (2013) define this consumer power as demand-based power. This power still exists in the digital era,

(10)

Relationship Marketing on Social Media – Master Thesis Business Studies 10 but besides this power other power sources do exist. Other sources of consumer empowerment are information-based, network-based, and crowd-based power. Information-based power relates to the ease of accessing information and the production of consumer-generated content. In case of

network-based power the focus shifts from the self to others and this power source centers on enhancing the content (Labrecque et al., 2013). These two power sources are manifested in the examples. The consumer produces content on an online channel and in the first example other consumers react and enhance the content, on this channel. These examples reflect the power balance between consumers and organizations. With social media the power shifts to a situation in which organizations and consumer are conversation partners (Pitt et al.,2002; Sinclaire and Vogus, 2011).

In the first example the organization reacts and the emotions do not turn positive. In the second example the consumer asks a question and the organization reacts in the same proper way and as a consequence the consumer is thankful for this reaction and no other consumer is involved in the discussion. At Dutch organizations the webcare has grown since 2012. In 2013 organizations sent 66% more reactions on Twitter and 40% of the @-mentions at this medium received a response (Dietz, 2013).These are good examples of the absence of clear online marketing guidelines about how to react on consumer initiated messages on social media.

This study is about the consumer initiated messages consumers send on social media and the way organizations react on these messages. As argued by Mangold and Faulds (2009), consumers are reducing their reliance on traditional advertising as a source of information to guide their purchases. The internet and especially social media enable consumers to voice their opinion, both positive and negative (Sinclaire and Vogus, 2011). These media offer organizations the opportunity to infiltrate into consumer conversations that take place online (Mayzlin, 2006). The power consumers gain by voicing their opinion online is balanced in the way that organizations receive more information about their consumers (Labrecque et al., 2013). The volume of consumer messages online is big and the question rises to what extent organizations have to react on all these consumer generated content

(11)

Relationship Marketing on Social Media – Master Thesis Business Studies 11 on social media.

To investigate this topic the central research question of this study that will be answered is:

‘What is the influence of the presence of an organizational response on consumer initiated messages on social media on the consumer’s evaluations ‘attitude towards the organization’ and ‘purchase intention’?’

To get more insight in the moderating circumstances that can influence the revealed relationships some more research questions will be investigated in this study:

‘What is the moderating effect of the appropriateness of the organizational response in this relationship?’

‘What is the moderating effect of the consumers expectation of the organizational response in this relationship?’

To address these questions, a quasi-experiment will be executed. The context of this research will be the social medium Twitter. The study will be focused on reactions of organizations on consumer initiated messages and the extent in which the consumer appreciates this response. The potential elements that can have an effect on the consumers’ evaluations of the organization which will be included in this study are the consumer’s expectation and the appropriateness of the organizational response. The moderating effects of these variables on the relation between the presence of a response message from the organization and the consumers’ evaluations of the organizations will be analyzed.

This research contributes to the theoretical framework of relationship marketing. Personal contact through social media between organizations and consumers and the way this can contribute to more positive consumer evaluations are the main topics of this research. The theoretical relevance is related to the scope communication channels that is broadened by the rise of social media and the results of this study will give organizations more knowledge about how they can reach their

(12)

Relationship Marketing on Social Media – Master Thesis Business Studies 12 have to react on consumer initiated content on these media.

Practically this study is important, because social media become more and more important in influencing consumer evaluations (Mangold and Faulds, 2009), like purchase intentions and

attitudes. Managers are unable to get much guidance in how to integrate social media in their marketing strategies. To enhance the guidelines, more research concerning the effects of social media is needed. This study will be focusing on the effects for organizations of using social media in their communication with the consumer. The results of this study will gain an insight into the value of the time, energy, and money organizations invest in the web care activities using social media.

In this study, first the existing literature about relationship marketing, social media and organizations will be described. Then, the methods of this study will be dealt with, followed by the results of the research and the conclusions and discussion.

(13)

Literature and Hypotheses

To get more insight into how organizations’ social media communication efforts affect consumers’ evaluations attitude towards the organizations and purchase intention at the organization, we will discuss relevant literature which lead up to a set of hypotheses.

Relationship marketing

The effects of reacting as an organization on what is said by consumers about the organization online on consumer preferences are not yet fully revealed in existing literature. Some studies do have relevant results which will be used in order to create hypotheses about the specific use of social media for organizations. Brown et al. (2005) researched the antecedents of positive WOM intentions of consumers and conclude that interaction between organization and consumer leads to

commitment. A committed relationship between organization and consumer increases positive WOM behavior. This conclusion is in line with the existent theory about relationship marketing. Relationship marketing is a long term marketing strategy which is focused on attracting, maintaining, and enhancing customer relationships (Berry, 2002). From this marketing perspective acquiring new customers is only the first step of the strategy. If an organization is only focused on attracting new customers, whereas existing customers are not loyal, this strategy might be wasteful. Organizations which understand relationship marketing try to make their customers loyal customers with good service. In a time when consumers can easily switch from one service organization to another, it is important to protect the existing customer base (Berry, 2002). Since the rise of social media consumers are no longer the passive component in their relation with organizations.

Consumers are able to manage their vendors and due to this consumer empowerment organizations should focus on the increase of value both for the organization and for the consumers in managing

(14)

Relationship Marketing on Social Media – Master Thesis Business Studies 14 the consumer relationship (Malthouse et al., 2013) .

The effects of customers’ relationship with a service organization has been investigated by Hess, Ganesan, and Klein (2003). In order to minimize dissatisfaction and following the decrease of customer business it is important for organizations to understand the effects of failures and the ways to provide recovery. Organizations need to know which recovery is adequate for the consumer to prevent further damage. Results of their study confirm the supposed buffering role of customer relationship in protecting organizations from the negative effects of service failures on consumers’ satisfaction (Hess, Ganesan, and Klein, 2003). First this buffer effect of tolerance is explained by the long term expectations the consumer has due to the relationship. Because consumers expect an equity balance in the future, they do not ask for an immediate compensation. The second effect of customer relationship is the factor of stability of attributes. Consumers are convinced of the incidental attribute of the cause of the failure and will be higher satisfied after the recovery (Hess, Ganesan, and Klein, 2003).

Monitoring and reacting on consumers’ needs through social media, give organizations the opportunity to customize their relationships with the customers. Good customized relationships between organizations and consumer are specific to the requirements of one particular consumer (Berry, 2002). Jansen et al. (2009) pointed out the ease of contacting consumers for organizations with microblogging and its potential to enhance the customer relationship. In relation to Twitter, these conclusions promote an active Twitter strategy for organizations. In their study about the right Twitter strategy for organizations Burton and Soboleva (2011) conclude that the best strategy depends on the organizations strategy as a whole. Furthermore they emphasize the importance of more research on the opportunities Twitter open to organizations.

Word of mouth and electronic word of mouth

Product related conversations and their effect on consumer preferences, the word-of-mouth (WOM) process, has been subject for research for many years. Arndt (1976) explained this WOM process as the way consumers seek social support for the adoption and risk reduction of new products by group

(15)

Relationship Marketing on Social Media – Master Thesis Business Studies 15 action. The WOM process, in a commercial situation, contains consumers sharing their attitudes and opinions about products, services and organizations with others (Jansen, Zhang, Sobel and

Chowdury, 2009). Because consumers receive information from many different channels, the credibility of WOM makes this process important for organizations (Godes and Mayzlin, 2004) and makes WOM a powerful marketing channel for organizations in influencing consumers (Jansen et al., 2009).

The reason why WOM is so powerful is related to multiple factors. According to Keller (2007) the average American consumer participates in 121 WOM conversations every week. The enormous amount of WOM conversations is one of the reasons why WOM is so powerful in marketing. In his research Keller pointed out the polarity of WOM messages as well. 63% of the messages are positive and just 9% are negative. Another reason for the impact of WOM is the credibility of the messages according to the receivers because of the personal trust the conversation partners have

(Keller,2007). Bone (1995) agrees on the accessibility and credibility of WOM conversations as powerful aspects of WOM. WOM can influence several outcomes: awareness, expectation,

perceptions, attitudes, behavioral intentions and behavior (Buttle, 1998). WOM has multiple effects on both direct and delayed product judgments (Bone, 1995). In this way WOM does not only affect customers’ choice to buy a product, but the post-purchase perception of products as well (Gruen, Osmonbekov, and Czaplewski, 2005). Besides product judgments, WOM affects brand attitude (Lee, Rodgers, and Kim, 2009).

However, the effects of WOM can be moderated by several factors. In their study Herr, Kardes, and Kim (1991), investigated the moderating factors like prior impressions of the product available, the consumer’s memory and the presentation of extremely negative attributes of the product. They presented the valence of WOM and the vividness of presenting WOM information as direct antecedents of consumer persuasion. They concluded that negative WOM was more

informative for consumers than positive WOM in order to distinguish high or low quality products. The experiment of Herr, Kardes, and Kim (1991) also revealed the impact of vividness of the

(16)

Relationship Marketing on Social Media – Master Thesis Business Studies 16 information on judgment. Vividly shown messages influence product judgment more than pallidly shown messages do. The moderating variables in their experiment are the availability of a well-defined prior impression of the product in the memory of the consumer. The strong WOM effect reduces in case of the availability of this kind of memories. The effects of WOM disappeared when extremely negative product attributes were presented (Herr, Kardes, and Kim, 1991).

Since the internet and electronic devices dominate all our lives and take over several communication processes, WOM has had its electronical transformation as well (Tran, Strutton, and Taylor, 2012; Doh and Hwang, 2009). Electronic Word of mouth, eWOM, is the exchange of customer know-how online (Wu and Wang, 2011) and is a reliable source of information to consumers (Gruen et al., 2006). Due to the Internet, the scope of eWOM is much broader than that of WOM.

Consumers have more geographical en temporal freedom when using the internet to share their opinions about brands, organizations, and products. Furthermore, the written online WOM has more permanence than the traditional offline WOM (Jansen et al., 2009). Participating in eWOM gives the consumer more power (Labrecque et al., 2013), because he can reach not only his friend or family member by phone or face-to-face contact, he can reach every person who is interested in his life or in the topic he is talking about online. Marketing managers are no longer the only persons who decide what is said about their brand (Winer, 2009) thanks to this shift in power-balance.

Theories developed for WOM have been applied on the electronic process. Brown et al. (2005) do not even make a distinction between WOM and eWOM and purpose WOM as a communication process among individuals in person, or through some communication medium. Some of the effects of eWOM are attitude towards the product, attitude towards a website, and purchase intention (Doh and Hwang, 2009). Based on the results of their study Gruen, Osmonbekov, and Czaplewski (2006), conclude that eWOM significantly affects the perceived overall value of an organization’s offering. Moreover, they found both a direct relationship between consumer to consumer know-how exchange and loyalty intentions, as an indirect relationship mediated by the overall value of the organizations’ offering (Gruen, Osmonbekov, and Czaplewski, 2006).

(17)

Relationship Marketing on Social Media – Master Thesis Business Studies 17 Why do consumers engage in the eWOM process? In their research about what individuals drives to start and maintain blogs Nardi, Schiano, Gumbrecht, and Swartz (2004) took interviews. They found five major motivations for blogging: documenting life, expressing opinion and commentary,

expressing deeply felt emotions, designing ideas in words, and forming communities. Hennig-Thureau et al. (2004) also investigated the motivations of eWOM. They used a sample of 2000 consumers and concluded that ‘consumers’ desire for social interaction, desire for economic

incentives, their concern for other consumers and the potential to enhance their own self-worth’ are the main reasons for consumers for articulating their opinions through the internet (Hennig-Thurau et al., 2004). The aspects that is not represented in their study, is the one of having contact with an organization as a consumer. This aspect is relevant because an increased number of consumers know how to find organizations online. For example, in 2013 the Dutch consumer sent 56% more messages to organizations via the online microblogging medium Twitter than in the previous year (Dietz, 2013). To what extent consumers use social media and the internet in order to have contact with

organizations will be one of the main issues of the current study.

Hennig-Thurau et al. (2004) advise managers to enhance the community aspect of online platforms to encourage eWOM. This study is not focused on platform operators, but on the effects of the presence of organizational responses on consumer initiated messages on social media in order to increase the relationship marketing theory.

Attitude and purchase intention

In this study the focus will be on attitude towards the organization and purchase intention. The relation between WOM and these consumer evaluations have been studied for years and before the Internet era, Arndt (1976), concluded that a positive relation between WOM and purchase intention exists. Based on the results of more recent studies, the effect of eWOM and advertising on purchase intention is less clear. Jones, Aiken, and Boush (2009), reveal that eWOM and advertising alone is not

(18)

Relationship Marketing on Social Media – Master Thesis Business Studies 18 enough to change the purchase intention of someone who has negative prior experiences. The relationship between eWOM and attitude is positive. Positive online reviews enhance the attitude towards the brand (Lee, Rodgers, and Kim, 2009).

As concluded in previous studies, positive WOM contributed to purchase intention and negative WOM decreased the purchase probability (Arndt, 1967; Lee, Rodgers, and Kim, 2009). Conversely, extremely negative messages are very powerful (Herr, Kardes, and Kim, 1991). In more recent studies, negative messages have been seen as more influencing than positive messages and the impact on attitude towards the brand of negative eWOM is much greater (Park and Lee, 2007; Lee, Rodgers, and Kim, 2009). Doh and Hwang (2009) add to this conclusions that the effects of eWOM are more likely to appear to a consumer, when the eWOM messages are in the same direction, all positive or all negative. Moreover, they conclude that a few negative messages contribute to the credibility of the messages. They also suppose that negative messages are more powerful than positive ones, but more research is needed to support this statement. Because one single negative message between other messages which are positive, is not harmful but does increase the credibility of the eWOM (Doh and Hwang, 2009). A different conclusion is revealed based on the study of Jones, Aiken, and Boush (2009) about the effects of eWOM on purchase intention. They conclude that prior unfavorable experience is more powerful than eWOM and advertising alone to change someone’s purchase intention. However, based on previous research and literature, we suppose that a positive relation between eWOM and attitude and eWOM and purchase intention exists (H1).

H1: The valence of consumer initiated messages about an organization on social media is positively correlated with the consumers’ evaluations about that organization.

(19)

Failure recovery

However, this study is not focused on the effects of reading eWOM, but on the effects of articulating opinions and receiving, or not receiving, an organizational response on this eWOM message on consumer preferences. In their study about travel marketing Bronner and De Hoog (2011) revealed five motivation categories for eWOM contribution of consumers. Besides self-directed, helping others, social benefits and consumer empowerment, helping companies is a motivation for consumers to share their experiences online (Bronner and De Hoog, 2011). The feedback of

consumers is very important for organizations, because receiving consumer feedback enables them to correct their mistakes (Huppertz, 2007). To stimulate consumers’ complaining intentions it is more important to increase the benefits than to decrease the costs of complaining (Huppertz, 2007). Although, the Internet enables consumers to complain easily to an organization. The consumer need not go to the organizations office or store, or post a letter of complaint. By means of organizations activity on social media and enabling consumers to complain online, the costs to complain are decreased, but the benefits do not increase necessarily.

Negative information is more informative than positive information (Herr, Kardes, and Kim, 1991) and the effects of negative eWOM are stronger than the effects of positive eWOM (Arndt, 1967). For this reason we suppose the effects on consumer preferences in case of negative consumer messages on social media to be stronger than in case of positive consumer message.

Organizational response

Besides the facts that sender and receiver in the eWOM process are separated in space and time and that sender and receiver do not necessarily have a relationship, one of the biggest differences between traditional WOM and eWOM is the broader scope of consumers communication, due to the Internet (Steffes and Burgee, 2009). By using forums, microblogging services, or other channels of the Internet, consumers are more powerful in voicing their opinion (Labrecque et al., 2013).

(20)

Relationship Marketing on Social Media – Master Thesis Business Studies 20 Consumers are able to ‘operate’ on the same level as marketers of organizations. Bickart and

Schindler (2001) state that information presented on an online forum, may have a greater influence than information that is sent by marketing persons, based on the greater credibility, higher relevance level, and the generation of empathy among readers.

Based on their study about eWOM, as customer-to-customer know-how exchange, Gruen et al. (2006) conclude that this kind of information has more effect on consumer behavior than

marketing-designed information. Besides the description of eWOM as online knowledge exchange of consumers Wu and Wang (2011) add the attribute of eWOM messages being non-commercial and created by consumers. In their study about the difference in influence between WOM and eWOM in consumers’ decision making process about experience goods, Steffes and Burgee (2009) concluded that consumers do not rely on the strength of the relationship they have with the information source. So, the extent of knowing each other online does not hold in this situation. In all these studies the credibility aspect is related to all being consumers without a profit motivation. From this point of view, eWOM seems to be a process where only consumers are participating in.

As a contradiction, many studies about WOM involve source credibility as one of the factors that influence the effects of WOM. Based on the message source theory, Wu and Wang (2011) investigated the influence of source credibility of eWOM on attitude towards the brand and purchase intention. They revealed a positive relationship. Also Wangenheim and Bayon (2004) studied the factors that influence the effects of WOM and they pointed source expertise as one of the factors that influence both attitude and consumers decision making. This would make the organization itself the most credible and expert source of information in eWOM conversations.

H2:The presence of an organizational response on social media on a consumer initiated message is positively related to consumer evaluations attitude towards the organization and purchase

(21)

Timing of the organizational response

Based on his research about complaint management in the hospitality and tourism sector Davidow (2000) advises managers to pay attention to the speed of organizational response and the redress to the complainant. Timing and redress will influence the satisfaction and content of word of mouth (Davidow, 2000). In a later study, Davidow (2003) suggests the relevance of the method of complaining, phone, letter, or e-mail, as a topic to execute further research.

From a relationship perspective the presence of organizations on social media implicate an intimate and immediate conversation with consumers (Kietzmann et al., 2011). Because of the sharing and interaction characteristics of social media, Kaplan and Haenlein (2010) advise to engage in conversation on social media actively to develop relationships. Based on these conclusions we suppose a fast organizational response is a positive factor to influence consumers’ evaluations (H3).

H3: A fast organizational response on social media on a consumer initiated message is positively related to consumer evaluations attitude towards the organization and purchase intention at the organization.

Consumers’ expectation and perceived appropriateness of the organizational response

To complete the factors that influence the right Twitter strategy for organizations, a few moderating variables will be investigated. The aspects ‘perceived appropriateness of the organizational response’ and ‘consumers’ expectation of a response’ will be investigated in this study. In order to influence the attitude and purchase intentions of consumers by reacting on tweets of consumers as an organization, does it make sense that a consumer is expecting a reaction? And to what extent does

(22)

Relationship Marketing on Social Media – Master Thesis Business Studies 22 the content of the organizational response make sense in the consumers’ attitude and purchase intention?

Expectation

To what extent does expectation of the organizational response influence the relationship between the presence of the organizational response and the consumers’ evaluations attitude and purchase intention?

Expectation has been subject of research for a long time. In Oliver’s (1980) research on the antecedents and consequences of satisfaction in a framework based on the previous literature on the subject is investigated. He supports the existing theory of the positive relationship between the performance / expectation ratio and satisfaction. This means confirmation to the expectation will lead to satisfaction. Besides this relationship Oliver (1980) reveals the effects of expectation on other ‘post exposure elements’ attitude and purchase intention. Satisfaction is positively related to

purchase intention (Williams and Naumann, 2011).

In a recent study Kietzmann and Canhoto (2013) investigated the social media implications for public affairs professionals. Meeting or exceeding expectations is an important antecedent of consumer eWOM and affects satisfaction. They used the disconfirmation theory of Oliver (1980) as underlying framework. About the consumer’ expectation of an organizational response on the consumer initiated messages, they conclude the consumer expectancy of firms using social media platforms efficiently (Kietzman and Conhoto, 2013). Based on this conclusions the moderating role of expectation in the relation between organizational response and customer’s attitude and purchase intention will be investigated in this study.

(23)

H4: Confirmation to consumers’ expectation of an organizational response message is a positive moderator in the relationship of organizational response and the consumer’s evaluation.

Consumer’s perceived appropriateness of the organizational response

In a time consumers wrote letters in order to praise a company or to lodge complaints, Smart and Martin (1993) investigated the way organizations have to handle these letters to have satisfied consumers. They concluded every support is needed to correspond with consumers in an appropriate way. The increase of interactivity makes communication a more valuable element of marketing (Duncan and Moriarty, 1998). In their development of a communication-based marketing model based on brand relationship Duncan and Moriarty (1998) underline the importance of the integration perspective in managing all sorts of brand communications in a cohesive way. Integrated marketing communication has one voice no matter which marketing function or media platform is used. With this strategic consistency messages are appropriate for consumers.

Other aspects of appropriate messages are investigated by Haghirian, Madlberger, and Tanuskova (2005) whit the advent of advertising on mobile devices. They point out the individual way of reaching the consumer as one of the big opportunities mobile advertising give to organizations. Personalization and interactivity in the communication between the organization and the consumer is important in order to send appropriate messages and enhance the consumer-organization relationship.

For the relations in this study a positive moderating role of consumers perceived appropriateness of the organizational response message on relation between presence of the organizational response and consumer evaluations attitude and purchase intentions is expected.

(24)

H5: Consumers’ perceived appropriateness of the organizational response is a positive moderator in the relationship between the presence of an organizational response and consumers’

evaluations.

The hypotheses are visualized in two models. The first model is related to the correlation between the valence of consumer initiated messages and consumer evaluations attitude and purchase intentions. The second model is related to the relationship between the organizational response and consumer evaluations attitude and purchase intentions. Furthermore it reflects the moderating effects of consumers’ expectation and perceived appropriateness of the organizational response.

Figure 1.1 Hypothesis 1

(25)

Methodology

Design

In this study the impact of organizational response to consumer eWOM on consumers’ attitude and purchase intention is investigated.The frame and communication channel in which this impact is tested is the online microblogging medium Twitter. To investigate the impact for organizations of reacting on eWOM via social media a quasi-experimental design is executed based on a

questionnaire. The data were collected in a quantitative way.

The independent variables in this study are valence of the consumer eWOM message on social media and the presence of the organizational response on this messages. The consumers attitude towards the brand or organization and the consumers purchase intention are the dependent variables of this research. Moderating variables in this study are the consumers expectation of a response and the perceived appropriateness of the organizational response. Furthermore, the aspect of the timing of the organizational response is investigated.

Respondents were approached via Twitter not more than one week after they sent a tweet to the organization. This time could not be longer in order to reduce the chance respondents have forgotten what kind of message they had sent to the organization and if they received a response on that message from the organization. Another reason for this period was that the organization needed to get time to respond. If the questionnaire would have been sent earlier, the data could get biased. The questions that were asked to the respondents who join the survey depended on whether they received an organizational response or not. Further manipulations were not made. The respondents were approached in a real time situation. For this reason this research is a field study and a quasi-experiment.

(26)

Context: Twitter

The particular eWOM channels that will be investigated in this study are social media. Similar to the traditional media, there are various kinds of social media that facilitate the conversation between consumers and between consumer and organization. The context in which this study will take place is the online microblogging service Twitter. Twitter started in July 2006 and is a social medium that allows people to send small messages online, microblogging. Like Sinclaire and Vogus (2011) mentioned, Twitter is simple, because it only requires a smartphone and it differs from other social media because it is a real-time broadcast platform. The small messages that can be published by this platform are called ‘Tweets’. Tweets contain 140 signs at maximum. People who follow you

(followers) and people who search for you or search for the subjects you have sent messages about, can see these tweets and can react to it by sending a reply. It is possible to open an account both as an individual and as an organization. At the beginning of 2012 Twitter has over 465 million accounts and 175 million tweets are sent every day (infographicslab.com, 2012).

The reasons why Twitter is the right context to investigate the effects of the presence of an organizational response on consumer initiated online messages on the consumers’ attitude towards that organization and purchase intention at that organization are the following. First of all,

microblogging is a tool for consumers to share word of mouth (Jansen et al. 2009). Twitter is probably one of the most extreme channels on which consumers and organizations are present and are able to interact with each other. With an active twitter account from both sides, direct

communication is possible. In this way consumers can have online conversations with an

organization, just like he or she could have with friends, family and other consumers. In the second place, the fact that the accounts are open and the messages of consumers and organizations are visible for everyone, makes it possible to read the consumers’ messages and to approach them to take part in the study. Furthermore, the Netherlands have the most active Twitter accounts (Bowling, 2012). In their study about the right Twitter strategy for organizations Burton and Soboleva (2011)

(27)

Relationship Marketing on Social Media – Master Thesis Business Studies 27 conclude that the best strategy depends on the organizations strategy as a whole. Furthermore they emphasize the importance of more research on the opportunities Twitter open to organizations.

Data collection

In order to investigate the impact of reactions of organizations via Twitter on consumer initiated tweets, respondents were approached via this medium. This is a logical consequence of the subject of this survey: Twitter. When the consumer sent a tweet that contains the @-sign followed by the name of the organization, he could be approached for this study. In Twitter the @-signs before the username of an organization implies, that the consumer is ‘talking’ to the organization directly. To address a message to somebody on Twitter the @-sign followed by the username of that person or company has to be mentioned in the message (Lovejoy, Waters, and Saxton, 2012). If the consumers uses the @-sign in his message to the organization the organization would see this and could decide to react or not. The messages are public, which means that others who are following the consumer or the organization on Twitter are able to see the message as well (Lovejoy, Waters, and Saxton, 2012).

Apart from the content of the tweets about an organization, the @-sign makes messages more easily comparable. The content could be negative, positive, or neutral and did not necessarily need to contain a question. Not reacting on a comment or question sent by consumers is comparable with ignoring a client. Reacting implies responsibility and establishment of the dialogue between organization and consumer (Lovejoy, Waters, and Saxton, 2012).

Inclusion and exclusion criteria

For this study we had some criteria to include or exclude consumer initiated messages. By consumer-initiated, is meant that there was no Twitter marketing action from the organization on which the

(28)

Relationship Marketing on Social Media – Master Thesis Business Studies 28 consumer reacted. The message should come voluntarily from the consumer as far as we could ascertain. Both male and female respondents were included. When the consumer mentioned that he had a work relation with the organization, for example ‘I will go to my work @Hema’, this consumer was not approached to participate in this study. The reason for excluding this consumers is that they are employees beside consumer and that could influence the answers of this consumer.

Furthermore, consumers who did mention more than one @-sign followed by multiple organizations in their message were not approached, because it was not possible to distinguish the attitude and purchase intentions towards more than one organization in the invitation message and in the questionnaire.

The survey was executed in the Netherlands, so only Dutch tweets have been included. The following organizations are part of this study: Hi, Vodafone NL, Blokker, Wehkamp, Hunkemoller, Ikea, Coffeecompany, Gamma, Nespresso, Hema, Albert Heijn, Jumbo, C1000, MediaMarkt, Philips, Landal Netherlands, De Bijenkorf, Etos and KNVB. All these brands or organizations are active in the Netherlands. Some of them are competitors and active in the same industry. To avoid a one-sided result, organizations from different industries are included. Data were collected based on the corporate Twitter accounts, or the Twitter account of the organization in the Netherlands. Accounts of single affiliates were not taken into account.

Because anyone with a Twitter account has full access to all the tweets, cooperation with the companies included in this study was not necessary. They were not asked to be part of this study.

Respondent approach

After taking the inclusion and exclusion criteria into account, respondents were approached

randomly. To collect respondents, tweets were used. Consumers’ attention was obtained by sending them a personal tweet, using the @-sign followed by their username. The tweets were in Dutch, but

(29)

Relationship Marketing on Social Media – Master Thesis Business Studies 29 the text contains this content:

Because of the amount of words, the text was split into two tweets. To inform the respondent about which organization they have to fill in the questionnaire the name of that organization was

mentioned in the invitation tweet.

All respondents were lead to the same url. The path of the questionnaire depended on the answer they gave, related to the situation of getting or not getting a response from the organization. Software that has been used to diffuse the questionnaire and to collect the data is qualitrics.com.

Survey instrument

Based on the hypotheses the questionnaire of this study has been set up. The complete questionnaires for respondents who received and who did not receive a reaction from the organization are attached in Appendix 1 and Appendix 2.

When respondents opened the questionnaire through the url, the introduction text was shown. This text gave information about the study and explained why respondents were approached to fill in this questionnaire, the object of the survey, and the approximate time needed to answer the

(30)

Relationship Marketing on Social Media – Master Thesis Business Studies 30 questions. Furthermore the introduction text described the importance of the respondents’ opinion, the fact that all answers are good, that no questions could be passed, and that data would be treated confidentially.

To match the tweets and the results of the questionnaire we asked about the name and Twitter name, to have a double check. Date of birth and gender were registered. Whether or not the organization had responded was asked as well. After the questions about the dependent and

moderating variables respondents were asked about their future intentions to use Twitter in order to have contact with the organization.

Measurements

This study includes two dependent variables and two moderating variables. The dependent variables attitude towards the organization and purchase intention are related to consumer evaluations. By these variables effects of an active use of social media are measured.

Attitude towards the organization. Attitude towards the organization was measured using a 3-item

scale. It shows how a respondent thinks about the organization to which a tweet was sent. The question was ‘How do you think about the organization which you have contacted via Twitter?’. Three scales were showed. ‘Very bad to very good’, ‘Not very nice to very nice’, and ‘Very negative to very positive’. A seven point Likert scale was used (Mitchell, 1986).

Purchase intention. The dependent variable purchase intention was measured by the question ‘How likely is it that you will buy the products or use the service of this company in the future?’ the answers varied from very unlikely to very likely on a seven point Likert scale (Baker and Churchill, 1977).

(31)

Relationship Marketing on Social Media – Master Thesis Business Studies 31 The moderating variables that were measured in this study are expectancy of a response of the organization when sending a tweet and the perceived appropriateness of the organizational response tweet.

Expectation. Depending on the presence of a response of the organization the questionnaire path

differed. In both cases, first the moderating variable expectation was investigated. Two questions were used to test this variable: ‘I am surprised the company reacted on my message’ and ‘I expected the company to react on my message’. The seven point Likert scale was used containing answers varying from ‘totally disagree’ to ‘totally agree’. Like in the study of Oliver (1977) this is a straight forward manner to test the post exposure expectation.

Perceived appropriateness of the organizational response message. The respondents perceived

appropriateness towards the message they received from the organization on social media as a reaction on their own message has been measured by a three item scale. This scale is based on the attitude towards the message construct. The question was: ‘What is your judgment of the

companies reaction on Twitter that you got?’. The three items contained seven point Likert scales from ‘very bad’ to ‘very good’, from ‘very unkind’ to ‘very kind’, and from ‘very negative’ to ‘very positive’(Mitchell, 1986).

Timing. Like in the study of Wirtz and Mattila (2004) timing of the organizational response was

measured by a two item scale. The content of the items was adapted to the current study, because this research is not related to service failure like the research of Wirtz and Mattila (2004). ‘What do you think about the reaction time of the company?’. The answers could vary from ‘Not in time’ to ‘Exactly on time’ and from ‘Too slow’ to ‘Too fast’. For these items the seven point Liker scale is used too.

(32)

Interrater reliability

Consumer tweets were coded based on the content in terms of their valence, which reflects the degree of positivity of the eWOM. Interrater reliability is an important parameter to ensure credibility (Moreno, Kelleher, and Pumper, 2013). Interrater reliability was covered by the

assessment of the tweets by five coders, who were not previously informed about the goal of this study. In the current study the Cronbach alpha coefficient was .97.

(33)

Results

The survey was framed online in Qualtrics. The link of the questionnaire is as a field study spread through Twitter via the personal Twitter account of the researcher. 527 consumers were approached and this resulted in a dataset of 331 respondents. 38.6 % did not fill in the questionnaire completely and results from these questionnaires were not useful and deleted. Respondents with an age below 16 years old were excluded. The final dataset contained 200 respondents.

Data preparation

To help to prevent response bias, some items of the questionnaire have been reversed. Before the calculation of the reliability of the scales the negatively worded items needed to be reversed. One item of the expectancy scale.

To test the reliability of each scale Cronbach’s Alpha is used. A Cronbach’s Alpha score above 0,7 is considered to be reliable (Cortina, 1993). The expectancy scale of the respondents who did not receive a response of the organization and the timing scale were not reliable.

Sample profile

The dataset consisted 43,5% man and the mean age of the total dataset was 34,30 (SD: 9,25). The overall range of the age of the respondents was from 16 to 60 years old. One respondents did not fill in his age. 53,0% of the respondents received a response message from the organization they had sent a message to. There was no selection in gender when the respondents were approached to join this study.

(34)

Table 4.1: Scale reliability

Cronbach’s alpha Items in scale Eliminated items Valence 0,972 5 0 Expectancy, response 0,763 2 0 Expectancy, no response 0,904 2 0 Perceived appropriateness 0,918 3 0 Attitude towards company, response 0,944 3 0 Attitude towards company, no response 0,927 3 0

Timing 0,652 2 0

The greatest part of the sent messages by consumers was negative 46,5%. 35,5% of the consumer messages was positive and 18,0% was considered to be neutral. There was no significant age difference between the degree of positivity of the sent messages by consumers (F (2,196) =,75 , p =.48).

The Twitter intentions are positive. 65% of the respondents will contact the organizations again through Twitter in the future. This is more than the percentage of respondents who received a response on their tweet.

Testing hypotheses

Several tests are executed in the statistical program SPSS to test the hypotheses. Below the results of these tests are described.

H1: The valence of consumer initiated messages about an organization on social media is positively correlated with the consumers’ evaluations about that organization.

(35)

Relationship Marketing on Social Media – Master Thesis Business Studies 35 A one-way between-groups analysis of variance was conducted to investigate the impact of the valence of consumer eWOM messages on consumers evaluations’ attitude towards the organization and purchase intentions at the organization. Valence was divided in positive, neutral, and negative. There was a significant difference at the p < .05 level in the attitude scores for valence: F(2,197) = 10,77, p = .00. The effect size, using the eta squared calculation, was medium 0.9. Mean attitude score for the negative valence group (M= 4.49 , SD= 1.30) was significantly different from the positive valence group (M= 5.40, SD= .98) and the neutral valence respondent group (M= 5.18, SD= 0.96). The neutral valence group did not significantly differ from the positive valence group.

There was a significant difference at the p < .05 level in the scores for purchase intentions as well: F(2,197)= 7,91, p=.00. Eta squared was used to calculate the effect size and was medium .07. Mean purchase intentions score for the negative valence group (M=5.08, SD=1.88) significantly differed from the positive valence group (M= 5.93, SD= 1.22) and the neutral valence group (M=5.97, SD= 1.06). The purchase intentions scores of the neutral and positive group did not differ

significantly.

Figure 4.1 : Hypothesis 1 is supported for both attitude and purchase intention.

4 4,5 5 5,5 6 6,5 7

Negative valence Neutral valence Positve valence

(36)

Relationship Marketing on Social Media – Master Thesis Business Studies 36 Higher mean scores for consumer’s attitude and purchase intentions in case of neutral and positive consumer initiated messages than in case of negative consumer initiated messages. Hypothesis 1 is supported for both attitude and purchase intention.

To consider the influence of the presence of organizational response on consumer messages on social media on the consumer’s attitude towards the company and the consumer’s purchase intention, this study tests the reaction of organizations on consumer tweets on Twitter. The field study resulted into six categories, summarized in the table below. The General Linear Model test of the statistical program SPSS is used to test the hypotheses.

Table 4.2 Descriptive statistics

Negative eWOM Attitude Purchase Intention

Mean SD Mean SD Cell Size

Presence Org. Response No 4.42 1.46 5.19 1.76 42 Yes 4.73 1.15 4.98 1.98 51 Total 4.59 1.30 5.08 1.88 93 Timing Org. Response Same day 4.67 1.16 4.80 2.08 25 Later 4.79 1.19 5.28 1.84 25 Total 4.73 1.16 5.04 1.96 50

Neutral eWOM Attitude Purchase Intention

Presence Org. Response No 4.90 0.85 5.70 1.25 10 Yes 5.29 1.00 6.08 0.98 26 Total 5.18 0.98 5.97 1.06 36 Timing Org. Response Same day 5.20 1.02 6.08 0.76 13 Later 5.64 0.86 6.09 1.30 11 Total 5.40 0.96 6.08 1.02 24

Positive eWOM Attitude Purchase Intention

Presence Org. Response No _Yes 5.13 _5.79 0.97 _0.87 5.88 _6.00 1.17 _1.31 42 ₂₉ Total 5.40 0.98 5.93 1.22 71 Timing Org. Response Same day 6.19 0.74 6.50 0.80 12 Later 5.51 0.77 5.85 1.52 13 Total 5.84 0.82 6.16 1.25 25

(37)

H2:The presence of an organizational response on social media on a consumer initiated message is positively related to consumer evaluations attitude towards the organization and purchase

intention at the organization.

This hypothesis is tested in three situations. The response of the organization on negative, neutral, and positive consumer initiated messages. In case of a negative (F(1,91)=1,32, p= .25, M=4.73, SD=1.15) and a neutral (F(1,34)=1,23, p= .28, M=5.29, SD= 1.00) consumer initiated message, the presence of an organizational response does not significantly influence the consumer’s attitude. If the consumer initiated message is positive, the presence of an organizational response does significantly influence the consumer’s attitude (F(1,69)=8,78, p= .00, M=5.79, SD=0.87).

For all sorts of consumer initiated messages, negative (F(1,91)=,27, p=.59, M=4.98, SD= 1.98), neutral (F(1,34)=,92,p=.34, M=6.08, SD=.98), and positive (F(1,69)=,16, p=.69, M=6.00, SD=1.31), the presence of an organizational response does not influence the consumer’s purchase intentions.

Hypothesis 2 is partly supported. The only situation in which the presence of an

organizational response does positively influence the consumer’s attitude towards the organization, is when the consumer sent a positive message.

H3: A fast organizational response on social media on a consumer initiated message is positively related to consumer evaluations attitude towards the organization and purchase intention at the organization.

Because the construct of timing was not reliable in this study, the test of the influence of a fast organizational response is executed based on the real time the organization used to react on the consumer’s message. Responding on the same day and responding later than on the same day were distinguished. The timing of the organizational response does not significantly influence the

consumer’s attitude. Not for the negative (F(1,46)=1,15, p=.29, M=4.67, SD=1.16), the neutral (F(1,20)=3,69, p=.07, M=5.20, SD=1.02), and the positive (F(1,21)=1,26, p=.28, M=6.19, SD=1.26) a fast response is better in order to influence the consumer’s attitude.

(38)

Relationship Marketing on Social Media – Master Thesis Business Studies 38 The consumer’s purchase intention is not significantly influences by a fast organizational response as well. These results are for negative(F(1,46)=2,08, p=.16, M=4.80,SD=2.08), neutral (F(1,20)=,00, p=.99, M=6.08, SD=.76), and positive (F(1,21)=,16, p=.69, M=6.50, SD=.80) consumer initiated messages. Hypothesis 3 is not supported.

Table 4.3 Results

F-value

p-value

Negative Independent variables Dependent variables Presence of the organizational response Attitude 1,32 .25 Purchase intention 0,27 .59 Timing of the organizational response Attitude 1,15 .29 Purchase intention 2,08 .16

Covariates

Consumer's expectation Attitude 0,03 .87 Purchase intention 2,83 .10 Consumer's perceived appropriateness Attitude 44,58 .00

Purchase intention 11,52 .00

Neutral Independent variables

Presence of the organizational response Attitude 1,23 .28 Purchase intention 0,92 .59 Timing of the organizational response Attitude 3,69 .07 Purchase intention 0,00 .99

Covariates

Consumer’s expectation Attitude 0,05 .28- Purchase intention 3,99 .06 Consumer's perceived appropriateness Attitude 7,62 .01 Purchase intention 1,2 .29 Positive Independent variables Presence of the organizational response Attitude 8,78 .00 Purchase intention 0,16 .34 Timing of the organizational response Attitude 1,26 .28 Purchase intention 0,16 .69

Covariates

Consumer’s expectation Attitude 0,03 .15- Purchase intention 0,62 .44 Consumer's perceived appropriateness Attitude 17,96 .00 Purchase intention 3,08 .09

(39)

H4: Confirmation to consumers’ expectation of an organizational response message is a positive moderator in the relationship of organizational response and the consumer’s evaluation.

The covariate consumer’s expectation of an organizational response does not influence the relationship between the presence of an organizational response and the consumer’s evaluations attitude towards the organization and the purchase intentions at that organization. Significance levels do all exceed the p=.05 level (Table 4.3). Hypothesis 4 is not supported.

H5: Consumer’s perceived appropriateness of the organizational response is a positive moderator in the relationship between the presence of an organizational response and consumers’

evaluations.

The covariate consumer’s perceived appropriateness of the organizational response, was significantly positive related to the consumer’s attitude, in case of a negative consumer initiated message,

F(1,46)=44,58, p=.00, in case of a neutral consumer message, F(1,20)=7,62, p=.01, and in case of a positive consumer initiated message, F(1,21)=17,96, p=.00.

In case of a negative consumer initiated message the covariate consumer’s perceived appropriateness of the organizational response is significantly positive related to the consumer’s purchase intention at the organization, F(1,46)=11,52, p=.00. When the consumer sent a neutral, F(1,20)=1,20, p=.29, or positive message, F(1,21)=3,08, p=.09, this relationship was not found. Hypothesis 5 is partly supported.

(40)

Conclusions

Social media are integrated in the daily lives of consumers and via these media consumers are able to share their opinions with everyone who is interested in these opinions. The consequences this development has for organizations had in terms of relationship marketing and existing electronic Word of Mouth (eWOM) theory are investigated in this study.

From a relationship marketing perspective the long term relationship between organization and consumer is important and the organization should be focused on increased value for both organization and consumer. The extent in which an active social media strategy could affect this goal focused on the consumer’s attitude towards the organization and the consumer’s purchase intention at the organization this study investigated the impact of organizational response on consumer initiated messages on social media.

Valence of the consumer initiated message. First of all the direct correlation between the

valence of the consumer initiated message about an organization on social media and the consumer evaluations attitude towards that organization and purchase intention at that organization were investigated. These two variables are positively correlated, which means that a consumer who writes a positive message on a social medium about an organization has a positive attitude and has a higher purchase intention than a consumer who shares negative messages on social media.

This result shows that the negativity or positivity of the consumer will not be easily changed. The antecedents of using social media may be more important for consumers than the consequences and the evaluations of the organization after spreading his opinion. For the organization it is

important to focus on the reasons why a consumer spreads his negative, neutral, or positive feelings or opinions about the organization and change the situation for the future if this is needed in order to maintain the long term relationship with the consumer.

Presence of the organizational response. The central research question of this study was

(41)

Relationship Marketing on Social Media – Master Thesis Business Studies 41 on social media on the consumer’s evaluations ‘attitude towards the organization’ and ‘purchase intention’?’ Because of the lack of literature about whether or not organizations should infiltrate in the eWOM process of consumers by reacting on what is said about the organization on social media, the impact of these organizational responses is investigated. The results show a positive effect of the presence of an organizational response on consumer’s attitude in the situation that the consumer sent a positive message. Effects of responding on neutral or negative consumer initiated messages are not found.

Timing of the organizational response. In order to create the most value for organization and

consumer out of the use of social media, the importance of a fast organizational response was tested. The timing of the organizational response was distinguished as ‘the organization gave a reaction at the consumer’s message at the same day’ or ‘the organization gave its reaction later than on the same day that it received a message from the consumer via social media’. The results show no impact of the timing of the organizational response.

Covariates consumer’s expectation and perceived appropriateness of the organizational response. Results show that the organizational response should be appropriate in that situation of a

positive consumer message and a present organizational response in order to influence the consumer’s attitude towards the organization. Appropriateness is in other situations a significant influencer of the tested consumers’ evaluations, but the presence of the organizational response in other situations is not relevant. Results do not show an effect of the consumer’s expectation of an organizational response.

The scope of the results of this study is the Netherlands, so the conclusions of this study are the most applicable to Dutch organizations. It is possible that the same study will lead to other results in other countries.

(42)

Discussion

The objective of this study was to expand the existing relationship marketing and eWOM literature in order to achieve a better understanding of the way organizations should react on consumer initiated messages on social media to influence consumers’ evaluations about the organization. The results of this study reflect the enormous capacity consumers have for voicing their opinion online. Negative, neutral, and positive messages appear on the internet and give a lot of information about the sentiment of consumers about an organization. The theoretical contributions, the managerial implications, the limitations of this study and suggestions for further research will be discussed.

Theoretical contributions

From the relationship marketing perspective monitoring the consumer initiated message is valuable in order to understand the consumer to be able to customize the relationship between organization and consumer (Berry, 2002). Like Jansen et al. (2009) concluded, the results of this study reflect the opportunities social media offer organizations to contact consumers and to enhance the relationship with customers. The contribution of this study is that the results provide more information about the specific situation in which the consumer is the first party who talks to the organization. The only situation in which the results of this study confirm positive effects of an organizational response on the consumer’s attitude is when the consumer sent a positive message. A customized reaction on a positive message from the consumer is appreciated by the consumer.

From the eWOM theory perspective not only the consumer is empowered by the Internet and social media (Labrecque et al., 2013), the direct opportunities to read what is said about the organization and to react on this content to the consumer who did share the content is valuable for the organization as well.

Relationship marketing on social media : the effects of organizational response messages on consumer initiated messages via social media on consumers’ evaluations