Influencer Marketing -
Credibility of eWOM in Social Network Sites
A comparison between Twitter and FacebookMaster Thesis December 2013
Linda Maas s1700847
University of Groningen Faculty of Economics and Business Msc Business Administration Marketing Management & Marketing Research
Abstract The purpose of this research is to research the credibility of Word of Mouth on Social Network Sites. There will be an comparison between Twitter and Facebook and which factors influence the credibility. Other research already showed results of the credibility of electronic Word of Mouth, this research will extend existing research by specifically discuss the eWOM on the most popular Social Network Sites.
2
Abbreviation Spelled Definition
SNS Social Network Site An online platform that offers the service to build an own social network with social
relations between people.
Based on definition Deepa, Chopade, Ranjith (2012)
SNSs Social Network Sites Online platforms (in general) that offer the service to build an own social network with social relations between people.
Based on definition Deepa, Chopade, Ranjith (2012)
WOM Word of Mouth Positive or negative oral communication between people without a commercial intention about a brand, product, service or organization.
Based on definition Arndt (1967)
eWOM Electronic Word of Mouth Positive or negative communication between people without a commercial intention about a brand, product, service or organisations which is published on the internet.
Based on definition Hennig-Thurau et al. (2004)
3
Management summary
Every customer has a voice these days on the internet. When a customer is dissatisfied it is easy to let this inconvenience show on a SNS such as Facebook or Twitter. With this in mind and the beliefs that influencer marketing will become more and more important. Brands should be looking for people who are influential within the marketing in which the brands operates. Furthermore it is important to know what kind of messages are perceived as credible on the SNSs.
Understanding how Social Network Site users perceive the credibility of the posted messages is important for organizations which want to participate in the SNSs. Twitter and Facebook are two different SNSs and therefore a message with the same meaning could be received in in other way across the different platforms. For that reason different determinants like argument strength, source credibility and confirmation with prior belief are taken into account in the different research questions.
To be able to answer the research question, data is collected via an online questionnaire. In total 309 respondents with a Twitter or Facebook account filled in the questionnaire.
Factor analysis and several regression analyses are performed in order to research if perceived eWOM credibility and eWOM adoption differ on Twitter and Facebook. After the analysing and accepting of rejecting the hypotheses, the main question could be answered. The research showed that there are some differences in the eWOM credibility and eWOM adoption on Twitter and Facebook. Although there may some differences in the relative importance of the variables, this research showed that perceived eWOM credibility on both Twitter and Facebook have a positive effect on eWOM review adoption. Furthermore the determinants have shown to have a positive relation on both Twitter and Facebook regarding the eWOM credibility.
4
Preface
After a hectic few years I am happy to present my marketing master thesis which is part of my master degree in Business Administration.
I started my study Marketing Research in Groningen to increase marketing research knowledge. The study thought me far more than I could imagine before starting it. Furthermore I am very glad that I had the possibility to add Marketing Management to my study, because the management side of marketing has also become one of my passions. I am very interested in Social Network Sites such as Facebook, Twitter and Google+, and also Word of Mouth has my interest. I hope this passion about these topics is well reflected within this thesis.
In special I would like to thank my supervisor, dr. J.E.M. van Nierop, for the support and understanding of my situation. Because of him I was able to finish my thesis before the end of this year. Erjen, thank you very much for your patience, reflection, fast feedback and help. Furthermore I would like to thank my second supervisor E. de Haan who was able in such a short time to act as my second supervisor.
5 Why did I choose to write this thesis?
Nowadays we live in a world where everybody can be heard. Unlike the past, a dissatisfied customer is able to let everybody in the world show his inconvenience with a certain brand or company. Maybe you have been in that place where you want to out your frustrations or feelings of injustice with as many people as possible. The world wide web is just a mouse click away.
Because of my personal interest in both Word of Mouth (WOM) and Social Network Sites (SNSs), I was searching for a topic in which I could combine these two points of interest. In this paper the credibility of electronic word of mouth (eWOM) in SNSs is discussed. The motivation for choosing Twitter as one of the SNSs is because of the limited amount of text. I am interested in what the space for a message means for the credibility of a message. Because of the limited space, Twitter users need to abridge their message like a headline. With this paper I investigate whether people believe such a strong message or that people need more text and motivations. Furthermore it is interesting to investigate the differences among the different SNSs. Where Facebook is mainly used for friends, Twitter seems to be more “open”. In a time where the SNSs pop up and organisations are wondering which SNS they should integrate into the communication strategy, it will be helpful to know what determinants influences the credibility.
With 2014 in sight, trend watchers and marketers are carefully looking at the trends for the following year. Jan Willem Alphenaar, trend watcher and marketer, promotes that in 2014 Influencer Marketing will become (more) important. Brands should be specifically looking for people who are influential within the market in which the brand operates. The number of connections is not important, but it must be the right connections according Alphenaar. The influencer must however be transparent about the cooperation to create acceptation within the network. Thus, in 2014 the Influencer-roll is no longer reserved for the well-known writers, bloggers and other celebrities according Alphenaar (“Waar moet je als marketeer in 2014 op letten, 2013”)
6 Table of contents 1. Introduction ... 8 1.1 Background information... 9 1.2 Thesis structure ... 11 1.2.1 Problem statement ... 11
1.2.2 Theoretical and social relevance ... 11
2. Theoretical framework ... 13
2.1 eWom adoption ... 14
2.2 Argument Strength ... 15
2.3 Source credibility ... 16
2.4 Confirmation of Prior Belief ... 16
2.5 Facebook and Twitter ... 17
2.6.1 Facebook and Twitter, the differences. ... 19
2.7 The (main)effect of Social Network Sites ... 21
3. Research design ... 23 3.1 Research method ... 24 3.2 Data collection ... 25 3.3 Plan of analysis ... 26 4 Results ... 28 4.1 Analyses ... 28 4.1.1 eWOM adoption ... 32
4.1.2 Argument strength, source credibility and confirmation with prior belief ... 33
5 Conclusions ... 37
5.1 Conclusion ... 37
5.2 Limitations and recommendations for further research ... 39
References ... 41
Appendix 1: Questionnaire ... 47
7
1.
8
1. Introduction
The Social Network Sites (SNSs) are rapidly changing. New SNSs are coming up and existing networks keep on developing to become the most used SNS. Before digging into the theoretical background, the facts and figures about Facebook and Twitter will be discussed. These facts give insight into the extent of the networks which will be discussed in this research. Insites Consulting (2011) investigated the social media usage around the world. The data considering Twitter and Facebook which is used in this chapter is derived from the Insites Consulting research.
Since 2010 is Facebook the most popular website worldwide (Lee and Cho 2011, Alexa 2013 ) and thus automatically also the most popular SNS. This is underlined by the fact that only 4% of the European people are not aware of Facebook. An awareness of 96% is high but it is the activeness of the SNS members that counts. When looking to the activeness, 62 % of the people that are aware of Facebook are also a member (while 7% was once, before 2011, a member).
Twitter is rapidly growing and has an awareness of 80%, this means that 20% of the European people are not aware of Twitter. In contrast with Facebook only 16% is a member of Twitter (and 5% was once, before 2011, a member), but the usage and members are still increasing.
Although in Europe Twitter is the second biggest and most known SNS after Facebook, 65% of the Facebook members do not have a Twitter account. While only 2% of the Twitter members do not have a Facebook account. Furthermore it is interesting to know that 19% of the European people is both member of Facebook and Twitter.
9
1.1 Background information
A network in which people are connected and engage in word of mouth is a SNS according Bristor (1990). Both Twitter and Facebook meet these criteria.
Twitter is a microblogging tool in which the user share their message up to 140 characters. With microblogging users are able to send short messages which can be distributed by instant messages, mobile phones, email or the Web (Java et al. 2007). Twitter users are able to subscribe to an account of interest in order to receive all the messages posted by that account. The Twitter users receive all the subscribed messages in their timeline (Twitter, “What is Twitter”, 2012). Twitter is used as a microblogging tool to talk about daily activities and to share information. Furthermore Twitter is also being used to seek information (Java, Song, Finin and Tseng, 2007; Church, 2008)
Facebook is an online SNS which started off as a geographically-bound community (Ellison, Steinfield and Lampe, 2007). Facebook is founded in 2004 and has reached over 901 million active users in 2012 (Facebook, “Key Facts – Statistics”, 2012). These users connect with friends and family over Facebook to share everything that matters to them. Furthermore Facebook is used to discover what is happing in the world. This sharing happens on the timeline of an user, which was introduced in September 2011. Within this timeline it is controllable who is able to see the posted content. Facebook is a free platform. By showing relevant ads, Facebook stays free for everyone and is able to keep on developing and growing.
The relevant ads on Facebook are based on demographic factors such as location, age, gender, education etc. Additionally companies are able to sponsor stories by intensifying the event that users have shared on their timeline (Facebook, “Advertising”, 2012). Promoted tweets is Twitters way of advertising. These promoted tweets can be placed at search results of users timelines. The relevance of the promoted tweet is based on location (Twitter, “Promoted Tweets”, 2012).
10 organization, brand, product or service (Harrison-Walker,2001; Arndt 1967). However, Braezeale (2009) states that nowadays the spokesperson does not have to be commercial, it can be an actual consumer. In this paper we assume that WOM is an informal conversation between consumers about a certain organization, brand product or service.
WOM and electronic Word of Mouth (eWOM) are similar in many ways. The use of internet makes WOM even more important (Bickart and Schindler, 2001; Lee and Youn, 2009). The world is at your fingertips and everyone is just a few ‘steps’ away according the six degrees of separation theory. The main difference between WOM and eWOM is that eWOM can occur from person to person without having a preceding relationship with one another (Lee and Youn, 2009). Moreover Lee and Youn (2009) state that eWOM can be anonymous which allows consumers to share their opinions in all honesty.
Another difference is eWOM could be spread much faster because of the internet and therefore more is powerful than WOM (Shu-Chuan and Yoojung, 2011). Because the resemblance between WOM and eWOM, this paper will mostly discuss eWOM because of the internet aspect. As the SNSs are being used to share everything that matters for the users, eWOM is very common across the SNSs. Zhang, Sobel and Chowdury (2009) mention in their article that 19% of the messages on microblogs, like Twitter, contain a mention of a brand. 20% of these messages mentioning a brand manifest an opinion of the brand. Remarkable is the fact that more than 50% were positive and 33% contained a critical message of the brand.
Bansal and Voyer (2000) investigated that when a sender is perceived as an expert, the receiver is likely to believe the incoming WOM information. Because this paper will investigate the reliability of eWOM at Twitter and Facebook, also the sender of the messages will be taken into consideration. Source expertise is one of the cognitive qualities described by Wathen and Burkell (2002) which influence the online credibility. Furthermore Cheung et al. (2008) mention source expertise as one of the determinants that influences the perceived eWOM credibility. Cheung et al. (2008) investigated the review credibility in on-line consumer discussion forums.
11
1.2 Thesis structure
Within this introduction the problem statement will be formulated and the explanation why the problem is scientifically and socially relevant. In the following chapter the different research questions will be clearly addressed in the theoretical framework. Furthermore chapter 2 shows graphically the relationships found in the literature. Chapter 3 will explain the choice of the research and the data collection method. The results will be discussed in chapter 4, together with the analyses and the discussion empirical date about the representativeness, reliability and validity.
1.2.1 Problem statement
Understanding how SNS users perceive the credibility of the posted messages is important for organizations which want to participate in the SNSs. Twitter and Facebook are two different SNSs and therefore a message with the same meaning could be received in in other way across the different platforms. For that reason different determinants like argument strength, source credibility and confirmation with prior belief will be taken into account in the different research questions. In order to investigate the influence of the most popular SNSs the following problem statement is composed:
Does perceived eWOM credibility and adoption differ on
Twitter and Facebook?
In order to answer that question it is important to know what eWOM adoption and credibility is. This information will be discussed in §2.1. The factors that play a role with perceived eWOM credibility is also well covered in §2.2 to §2.4. Furthermore in §2.5 the SNSs Twitter and Facebook will be discussed. Firstly chapter 2 starts off with the theoretical framework.
1.2.2 Theoretical and social relevance
12
2.
13
2. Theoretical framework
In order to create a good overview, the conceptual model will be presented first instead of the hypotheses. With the model in mind, paragraph 2.1 to 2.5 will elaborate the formulated hypotheses.
Figure 1 shows the research model which is based on the literature and the hypotheses of the following paragraphs.
Figure 1: research model
14
2.1 eWom adoption
As consumers can be influenced in several ways, the Dual-process theory considers two different types of influences affecting the credibility of online reviews. Normative factors versus informational factors. Informational influence is based to the extent in which sources influence consumers simply by providing information. Normative influences however is based on the behaviour through social pressure (Hoyer and MacInnes, 2007; Cheung 2009). These influences are important to take into account when investigating the credibility of eWOM across the different platforms.
The research of Cheung et al. (2008) investigated both the informational determinants as the normative determinants as an effect on the perceived eWOM review credibility. These normative determinants, recommendation consistency and recommendation rating, significantly influenced perceived eWOM credibility. However, this research goes further into the fact how receivers react on a single message shared on Twitter and Facebook. The receivers are not looking for a certain review but are confronted with a certain message about a brand. Therefore we only take the informational determinants into account which significantly influence the perceived eWOM credibility.
In case of adoption of WOM, it is investigated that the influence of WOM is stronger than other marketing actions (Buttle, 1998; Bickart and Schindler, 2001) especially according to the purchase decisions (Engle, Kegerreis and Blackwell 1969). Angelis et al. (2012) mention in their research that WOM influences nearly 70% of all purchase decisions.
In the research of Cheung (2009) “eWOM credibility is defines as the extent to which one perceives a recommendation/ review as believable, true, or factual”. In contrast with the Cheung research this research will also take the trusting beliefs about a person into account since the person behind the social network profile is also very important. When looking at online reviews people will only take online advice that they perceive as credible (Cheung, 2009). Therefore it is interesting to research what message will be received as credible. Thus, the following hypothesis is formulated:
15
2.2 Argument Strength
Argument Strength, coherent with the quality of the message, is an important driver of message acceptance (Coulter and Punj, 2004; Cheung et al., 2009). With valid arguments a message will be acknowledged as credible by the receiver. Because of the difference between Twitter and Facebook in the message length, it will be interesting to know if the same message will be received in the same way in the different SNSs. In other words, is the message strength for the same message different if there is a possibility to elaborate? As it is possible on Facebook to enlarge a message and explain more, this research will investigate if a short Twitter message on twitter is received in the same way as on Facebook. Cheung (2009) showed that argument strength has a positive effect on the perceived eWOM review creditability. Furthermore “argument quality is likely to moderate the persuasiveness of messages” mentions Kao (2012).
Hereby in figure 2 an example of a weak and a strong argument message according Kao (2012):
Strong Weak
Figure 2: examples of a strong and weak message according KAO (2012)
16
2.3 Source credibility
Credibility is the quality of being believable or trustworthy, also known as believability. Tseng and Fogg (1999) identified four types of credibility; presumed, reputed, surface and experienced. In this research we will not elaborate about the different types but speak about credibility in general. However for the insight the different types will be shortly discussed. It is mostly assumed that friends tell the truth, when friends are seen as credibly this is presumed credibility. Reputed credibility describes how much a perceiver believes someone because of the label of a third parties (for example a doctor or Nobel Prize winner). People prejudging its outward appearance alone is seen as surface credibility. Experience credibility is believability based on first-hand experience. These four types of credibility will also be found in the different SNSs. Chu and Kim (2011) found that the higher the level of trust within social network contacts, the higher the chance they will engage in opinion seeking and giving and even passing behaviour.
Like argument strength, source credibility is also proven by Cheung (2009) to have a positive effect on the perceived eWOM review credibility.
2.4 Confirmation of Prior Belief
17
2.5 Facebook and Twitter
Before introducing the believed differences, this research introduces both SNSs with the main variances. Since there is little existing scientific literature on Twitter and Facebook in general this paragraph makes use of the official annual reports.
Facebook announced in their annual rapport (Form 10-K) that their users, particularly younger users, are “aware of and actively engaging with other products and services similar to, or as a substitute for, Facebook”. They believe that ‘some of their users have reduced their engagement with Facebook in favor of increased engagement with other products and services such as Instagram’.
Since the use of Facebook is declining it is interesting to research what SNS could be best used for stimulation WOM and in which form. What kind of eWOM will be best adapted by the SNS users?
Facebook also announced in their annual rapport (Form 10-K) that they believe that recommendations of friends have a powerful influence on consumer interest and purchase decisions. With the help of Facebook they offer marketers “the ability to include social context with their marketing messages. Social context is information that highlights a friend’s connections with a particular brand of business”
According Twitter’s information (Twitter, 2013) about the audience there are 200 million active users. Those users send an average of 400 million Tweets every day. The Tweets vary “from your newest mobile app to a popular event you’re involved in to shopping for your products to news about your company.”
Twitter is used for multiple purposes like keeping in touch with acquaintances like friends and colleagues and gathering information for both business as personal interest. Furthermore to spread a message to a huge range of followers ( Zhao and Rosson 2009; Jhih-Syuan 2011). Kwak et al. (2010) took a closer look at retweets and discovered that the average range of a retweet is 1000 users whereby the number of followers does not matter. As soon as a message is retweeted, the message gets 4 ‘hops’ away from the original sender.
18 Twitter furthermore refers on their site to the study performed by the business named Compete (Audiences on Twitter, 2013), which is held in 2011 and elaborates on the facts why users follow brands on Twitter:
94% Discount and promos 88% Free stuff
87% Fun and entertainment 79% Updates on upcoming sales 79% Access to exclusive content
19
2.6.1 Facebook and Twitter, the differences.
The same qualitative good message on Twitter and Facebook will probably be seen differently since Facebook has more space to amplify the message. When a message is not explained further the reader might think that there is no further evidence and the argument strength will decrease. Shu-Chuan and Kamal (2008) remark that in greater message elaboration, the quality of the arguments has a greater impact compared with low message elaboration. Therefore, the perceived eWOM review credibility is stronger on Facebook than on Twitter when the possibility of elaboration on Facebook is used. However, when the same message (max 140 characters) is displayed on Twitter and Facebook it is believed that the credibility on Twitter is stronger because there is no elaboration expected. Thus, the following hypothesis for this research in formulated:
H2 The argument strength on Twitter has more positive influence on perceived eWOM credibility than argument strength on Facebook.
The messages on the Facebook timeline will be mostly assumed as true since Facebook is a network among friends, which is called the presumed credibility. At Twitter it is free to follow everyone, friends but also experts or other interesting people based on their biography on Twitter.
For that reason it is assumed that the credibility on Facebook is stronger. This is underlined by Brown and Reingen (1987) who found that weak ties have a crucial role in the flow of WOM across groups, while strong ties are significant at the micro level of referral behaviour. Consequently strong ties are perceived as more influential than weak ties in decision-making.
20 Gilly et al. (1998) mentioned that WOM between people with the same background is more efficient than when the background differs. Furthermore strong ties, like close friends on Facebook, have more influence than weak ties like organizations and acquaintances (Brown, 1987). Additionally, the connection on Facebook points out that there is a mutual trust because the friendship is accepted by both of the users. Whereby Twitter it is possible to follow a person without the need of acceptance from the other user or a follow-back, in this case it is called a weak tie.
The Facebook pages by brands or communities are not taken into account. These pages are open for everyone to like. In the research of De Vries et al. (2006) the effects of social media marketing on these brand pages is discussed. This research focuses on the eWOM of people within the network.
H3 Source credibility on Facebook has a stronger positive influence on perceived eWOM credibility than source credibility on Twitter.
The messages among different SNSs differs because of the difference of people that are in the SNSs. It is expected that people are more likely to believe information if it is conform the ideas of the reader. It is believed that at Twitter people choose to follow certain people because of their ideas, interesting messages and expertise, whereas Facebook is used as a network among friends, family and acquaintances. As Chu and Kim (2011) mention that people within the same SNS have the same socio-demographic characteristics like gender and age, but also beliefs and attitudes will be comparable. Therefore it will also be likely that people within a strong-tie network have the same opinions and beliefs. Facebook is for that reason the most common network to find messages that are equal to prior beliefs. On Twitter, with a wide range of people and opinions, it could be that the recommendation disconfirms the prior belief. In that case the recommendation will not be enhanced (Cheung et al. 2009).
21
2.7 The (main)effect of message credibility Social Network Sites
Google stated in their 10-K form (2010) they face competition from SNSs such as Facebook and Twitter. The report mentions that in some cases it is more likely that product referrals are more looked into at SNSs as Twitter and Facebook than search for the specific information through a search engine. This could be due the fact that these SNSs are seen as credible. It will be interesting to know if the credibility of the messages has something to do with that.
The follow concept of Twitter makes the (news)feed trustworthy since people tend to follow persons or brands they know personally or especially elected to follow. In this thesis we research the difference between the SNSs Facebook and Twitter. As Zhao and Rosson (2009) mention Twitter is more open and therefore easier to follow people with similar interest. There is no existing connection necessary to follow someone on Twitter. Although Twitter offers the possibility to lock an account, in practice only a few accounts are ‘locked’. On Facebook the network is more protected by mutual acceptance. Therefore it is more likely that the Facebook network consists of people they trust and therefore also their messages.
Within this research it is believed that eWOM on both Twitter and Facebook is perceived as credible. In order to research the relation of the general SNS message credibility on the perceived eWOM review credibility and to define the difference between Facebook an Twitter the following hypothesis is formulated:
H5: SNS message credibility on Facebook has a stronger positive influence on perceived eWOM credibility than SNS message credibility on Twitter.
Brown and Reingen (1987) mention that there are different kinds of ties, ranging from strong-ties like close friends to weak-strong-ties as acquaintances. Shu-Chuan and Yoojung (2011) found in their research that tie strength is positiviley related with eWOM. This is including opinion seeking which is interesting regarding the eWOM about the colour which is used in this research. The tie-strength is believed to have a significant contributory effect on the relationship and is therefore taken as the moderating effect. Therefore the following hypothesis if formulated with the moderating effect:
22
3.
23
3. Research design
In this chapter the methodology is discussed. Firstly, the conducted research will be described together with the chosen method. Secondly, the data selection will be discussed and at last the plan of analysis before the results will be shown.
The conceptual model will be tested with an product that comes with an experience. The chosen product is a smartphone and in particular the iPhone. When choosing a new phone there are nowadays a lot of choices. Most of the people make use of a smartphone but the brand and software differs. For that reason people may want to know more about the product before buying it. How one experience the product, how long the battery really lasts etc.. In contradiction with a low budget product which is bought monthly, for example razor blades which can be easily evaluated before purchasing, the need for eWOM will be higher with an experience product (Christiansen and Tax, 2000). Smartphones, and iPhones in particular, are seen as relatively expensive and therefore are a high perceived risk in a financial context, which contributes to the importance of WOM (Bansal and Voyer, 2000).
24
3.1 Research method
this research makes use of an online questionnaire in order to find out in which SNS, Twitter or Facebook, will positively tribute to a higher perceived eWOM credibility. The questions in this research are based on the research by Cheung et al. (2009) and are adapted to fit the conceptual model which is based on the literature research. Firstly the questions which are derived from previous research were translated into Dutch because the questionnaire will be distributed in The Netherlands. The questionnaire will be distributed online using the questionnaire software Limesurvey. Limesurvey offers the possibility to add different routes so respondents only need to see the questions that are relevant for them. Also the respondents who do not have a Twitter of Facebook account are filtered out immediately and will be thanked for their willingness to participate.
The first section of the questionnaire consist of general questions about Twitter and Facebook. The second section of the questionnaire consists of five-point Likert scale questions. These questions were derived from the research of Cheung et al. (2009). However in this research is chosen to use less items per construct to prevent fatigue among the respondents. This was a deliberate decision by which the measurability is taken into account. Therefore three items per construct is chosen.
The likert scale questions will be accompanied by a screenshot from the message they will have to judge. This is the same message for all the questions. There will only be a difference in the look-and-feel of the message so the respondent can relate to the network (Facebook or Twitter). As mentioned before the message will have the same length in this experiment in order to see how the message is evaluated on both SNSs.
Within this experimental design each respondent had an equal chance for facing the questions regarding the independents variables (argument strength, source credibility and conformation with prior belief) of Facebook and Twitter. With a random assignment it is ensured that the groups are as comparable as possible with respect to the dependent variable: perceived eWOM credibility (Cooper and Schindler 2006). Since the respondents need to empathize with the situation, the Facebook users see the message in a Facebook-surrounding while people with a Twitter account see the message in a Twitter-surrounding. If the respondents have a Facebook and a Twitter account they were asked to answer all the questions about Twitter and Facebook.
25 The total questionnaire is shown in appendix 1. In the following table (table 2) the different variables regarding the hypotheses are shown. The table gives an insight when in the following chapter the analytics and results will be described.
Construct Items Measurement
level
eWOM adoption
To what extent do you agree with the message? 5-point likert Information from this message contributed to my knowledge 5-point likert The message made it easier for me to make purchase decision. 5-point likert Perceived eWOM
credibility
I think the message is factual 5-point likert I think the message is accurate 5-point likert I think the message is credibile 5-point likert Argument
Strength
The message is convincing 5-point likert
The message is strong 5-point likert
The message is persuasive 5-point likert
Source credibility
The message sender is reputable 5-point likert The message sender is highly rated 5-point likert The message sender is trustworthy 5-point likert
Confirmation with prior belief
Information from message contradicted what I had known before reading it
5-point likert
The message supported my impression of discussed product 5-point likert The message reinforced information I had previously know
about the product.
5-point likert
Table 2: variables questionnaire (appendix 1)
3.2 Data collection
An online survey is conducted to gather as many respondents as possible who are using Twitter and Facebook. Before the questionnaire went online for everyone the questionnaire was tested by a minor group of test-respondents to check if there were some ambiguities, errors or other issues. After this test round some remarks were added to the questions so there could not be any misunderstanding. No major adjustments were made.
26 Secondly a respondents panel from thesistools.com / Van Rixtel is asked to fill in the questionnaire to complement the needed respondents in order to get a representative research. The target number of respondents was at least 100 respondents per SNS and a total of approximately 300 respondents. Respondents had the possibility to complete the questionnaire from 20 November 2013 to 04 December 2013. In these 14 days, 541 respondents started the questionnaire. Within these 541 replies there were some incomplete questionnaires and respondents who did not have a Facebook or Twitter account. This resulted into a total of 309 respondents who filled in the questionnaire completely and have a Facebook or Twitter account.
3.3 Plan of analysis
Statistical program SPSS 20 is used to test the hypotheses. Furthermore Microsoft Excel is used to create graphs and figures. The population for this research consists of the Dutch Twitter and Facebook-users. Within this population 309 respondents did cooperate in this research.
In particular the factor analysis and Cronbach’s Alpha will be executed on the various items. The factor analysis is a multivariate statistical technique which is used for a large number of variables to define the underlying structure. These underlying variables are the factors. The Factor analysis is used for data reduction.
If the correlations between the variables are small, it may not be appropriate to conduct a factor analysis. In order to check if the factor analysis is suitable the Bartlett’s test of sphericity can be used or the Kaiser-Meyer-Olkin (KMO).
Firstly the Principal Component Analysis is conducted and all the variables (table 2) are used. The KMO-score of the analysis is 0,831 which means that the factor analysis is appropriate. The Bartlett’s Test of Sphericity redulted in a Chi2 value of 2203,281 (p < 0,0000), which means that the factor analysis is appropriate.
27
4.
28
4 Results
In this chapter the data of the questionnaire will be analysed. Both the hypotheses as other interesting facts will be displayed in the analyses.
4.1 Analyses
To select the right target (Facebook and Twitter users) and gain insight in the user rate, the respondents needed to answer the question if they were familiar with Twitter and/or Facebook. The target group, Facebook and/or Twitter users, and therefore the number of respondents which are used in this research is 309.
In total 306 respondents were familiar with Facebook and 168 were familiar with Twitter. A total of 165 respondents indicated that they both know Facebook and Twitter. This also leads to the conclusion that only 3 of the 309 respondents are not familiar with Facebook. As expected Twitter is a bit less known, 141 respondents are not familiar with Twitter. The percentages of the respondents who are familiar with Facebook and/or Twitter are shown in figure 3.
Figure 4 shows that 99,3 of the respondents who are familiar with Facebook also have a Facebook account. While 72,6% of the respondents who are known with Twitter also have a Twitter account. This is graphically shown in figure 5. An explanation of such a high percentage for both Twitter and Facebook account holders can be due to the fact that respondents with a Facebook and/or Twitter account were invited the fill in the questionnaire. Therefore these figures cannot be used for brand awareness in total and are only used for insight in this research.
Figure 4: distribution respondents regarding Facebook account
Figure 3: Familiair with Twitter and/or Facebook Figure 5: distribution respondents regarding Twitter account 45,6% 1,0% 53,4% Facebook Twitter 99,3% 0,7% Facebook account No Facebook account
72,6%
27,4%
29 Furthermore it is interesting to see how the Facebook and Twitter networks are divided by people within the network. The respondents was asked to divide their network into friends, family, acquaintances, business, interesting people (that they don’t know) and other. The mean of these percentages are shown in figure 6. Where the Facebook network mostly contains out of friends, family and acquaintances, the Twitter network is used in another way and therefore the respondents mostly follow interesting people (that they do not know) or accounts that are interesting businesswise. This could also be really interesting for the relation of eWOM credibility and tie-strength.
Figure 6: the dividing of people within the SNSs Twitter and Facebook.
When looking at the gender of the respondents in figure 7, it is noticeable that the females have a majority with 68,90%. However the results of the independent t-tests shows there is no significant difference between male and female when looking to the perceived credibility on Twitter and Facebook. The ‘elderly’ from the age of 61 are less represented with 7,11% in this research. This could be due the fact that the questionnaire was only possible to fill in by respondents with a Facebook and Twitter account. Joining a SNS could be quite a step for people of that age. The total circle-diagram is shown in figure 8.
Figure 7: Male / Female respondents Figure 8: Respondents by age
41,0% 19,7% 25,7% 7,3% 3,8% 2,7% 24,5% 5,4% 11,5% 25,0% 28,7% 5,0%
Friends Family Acquaintances Business Interresting people (that you don't
know) Other Facebook Twitter 31,1% 68,9% Male Female 20,7% 29,1% 14,9% 16,5% 11,7% 6,8%
0,3% younger than 21 years old
30
When looking at the colour preferences of the respondents, the colour gold is the least popular as expected after the pre-test. The results of the choices are shown in figure 9. When looking at the figure it is clear that there is only a minor difference between the shifts after the displayed eWOM. With a correlation of ,960 and a minor differce of -,003 with the paired-samples t-test and a significance of ,764 we need to accept that there is no difference in the choices after the eWOM.
Figure 9: Colour choice before and after the eWOM
The following figure shows the differences between Facebook and Twitter, when looking at figure 10, it shows that 34,9% of the Facebook messages is seen as reliable. On Twitter this is 23,8%. When conducting a paired samples t-test it is shown that there is a significant difference between the reliability on Facebook and Twitter.
Figure 10: General reliability messages on Facebook / Twitter
53,4% 54,0%
37,2% 35,6%
9,4% 10,4%
Colour preference (before eWOM) Purchased coulor (after eWOM)
31
For more insight the respondents were asked if they ever see reviews within their network. Figure 11 shows that 47,7% of the Facebook users and 37,7% of the Twitter users do see a review on the SNS. On Facebook 145 respondents did ever see a review in their timeline. Out of those 145 people 21% found a review useful for a purchase decision. When asking the 46 Twitter users who ever did see a review on Twitter if the review helped with a purchase decision only 17,4% answered positive. The percentages of the reviews that helped with a purchase decision are shown in figure 12.
Figure 11: Do you ever see reviews on Facebook / Twitter
Figure 12: Did a review on Facebook / Twitter help with the purchase?
Factor analysis
In the research of Cheung et al. (2009) the variables Argument Strength, Source credibility, Confirmation with Prior Belief, Perceived eWOM credibility and eWOM adoption are conducted out of various questions. Within this research the questions are based on the research of Cheung et al. (2009) therefore the factor analysis will be used if the questions can be grouped into the same number of variables.
The factor analysis needs to be performed two times, once for the questions regarding Facebook and once regarding the questions about Twitter. For both tests the KMO and Bartlett’s Test of Sphericity shows (table 2) that de factor analysis is reliable and significant.
KMO Bartlett’s Test
Facebook 0,831 Sig. ,000
Twitter 0,844 Sig. ,000
Table 2: KMO and Bertlett’s test
When performing the factor analysis for the Twitter items the three factors as the independent variables in the research of Cheung et al. (2009) can be formed. The table with the results is shown in Appendix 2. The total variance for Twitter is 75% with 3 factors. In case of the factor
32
analysis for the Facebook items there are also three factors (with a total variance of 68%). However the item ‘Information from message contradicted what I had known before reading it’ which belongs in the Twitter factor analysis to ‘source credibility’ should belong to the factor ‘confirmation with prior belief’ according this analysis. Since this is not the case with Twitter and the research of Cheung (2009) the Conbach’s Alpha will be taken into consideration.
Looking at the Cronbach’s Alpha it is chosen to delete the “message is persuasive” and “the message reinforced the information I had previously known” for both the Twitter and Facebook factors. With deleting these items the Chronbach’s Alpha derived from poor/acceptable to good. The figures are shown in table 3.
Message is persuasive The message reinforced the information I had previously known
Facebook Twitter Facebook Twitter
Cronbach’s Alpha ,659 ,838 ,531 ,700
Cronbach’s Alpha if item deleted
,825 ,890 ,807 ,770
Table 3: Chronbach’s Alphas
This difference could be explained due to the reduced items per construct. After the test questionnaire it was a deliberate decision to reduce the items per construct to a maximum of three in order to prevent respondents from irritation and fatigue.
It is a nice given that this research could also work with the variables argument strength, source credibility and confirmation with prior belief. The factor analysis also seemed appropriate for the items belonging to the variables perceived eWOM credibility and eWOM adoption.
4.1.1 eWOM adoption
To test the relation between perceived eWOM credibility and the eWOM adoption, as formulated in hypothesis 1, a regression analysis is conducted.
33
Facebook Twitter
Unst. Beta Beta p-value Unst. Beta Beta p-value
Constant 1,265 1,000 -1,481 1,000 Perceived eWOM credibility ,554 ,554 ,000 ,612 ,612 ,000 R Square ,307 ,375 Anova Sig. ,000 ,000
Table 4: Regression analysis eWOM adoption (H1)
The R-square indicates that 38% of the dependent variable ‘eWOM adoption’ can be explained by the variable ‘perceived eWOM credibility’ on Twitter. Furthermore the ANOVA table shows that the model can statistically significantly predict the outcome variable (P = 0,000). Therefore we can state that perceived eWOM credibility has a positive effect on eWOM review adoption. This is both the case for Facebook and Twitter.
4.1.2. Argument strength, source credibility and confirmation with prior belief
When looking at the regression analysis for the relation between the determinants (argument strength, source credibility and confirmation with prior belief) on Twitter the variables explain 49% of the model. The model is significant, however when looking at the variables, the factor source credibility is not significant.
Facebook Twitter
Unst. Beta Beta p-value Unst. Beta Beta p-value
Constant 1,680 1,000 2,052 1,000 Argument strength ,388 ,388 ,000 ,531 ,531 ,000 Source credibility ,225 ,225 ,000 ,104 ,104 ,154 Confirmation with prior belief ,200 ,200 ,000 ,273 ,273 ,000 R Square ,365 ,491 Anova Sig. ,000 ,000
Table 5: Regression analysis perceived eWOM credibility (H2, H3, H4)
When this regression analysis is compared to the model of Facebook, it is clear that both models are significant. The R-square of the Facebook-model is 0,365. Furthermore are all the variables of Facebook significant. For both Twitter and Facebook it is the case that argument strength has the most effect on perceived eWOM credibility. On Facebook this is followed by source credibility while on Twitter the second highest effect is confirmation with prior belief.
34 The R-square indicates that 37% of the dependent variable ‘perceived eWOM credibility’ can be explained by the variables argument strength, source credibility an confirmation with prior belief on Facebook. Furthermore the ANOVA table shows that the model can statistically significantly predict the outcome variable (P = 0,000). Furthermore table 5 shows the regression analysis for Twitter on eWOM credibility. When looking at the R-square it indicates that 49% of the perceived eWOM credibility can be explained by the variables argument strength, source credibility an confirmation with prior belief on Facebook. However source credibility is not significant. As such hypothesis 3 for Twitter is rejected.
4.1.3 SNS message credibility and tie strength
Now the determinants are discussed, the direct relation between the credibility of messages on Twitter and Facebook and the perceived eWOM credibility will be taken into account. This relation is tested in a new regression model. Within this model the main effect of SNS message credibility and tie strength will be taken into account. Furthermore the moderating effect of the tie strength which is shown in hypothesis 6. This is shown in table 6.
The reason for two separate regression models on perceived eWOM credibility is due to the fact that when the model is build with all the variables almost no interesting relations could be derived (because of insignificancy). Besides the R-square of this separate models indicates that as much as 90% of the perceived eWOM credibility can be explained by the variables credibility, tie-strength and the moderating effect in the case of Facebook. All the Facebook variables have a positive and significant effect on the perceived eWOM credibility.
Facebook Twitter
Unst. Beta
Beta p-value Unst. Beta
Beta p-value
Constant -,576 ,000 -,773 ,000
SNS message credibility ,060 ,044 ,019 ,186 ,139 ,000
Tie-strength ,096 ,099 ,000 ,059 ,063 ,077
Moderator Tie strength* SNS message credibility
,261 ,918 ,000 ,312 ,870 ,000
R Square ,904 ,857
Anova Sig. ,000 ,000
35 When looking at the Twitter-model the variables SNS message credibility and the interaction (tie-strength*SNS message credibility) have a positive effect on the perceived eWOM credibility. Tie-strength has a p-value of ,077 and therefore is not significant. With the interaction the relation between SNS message credibility and perceived eWOM credibility is stronger for strong ties than for weak ties.
36
5.
37
5 Conclusions
Firstly the hypothesis will be discussed shortly. After that the main question will be answered in the conclusion. In paragraph 5.2 the limitations of this research and recommendations for further research will be discussed.
5.1 Conclusion
Within this research the confirmation is found that perceived eWOM credibility has a positive effect on eWOM review adoption. This is the case for both the messages on Twitter as on Facebook.
This research concluded that argument strength is the most important factor for perceived eWOM credibility on both Twitter and Facebook. When looking at the positive influence it can be concluded that the argument strength on Twitter is stronger than on Facebook. Herewith we can accept hypothesis 2.
On Facebook the source credibility was the second most important factor for perceived eWOM credibility. With advertising on Facebook and using the network of the target (to gain more ‘likes’) it is important to check if the source is credible on that area (select on study, work or interest) this will lead to higher perceived eWOM credibility. Within Facebook-advertising it is possible to select on these different items. At last the confirmation with prior belief showed a positive effect on the perceived eWOM credibility.
On Twitter the confirmation with prior belief is the second most important factor for perceived eWOM credibility. This could be used in advertising on Twitter by checking the sentiment of tweets about the specific product or service. This way the advertisement could be used with twitter users with the same belief. The relation between source credibility and perceived eWOM credibility on Twitter is not proved.
Hypothesis 3 cannot be compared because the source credibility on Twitter is not significant. Therefore we cannot conclude if the positive influence of source credibility on perceived eWOM review credibility is stronger on Facebook than on Twitter.
38 The statistics shows that in general the messages on Twitter are more credible than the messages on Facebook. Therefore we cannot accept hypothesis 5.
At last the moderating effect of the tie strength has a stronger impact on Facebook than on Twitter, therefore we can accept hypothesis 6. Additionally we can state that the positive moderating effect of tie strength is stronger on Facebook than on Twitter in relation to the perceived eWOM review credibility. How stronger the tie, the stronger the positive effect of SNS message credibility on perceived eWOM credibility.
With accepting or rejecting the hypotheses the main question of this research is answered. Yes, there are some differences in the eWOM credibility and eWOM adoption on Twitter and Facebook. Although there may some differences in the relative importance of the variables, this research showed that perceived eWOM credibility on both Twitter and Facebook have a positive effect on eWOM review adoption. Furthermore the determinants (only source credibility was not significant) have shown to have a positive relation on both Twitter and Facebook regarding the eWOM credibility.
Using this knowledge businesses could anticipate better on the perceived eWOM by giving more attention to the advertisements. When advertising the business has the possibility to show the advertisement at the target together with people within the targets network who likes the page.
Figure 13 shows an example of a Facebook advertisement that is shown on a page with the likes of people, within the target his network, who likes the page. For extra effect the business should select the right targets and make the argument in the message as strong as possible. Furthermore the source credibility and confirmation with prior belief have to be taken into account. Figure 13: Example Facebook advertisement
39 Argument strength and confirmation with prior belief also needs to be taken into account when creating an ad or message on Twitter. However, the space in Twitter messages is limited thus it is recommended to focus on argument strength. Since argument strength has shown to have the strongest relation to perceived eWOM adoption. Unfortunately the advertisement options on Twitter are not as comprehensive as the possibilities on Facebook. Although the eWOM on Twitter is perceived as more credible than the messages on Facebook. With creating a buzz the businesses should take into account that confirmation with prior belief and the SNS message credibility with the interaction effect of tie-strength do have a positive effect on the perceived eWOM credibility.
5.2 Limitations and recommendations for further research
This research has a few limitations. Firstly some items were removed to prevent fatigue among the respondents who need to fill in the questionnaire. In further research it is recommended to use all the items as suggested in the research of Cheung et al. (2009).
As research has indicated in the past WOM has shown to be more effective than other traditional marketing tools (such as advertising) (Buttle, 1998; Bickart and Schindler, 2001). Therefore it is unexpected that this research showed that there is no significant change before and after the eWOM regarding the colour of preference. Especially because the determinants have shown to have a positive effect on the perceived eWOM credibility. This could be due to the fact that the review was especially about the colour of the product. For further research another kind of review could be taken into consideration. Instead of a colour it could be more technical. It would be nice to see if the persuasion of respondents differ between an opinion and a more ‘technical’ fact.
40
Conclusions
41
References
Articles
Angelis, M., Bonezzi, A., Peluso, A., Rucker, D., & Costabile, M. (2012). On Braggarts and Gossips: A Self-Enhancement Account of Word-of-Mouth Generation and Transmission. Journal Of Marketing Research (JMR), 49(4), 551-563. doi:10.1509/jmr.11.0136
Arndt, J. “Role of Product-Related Conversations in the Diffusion of a New Product.” Journal of Marketing Research 1, 4 (1967): 291–295.
Atkin, C. and Block, M. (1983), “Effectiveness of Celebrity Endorsers,”Journal of Advertising Research, 23, March, pp. 57-61
Bansal, H.S. & Voyer, P.A. (2000), Word-of-mouth processes within a services purchase decision context, Journal of Service Research, Vol. 3 No. 2,166-77.
Bickart, B., and R.M. Schindler (2001), Internet Forums as Influential Sources of Consumer Information, Journal of Interactive Marketing 15, 3: 31-40
Breazeale, Michael (2009), Word of Mouse, International Journal of Market Research, Vol. 51 Issue 3, p297-318
Buttle, F. (1998), Word of mouth: understanding and managing referral marketing. Journal of Strategic Marketing, 6, 3, pp. 241-254
Brown, J., & Reingen, P. H. (1987). Social Ties and Word-of-Mouth Referral Behavior*. Journal Of Consumer Research, 14(3), 350-362.
Cheung, Christy M.K.; Lee, Matthew K.O. & Rabjohn, Neil (2008), the impact of electronic word-of-mouth; the adoption of online opinions in online customer communities. Internet Research, Vol. 18, Issue 3, 229-247
Cheung, M.Y.; Luo, C.; Sia, C.L.; and Chen, H. (2009) Credibility of electronic word-of-mouth: Informational and normative determinants of on-line consumer recommendations. International Journal of Electronic Commerce, 13, 4 (summer 2009), 9-38
42
online]. May 2012;26(2):83-91. Available from: Business Source Premier, Ipswich, MA. Accessed December 14, 2013.
Shu-Chuan, C., & Yoojung, K. (2011). Determinants of consumer engagement in electronic word-of-mouth (eWOM) in social networking sites. International Journal Of Advertising, 30(1), 47-75. Christiansen, T., & Tax, S. S. (2000). Measuring word of mouth: the questions of who and when?.
Journal Of Marketing Communications, 6(3), 185-199. doi:10.1080/13527260050118676 Coulter, K. S., & Punj, G. N. (2004). The effects of cognitive resource requirements, availability, and
argument quality on brand attitudes. Journal Of Advertising, 33(4), 53-64
Eun Sook, K., & Yongjun, S. (2011). FOLLOW ME! GLOBAL MARKETERS' TWITTER USE. Journal Of Interactive Advertising, 12(1), 4-16
Shu-Chuan, C., & Yoojung, K. (2011). Determinants of consumer engagement in electronic word-of-mouth (eWOM) in social networking sites. International Journal Of Advertising, 30(1), 47-75. Church, Geno. To tweet or not to tweet: How and when to use Twitter in PR efforts. Public Relations
Tactics, Jul2008, Vol. 15 Issue 7, p10-10, 1p
Deepa S., D., Chopade, N., & Ranjith, P. V. (2012). Social Networking Sites — A New Era of 21st Century. SIES Journal Of Management, 8(1), 66-73
Ellison, N.B., Steinfeld, C., & Lampe, C. (2007). The benefits of Facebook “Friends:” Social capital and college students’use of online social network sites. Journal of Computer-Mediated Communication, 12(4), article 1. Retrieved September p, 2007.
Engle, J. F., Kegerreis, R. J., & Blackwell, R. D. (1969). Word-of-mouth Communication by the Innovator. Journal Of Marketing, 33(3), 15-19.
Erdogan, B. Zafer (1999): Celebrity Endorsement: A Literature Review, Journal of Marketing Management, 15:4, 291-314
Gilly MC, Graham JL, Wolfinbarger MF, Yale LJ. 1998. A Dyadic Study of Interpersonal Information Search. Journal of the Academy of Marketing Science 26(2): 83.
43
Hennig-Thurau, FT, 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 mierneti Journal of Interactive Marketing, 18(1), pp. 38-52.
Huberman, B.A., Romero, D.M., & Wu,F. (2009), Social networks that matter: Twitter under the microscope, First Monday 14, (1-5)
Hoch, S.J. and Deighton, J. (1989), “Managing what consumers learn from experience”, Journal of Marketing. 53 (April), 1-20.
Jansen, B.j.; Zhang, M.; Sobel, K.; and Chowdury, A. (2009), Twitter power: Tweets as electronic word ofmouth. Journal of the American Society for Information Science and Technology, 60:1 20 Java, X. Song, T. Finin, and B. Tseng. Why we twitter: understanding microblogging usage and
communities. In Proc. of the 9th WebKDD and 1st SNA-KDD 2007 workshop on Web mining and social network analysis. ACM, 2007.
Jhih-Syuan L, Peña J. ARE YOU FOLLOWING ME? A CONTENT ANALYSIS OF TV NETWORKS' BRAND COMMUNICATION ON TWITTER. Journal Of Interactive Advertising [serial online]. Fall2011 2011;12(1):17-29.
Kao, D. (2012). Exploring the effect of regulatory focus on ad attitudes: The moderating roles of message sidedness and argument quality. International Journal Of Psychology, 47(2), 142-153. doi:10.1080/00207594.2011.628672
H. Kwak, C. Lee, H. Park, and S. Moon. What is twitter, a social network or a news media? Pages 591– 600. ACM, 2010.
Lee, Mira; Youn, Seounmi. Electronic word of mouth (eWOM) – how eWOM platforms influence consumer product judgement. International Journal of Advertising, 2009, Vol. 28 Issue 3, p473-499, 27p
Lee, S., & Cho, M. (2011). Social media use in a mobile broadband environment: examination of determinants of Twitter and Facebook use. International Journal Of Mobile Marketing, 6(2), 71-87.
44
Maheswaran, D., and J. Meyers-Levy, “The influence of message framing and issue involvement,” Journal of Marketing Research, (27:3), 1990, 361-367
McKnight, D.H. & Kacmar, C. (2006) Factors of information credibility for an Internet advice site. Proceedings of the 39th Hawaii International Conference on System Sciences, Vol. 6, 113. Murray, K. B. (1991). A test of services marketing theory: Consumer information acquisition activities.
Journal Of Marketing, 55(1), 10
Petty, Richard E., Cacioppo, John T., and Schuman, David (1983), “Central and Peripheral Routes to Advertising Effectiveness: The Moderating Role of Involvement,”Journal of Consumer Research, 10, December, pp 135-146
Shu-Chuan, C., & Kamal, S. (2008). The effect of perceived blogger credibility and argument quality on message elaboration and brand attitudes: and exploratory study. Journal Of Interactive Advertising, 8(2), 1-31.
Trusov, Michael, Randolph E. Bucklin, and Koen Pauwels, (2009), "Effects of Word-of-Mouth versus Traditional Marketing: Findings from an Internet Social Networking Site", Journal of Marketing, 73 (5), 90-102.
Tseng, S. & Fogg, B.J. (1999). Credibility and computing technology. Communications of the ACM, 42(5), 39–44.
Vázquez-Casielles, R., Suárez-Álvarez, L., & del Río-Lanza, A. (2013). The Word of Mouth Dynamic: How Positive (and Negative) WOM Drives Purchase Probability: An Analysis of Interpersonal and Non-Interpersonal Factors. Journal Of Advertising Research, 53(1), 43-60. doi:10.2501/JAR-53-1-043-060
Wang A. The Effects of Expert and Consumer Endorsements on Audience Response. Journal Of Advertising Research [serial online]. December 2005;45(4):402-412. Available from: Business Source Premier, Ipswich, MA. Accessed June 19, 2013.
Zhang, W., & Watts, S. A. (2008). Capitalizing on Content: Information Adoption in Two Online communities. Journal Of The Association For Information Systems, 9(2), 72-93.
45
Books
Cooper, D.R., Schindler P.S. (2006). Business Research Methods (9th ed.) McGraw-Hill international edition.
Hair, J. F. Jr., Black, W. C., Babib, B. J., Anderson, & R. E., Tatham, R. L. (2006). Multivariate data analysis (6th ed.). Upper Saddle River, NJ: Pearson.
Internet
Alphenaar, J.W., Retrieved October 3, 2013, from
http://www.intermediair.nl/vakgebieden/marketing/waar-moet-je-als-marketeer-2014-op-letten
Facebook, Advertising, Retrieved May 25, 2012, from
http://newsroom.fb.com/content/default.aspx?NewsAreaId=139
Facebook, Form 10-K annual rapport, December 31 2012, Retrieved June 16, 2013, from
http://investor.fb.com/secfiling.cfm?filingID=1326801-13-3
Facebook, Key Facts – Statistics, Retrieved May 25, 2012, from
http://newsroom.fb.com/content/default.aspx?NewsAreaId=22
Google Inc, Form 10-K annual rapport, December 31 2010, Retrieved Jule 1, 2013, from
http://investor.google.com/documents/20101231_google_10K.html
Insites Consulting, “Social media around the world 2011”, Steven van Belleghem, September 14, 2011, Retrieved May 8, 2012 from http://www.slideshare.net/stevenvanbelleghem/social-media-around-the-world-2011
Twitter, Promoted tweets (2012), Retrieved May 25, 2012, from
https://business.twitter.com/nl/advertise/promoted-tweets/
Twitter, What is Twitter (2012), Retrieved May 25, 2012 from
Twitter, Audiences on Twitter (2013), Retrieved July 6, 2013 from
46
47
Appendix 1: Questionnaire
The questionnaire is in Dutch since the population of this research is the population in The
Netherlands who uses Twitter or Facebook. Within the questionnaire were some routings, but for the insight all the questions are visible in this appendix.
Met welke van de volgende social media bent u bekend? Twitter
Twitter & Facebook Geen van beiden
Heeft u een Facebook account? Ja
Nee
Heeft u een Twitter account? Ja
Nee
Hoe omschrijft u de mensen binnen uw Facebook netwerk?
Verdeel 100% over de categorieën (Niets in de categorie? Vul dan 0 in). Vrienden
Familie Kennissen Zakelijk
Interessante mensen (die u niet direct kent) Anders
Hoe omschrijft u de mensen binnen uw Twitter netwerk? (De mensen die u volgt). Verdeel 100% over de categorieën (Niets in de categorie? Vul dan 0 in).
Vrienden Familie Kennissen Zakelijk
Interessante mensen (die u niet direct kent) Anders
Algemene vragen
Bij de volgende vragen uzelf graag inleven in de situatie.
Indien u een kleur moet kiezen voor de iPhone 5s, welke kleur heeft dan uw voorkeur?
48
Ga ervan uit dat u een iPhone wilt kopen, maar u bent er nog niet helemaal uit wat betreft de kleur.
Neem vervolgens iemand binnen uw Facebook netwerk in gedachten waarvan u de berichten regelmatig liked (leuk vindt). Vervolgens ziet u diegene het volgende oprechte bericht plaatsen:
Hoe staat de persoon die u in gedachten heeft tegenover de iPhone? Zeer negatief
Negatief Neutraal Positief Zeer positief
In welke categorie valt de persoon die u in gedachten heeft?
Vrienden Familie Kennissen Zakelijk
Interessante mensen (die u niet direct kent) Anders
Welke binding heeft u met de persoon in gedachten? Zeer sterk
