The influence of Facebook and Twitter on the perception of social presence, e-trust and purchase intention in e-commerce

(1)

Master Thesis

The influence of Facebook and Twitter on the perception of

social presence, trust and purchase intention in

e-commerce.

29 januari 2015 University of Amsterdam Nienke Bootsma - 10982590 Supervised by dhr. drs. F. Slisser

(2)

(3)

3

Statement of Originality

This document is written by Nienke Bootsma, who declared to take full responsibility for the contents of this document. I declare that the text and the work presented in this document is

original and that no sources other than those mentioned in the text and its references have been used in creating it. The Faculty of Economics and Business is responsible solely for the

(4)

Acknowledgements

This Master thesis is the final assignment for my Master Business Administration at the University of Amsterdam. It describes the influence of social media platforms, Facebook and Twitter, on social presence, e-trust and purchase intention in an online shopping environment. Completing this master thesis would not have been possible without the help of a number of people. Therefore, I would like to express my feelings of gratitude for their help. First of all, I would like to thank my thesis supervisor, dhr. drs. F. Slisser. His guidance, enthusiasm, critical view and feedback gave me additional insights and confidence to write this thesis. Furthermore, I would like to thank Bart van Endhoven, who supported me with the data analysis and structuring my thesis. I would also like to thank my boyfriend, Rogier van der Vat, for his unconditional support during this exciting, but stressful period of writing my thesis, and my lovely parents and sisters, for always being there for me. And thanks to my friends for the support and spreading the survey around. Last but not least, I would like to thank all my respondents for filling in the survey.

Nienke Bootsma

(5)

5

Abstract

Despite the proven lack of presence of human elements, online shopping is becoming more and more popular. Several new technologies increase the degree of social presence - the salience of the others in a mediated communication – on e-commerce platforms. This study focused on the influence of social media platforms, Facebook and Twitter, on social presence, building forward on the theory described by Lu & Fan (2014), who presented a model where social presence had a positive influence on e-trust and e-trust had a positive influence on purchase intention in e-commerce. These results are reproduced in this study and the influence of Facebook and Twitter is measured.

The research question studied in this research is ‘To what extent have online sellers’ webcare interaction on Facebook and Twitter an influence on social presence, e-trust and purchase intention and does it differ between Facebook and Twitter?’. In order to test this research question a quantitative survey was filled in by 182 respondents. For the survey the fictional webshop ‘E-shop.com’ was introduced and the respondents were randomly assigned to one of the four different situations. In the first situation, E-shop.com didn’t have Facebook or Twitter. Second, E-shop.com only had Facebook. Third, E-shop.com only had Twitter. And in the latter situation, E-shop.com had as well Facebook as Twitter.

Social presence consists of 3 dimensions, which are social presence of the web, perception of others and social presence of interaction with the seller (Lu & Fan, 2014). For the first dimension the influence of social media buttons on the contact page of E-shop.com was measured. For the second and third dimension of social presence, the influence of proactive and reactive webcare was measured.

The results show that social presence of the web can be increased by integrating a Facebook button, a Twitter button or both of these social media buttons on the contact page of a webshop. Perception of others can be increased by using reactive webcare on Facebook or Twitter or proactive webcare on Facebook. Social presence of interaction with the seller can be increased by using reactive webcare on Facebook, Twitter or both of these platforms. When using proactive webcare, social presence of interaction with the seller will only increase when it is used on both Facebook and Twitter.

(6)

1. Introduction

In the past years, online shopping (OS) showed a tremendous growth and it is estimated this grow will continue in the upcoming years (statista.com). It might be due to its efficiency, that OS is increasing popularity. In comparison with physical shopping (PS), OS has some

advantages, namely: it makes it easier to compare several product offers, it is time saving and the bought products will be delivered at home. However, shopping always has been a social activity, whereby social interactions with others influence consumers purchase decisions (Godes et al., 2005). Present days, these face-to-face social interactions are absent on the OS platforms, so called e-commerce platforms.

Lu & Fan (2014) showed that the major weakness of OS is indeed the lacking presence of human elements and might even hinder the growth of e-commerce platforms. When approaching a new experience people try to apply the rules that have governed previous similar experiences. For e-consumers, which means consumers that shop online, this will be the experience of PS. However, when shopping online they will miss the geographical, temporal, tactile and social experiences of PS (Rutter, 2001).

To understand the difference in presence of human elements between PS and OS, the concept of social presence can be used. With the social presence theory, Short et al. (1976) showed that a medium’s social effects are principally caused by the degree of social presence which it affords to its users. In this theory social presence is defined as the salience of the others in a mediated communication (Short et al., 1976).

Apparently a social component is missing on e-commerce sites. Social commerce is a new evolution in e-commerce. Social commerce increases the degree of social presence in the online environment and hereby it reintroduces the social aspect of shopping to e-commerce. Social commerce stands for new design features built upon social media and Web 2.0

technologies. These new design features help to enhance consumer participation and allow to collect socially rich information, resulting in a more trustworthy and sociable online

transaction environment. Social commerce combines commercial and social activities by deploying social technologies into e-commerce platforms (Lu & Fan, 2014).

Although OS lacks presence of human elements, it’s becoming more and more popular. New technologies, like social media, make it easier for e-commerce platforms to

(9)

9

increase the degree of social presence. However, surprisingly the role of social media platforms like Facebook and Twitter is not clearly defined yet. This study is conducted in order to fill this gap in the literature. It would be interesting to know if these social media platforms contribute to the degree of social presence, trust and purchase intention of an e-commerce platform and if there is any difference between these platforms. Apart from academic relevance, it also helps marketing managers to create a social media strategy that increases social presence, e-trust and purchase intention.

2. Literature review

2.1. Social presence on e-commerce platforms

2.1.1. Defining social presence

In the introduction the main difference in social presence between PS and OS is described shortly. Also a short definition of social presence – the salience of the other in a mediated communication – was provided. This paragraph will now elaborate a more extensive description of social presence and why it’s important on e-commerce platforms.

First of all, it is important to know that presence consists of two interrelated

phenomena, which are physical presence and social presence. The first, physical presence, is about the sense of ‘being there’. The second, social presence is about ‘being together with another’ (Biocca et al., 2003). While physical presence consists of automatic responses to physical cues and models that create the illusion of place and time, social presence consists of primitive responses to social cues, the simulation of ‘other minds’ and the feeling of others. In this research the focus will be on social presence. Biocca & Harms (2002) gave the following elaborative definition of social presence:

Mediated social presence is property of people, not of technologies, but it is a moment-to-moment phenomenal state facilated by a technological

(10)

course of a mediated interaction from a low level awareness that another being is co-present to more intense sense of the accessibility of psychological

modeling of the other’s intentional states (i.e., the attributional modeling of the other mediated mind).

At the lowest levels social presence is characterized by perceptual awareness, a peripheral sense of spatial co-presence of the other’s mediated body, and minimal, automatic attributions about the internal states of the other such as basic categorization of the other’s identity, sentience, and attention. Over the course of mediated interaction, social presence may include increasing sense of the accessibility of the other, perceived as increasing psychological and

behavioral engagement. Psychologically the user may have a greater sense of access to intentional, cognitive, or affective states of the other.

Finally, consider, as the level of social interaction, social presence can be characterized as subjective or intersubjective judgment of mutual accessibility of the other, such as mutual attention, mutual comprehension, shared emotional states, and interdependent behavior. (Biocca & Harms, 2002, p. 11)

Biocca & Harms (2002)’ definition state that social presence can have different levels in people’s perception. Depending on the nature of the mediated communication.

The fact that there is no physical presence and interaction with OS is the main difference between PS and OS. All interactions that take place with OS are mediated. This means people do not sense and interact with persons immediately, but they experience the other as if they were co-present and socially engaged. Depending on the characteristics of the person, the nature of the interaction and the properties of the medium, people will experience a certain level of social presence in these mediated interactions (Biocca & Harms, 2002).

A concept closely related to the concept of social presence is media richness. This concept explains ‘the properties of the medium’ mentioned above and explains the difference between direct and mediated communication. The media richness theory states that media differ in the degree of richness they possess. Richness means the amount of information that

(11)

11

can be transmitted in a given time. The amount of information that can be transmitted differs per medium (Kaplan & Heanlein, 2010). Face-to-face communication for example can be seen as very rich, while e-mail conversations are far less rich, due to the fact that with this medium the persons are not visible and the answers are delayed. Overall, a higher perception of social presence is expected for rich media compared to less rich media.

Lu & Fan (2014) propose a three- dimensional conceptualization of social presence in e-commerce. Their conceptualization consists of social presence of the web, perception of others and social presence of interaction with sellers. These three dimensions of social presence clarify that online consumers need to browse information by interacting with the web, make sense of other consumers and communicate with the sellers.

Social presence of the web

The first dimension, social presence of the web, refers to the capability of a website to convey a sense of human warmth and sociability (Lu & Fan, 2004). This dimension concerns

functionalities and content of an e-commerce platform that gives the visitor of this platform a feeling social presence of the seller. Basically all personal and social functionalities or content of the e-commerce platform count for this dimension. These content and functions can for example be an avatar that welcomes the customer on the website, a socially rich text or a picture of the employees. This dimension is not about interaction between consumers and sellers, but just about the personal and social functionalities the website offers.

Perception of others

The second dimension, perception of others, refers to the extent to which other social factors appear to exist and react to the users in online communities (Lu & Fan, 2004). This dimension is about the ‘hints’ consumers get on the existence of other consumers on the same

e-commerce platform. This could be a review of a certain product by other e-consumers or a functionality on the e-commerce platform that shows who already bought a product. But it could also be a discussion between consumers on social media. This dimension is about the possibilities an e-commerce platform offers it’s consumers to status updating,

(12)

Social presence of interaction with sellers

The third dimension, social presence of interaction with sellers, refers to chat tools that make online interactions between consumers and sellers possible (Lu & Fan, 2014). This dimension adds to the conceptualization of social presence that it’s not only about imaginary

interactions, but also about real interactions between consumers and sellers. The perception of social presence of interaction with sellers on e-commerce platforms could for example be increased by a chat function for sales and customer support. In order to sustain a good relationship with sellers, a few methods are often employed by sellers, such as, choosing special words and sending the emoticons like smile. Lu & Fan (2014) conclude that the computer meditated communication chat tools convey a sense of social presence. Which means that consumers can make sense of the attitudes, benevolence, and integrity of sellers via these computer mediated interactions, thereby forming beliefs about the ecommerce website, such as if they can trust the seller.

2.1.2. The relevance of social presence in e-commerce

Social presence is important to consumers when they come to think about the characteristics of the online seller and the quality of the products (Short et al., 1976). Biocca et al. (2003) showed that consumers are motivated to use several media to get the feeling of social

presence in a wide range of activities, like getting to know someone, exchanging information and opinions and making decisions. Increased social presence will lead to a better perception of the seller (Short et al., 1976). Hereby it will increase e-trust and consequently purchase intention (Lu & Fan, 2014).

2.2. E-trust

2.2.1. Defining e-trust

Trust can be described as the relationship between an agent and another agent or object, that is based on the agent’s beliefs about the capabilities of the other agent or object (Taddeo, 2009). Another commonly used definition of trust is ‘the willingness to rely on an exchange

(13)

13

partner in whom one has confidence’ (Moorman et al., 1992). Mayer et al. (1995) described trust as the willingness to be vulnerable to the actions of another person of people. All these definitions of trust have in common that a person trusts another person or organization when he or she expects that the commitments undertaken by the other person or organization will be fulfilled. Knowing this, we can separate cognitive and affective trust. Cognitive trust is based on the consumers’ judgment of the reliability and capabilities of the seller and affective trust is based on a bond among individuals (Ribbink et al.,2004). Trust in an online medium can be called e-trust.

Mayer et al. (1995) showed that trust could be seen as the product of a set of three trustworthiness beliefs. The first, integrity, is about the belief that the trusted party attaches to rules, for example keeping promises. The second, ability, means the beliefs about the skills and competence of the trusted party, for example delivery in time. The third, benevolence is the belief that the trusted party wants to do good to the consumer, asides from wanting to make profit.

2.2.2. The relevance of trust in e-commerce

Ribbink et al. (2004) showed that, since online transactions involve many uncertainties for the customer, trust is a condition for exchange. The uncertainties that online consumers face while involving in online transactions are for example the lack of direct contact with the seller and they have to hand over sensitive personal information, like credit card numbers, in order to complete a transaction. Even when online consumers only enter an e-commerce platform, without purchasing anything, personal information may be automatically collected (Gefen, 2002). Also Gefen & Straub (2004) showed the importance of trust to overcome uncertainties in the online environment. According to Gefen & Straub (2004), people always try to reduce their social uncertainty, which means they seek ways to understand and predict the behavior of other human beings. Social uncertainty can be reduced through rules, customs and trust. And as in online environments the rules and customs are often not defined, trust is an important condition for exchange.

Gefen (2002) showed two outcomes of consumer trust in e-commerce. First of all, consumer trust increases the willingness to buy something from the online seller, i.e. purchase

(14)

intention. The second outcome of consumer trust in e-commerce is the willingness to

window-shop at the e-commerce platform, in other words, visiting the website. So, e-trust is believed to increase online customer purchase intention.

According to Riegelsberger et al. (2003), online consumers process several trust cues when assessing an online sellers’ trustworthiness prior to a transaction. These trust cues can be cues based on cognitive trust and affective trust. Cognitive trust cues are based on cues that show that the seller has a rational interest to act as promised; this could for example be third party certifications and links to well-established brands. Moreover, several studies have shown that affective trust cues, such as interaction with the seller, also have a strong impact on consumer decision-making (Riegelsberger et al., 2013).

Gefen & Straub (2004) showed that trust relates to other people and is nourished through interactions with them. Therefore they describe social presence as an important factor for e-trust. Social presence influences trust, as trust is built through constructive interactions with other people. This means that the perception of a high degree of social presence is likely to contribute to the building of trust (Gefen & Straub, 2004). In this logic, the perception that there is personal contact with the seller via computer-mediated communication, which means high social presence, should increase e-trust.

Lu & Fan (2014) also described the positive effect of social presence on e-trust. Thereby the authors described a significant positive impact of trusting beliefs on purchase intention in the context of online social commerce. The authors showed that the three dimensions of social presence, social presence of web, perception of others and social presence of interaction with the sellers, have a positive influence on e-trust. The more consumers trust the online seller, the higher the purchase intention. So, e-trust can be seen a mediator in the relation between social presence and purchase intention.

2.3 Purchase intention

2.3.1. Defining purchase intention

According to Fishbein & Ajzen (1975) intentions are determined by the attitude a person has to the behavior. This means, intentions can predict behavior. In this study, we measure the

(15)

15

attitude a person has towards buying a product on an e-commerce platform, which means we measure purchase intention. Purchase intention can be defined as the intention a person has to buy a certain product.

Jiang et al. (2010) described the stimulus-organism-response (S-O-R) model as a theoretical framework to explain how several features affect e-consumers and their behavior. The S-O-R model shows that several environmental cues act as stimuli that affect a

consumers’ cognitive and affective reaction, which will form their attitude and affect their behavior. According to this model different stimuli (e.g. interaction with a sales person) have an influence on the response (e.g. purchase intention on e-commerce platforms) (Jiang et al.,2010).

2.3.2. The relevance of social presence and e-trust for purchase intention

Van der Heijden et al. (2003) described two dimensions that OS from PS. The first dimension is interaction, which is face-to-face in PS and mediated in OS, which causes a difference in the perception of social presence. The second dimension is trust, according to van der Heijden et al. (2013) a greater degree of trust is required for OS in comparison with PS.

It may be clear that both social presence and e-trust influence purchase intention. Social presence influences purchase indirect with e-trust as mediator. E-trust has a direct effect on purchase intention. Trust is seen as an important determinant for purchase intention. As trust is a prerequisite for social behavior, especially regarding important decisions, like spending money, it’s expected to influence purchase behavior and it’s proven to influence purchase intention (Gefen, 2000; Lu & Fan, 2014).

2.4. Social media

2.4.1. Defining social media

Over the last years the internet has shifted more and more towards user-generated

technologies, like social media networks (Smith, 2009). Social media can be described as highly interactive platforms, where individuals and communities share, co-create, discuss and modify user-generated content (Kietzmann et al., 2011). The social revolution has caused

(16)

several changes in the relation between consumers and sellers. First of all, the consumer opinions about products are not only formed by what the seller tells them about the product, but also by the growing amount of user information online. Second, consumers have a growing network of influence. Third, the opinions consumers find online also influence their offline opinions, as they move into traditional media (Smith, 2009). These changes lead to consumer empowerment, which means that it’s important for companies to listen to their consumers. Therefore, a growing amount of companies is getting active on social media, trying to interact with their consumers.

A growing amount of companies are interacting with their consumers via social media. And, concluding the first three paragraphs of this chapter, the interaction between

e-consumers and sellers in order to increase social presence, e-trust and purchase intention is very important. Interaction with online sellers can be mediated via a chat function on the website, the telephone or e-mail. But since the upcoming use of social media platforms, this medium is getting more and more important for online sellers. Moreover, social media have higher media richness than a chat function or e-mail, which is expected to cause higher social presence (Kaplan & Heanlein, 2010). The telephone also has a high media richness, however this is quite expensive and labor-intensive. Therefore this study will focus on the effects of interaction between e-consumers and sellers via social media on social presence e-trust and purchase intention.

2.4.2. Facebook and Twitter

In the Netherlands, Facebook and Twitter are among the most used social media websites (Internetworldstats.com). Facebook is the largest social media website of the world. It allows users to create a profile where they can post information about themselves, use the chat function to send public or private messages, comment on others and it has a lot of online socializing functions like fan pages and company pages. Another large and interesting social media website is Twitter. It is focused on sharing opinions and information, but it also allows social interactions via private messages or comment on others (Hughes et al., 2012).

(17)

17

not on the private messages, as these are not expected to differ from the chat functions Lu & Fan (2014) studied.

It seems logical to see seller’s presence and interaction with their consumers on social media as a factor of influence on social presence, as defined by Lu & Fan (2014). This is, because the second and third dimension of social presence, perception of others and social presence of interaction with the seller, will be formed by interactions. Social media allows consumer-to-consumer interactions as well as consumer-to-seller interactions. Thereby the first dimension of social presence, social presence of the web, is about social features of the commerce platform itself. Which means this dimension might be influenced by signs at the e-commerce platform that the seller is present on social media, these signs could be Facebook and Twitter buttons.

Although it seems logical to mention the influence of social media on perception of others and social presence of interaction with sellers, Lu & Fan (2014) did not mention this specifically. In their article it is mentioned that chat functions had influence on the perception of interaction with the seller, however, these chat functions differ from social media in the sense that the interaction between consumer and seller is not publicly visible, while they are on social media. Therefore, social media interactions are not only expected to influence perception of social presence of interaction with the seller, but also perception of others, as other consumers can follow the conversation between consumers and sellers, and they are also able to join the conversation. Another difference in influence on social presence between the chat functions as mentioned by Lu & Fan (2014) and social media as tested in this study, is that social media is expected to have a higher media richness than chat functions, and therefore higher social presence is expected after interaction on social media, compared to interaction on a chat function.

Culnan et al. (2010) showed that social media platforms like Facebook and Twitter enable the creation of virtual customer environments. To gain business value, like building social presence, e-trust and purchase intention, e-commerce platforms should fully exploit the capabilities of social media platforms, like informing consumers about products and interact with consumers. Also Hughes et al. (2012) showed that social media is becoming one of the most popular tools for social interaction and information exchange between firms and their consumers.

(18)

Although Facebook and Twitter show many similarities, there are some differences.

Kaplan & Haenlein (2010) made a classification of social media websites by social presence/ media richness and self-presentation/ self-disclosure. See table 1 for their classification. This classification is used in this study, as it shows the expected difference in social presence between Facebook and Twitter. Social presence/ media richness stands for the intimacy, immediacy and the amount of information a social media website allows to be transmitted. Self-presentation/ self-disclosure is the ability that people have to control the impressions other people form of them in any type of social interaction. Facebook clearly belongs to social networking sites with medium social presence/ media richness, whereas Twitter is often seen as a micro-blogging website. Therefore, Twitter can be placed somewhere between blogs and social networking sites. Which means we would expect slightly less social presence for Twitter than for Facebook.

Table 1.

Classification of social media by Kaplan & Haenlein (2010).

2.4.3. Types of interactions on Facebook and Twitter

As described in the last paragraph, Facebook and Twitter both offer several manners to interact. However, all these different types to interact cover two broad types of interactions between online sellers and consumers. The first type of interaction is webcare. Webcare can mean a reaction from the seller to a specific request from a consumer (e.g. a question or complaint about a product). This type of webcare can be called reactive. On the other hand,

(19)

19

webcare can also be proactive. In this case the seller is reacting in response to messages from a consumer, that were not directed to the seller (e.g. the seller responds to a discussion between two consumers) (Noort et al., 2014).

The second type of interaction between online sellers and consumers on Facebook and Twitter is engagement (Sashi, 2012). Engagement interactions consist of interactions that do not answer questions or feedback from consumers, instead these interactions are mainly for fun and building a relationship between consumers and the seller. An example of engagement interaction is a consumer reacting on a status update from the seller or sharing a post from the seller in order to make a chance win something.

According to Noort et al. (2014) webcare is reacting on complaints or questions of consumers on social media platforms, in order to keep the consumer satisfied. Also Sashi (2012) argued that the interactive nature of social media helps to better serve consumers and satisfy their needs. This makes webcare comparable with the job of an employee in a physical store, helping a client out with questions or complaints. Therefore, it is expected that both types of webcare, reactive and proactive, have a positive influence on social presence, as interaction between consumers and sellers increase the perception of social presence.

Furthermore, engagement interactions are expected to have a positive influence on social presence, as it shows the presence of as well the seller as other consumers reacting on the post. Thereby it increases the perception of others and the social presence of interaction with the seller, as Lu & Fan (2014) described. According to the customer engagement cycle as proposed by Sashi (2012), engagement posts lead to connection, interaction and

satisfaction. However, engagement interactions will not be investigated in this study, as this will broaden the scoop too much. Therefore, this study will focus on the effects of proactive and reactive webcare interactions on social presence, e-trust and purchase intention.

2.5. Research question

After reviewing the literature, it can be concluded that little is known about the influence of webcare interactions between online sellers and e-consumers on Facebook and Twitter on social presence, e-trust and purchase intention in e-commerce. To fill this gap in the

(20)

consumers on Facebook and Twitter on the consumers’ perception of social presence, e-trust and purchase intention. Not only the influence of interactions with consumers and sellers on Facebook and Twitter was investigated, also the differences in social presence between Facebook and Twitter were tested. Thereby, the differences between the different types of webcare interactions, proactive and reactive webcare, were analysed. The influence of

Facebook and Twitter buttons on the e-commerce website on social presence of the web were also tested. The primary research question of this study will be:

To what extent have online sellers’ webcare interaction on Facebook and Twitter an influence on social presence, e-trust and purchase intention and does it differ between Facebook and

Twitter?

2.6. Academic and managerial relevance

From an academic perspective this study contributes to the existing knowledge about the positive influence of social presence on e-trust and the positive influence of e-trust on

purchase intention and it fills in the literature gap about the influence of webcare interactions between online sellers and e-consumers on Facebook and Twitter on social presence, e-trust and purchase intention in e-commerce. We already know that social presence increases

purchase intention, and this relationship is mediated by e-trust. However, we don’t know what the influence of webcare interactions between online sellers and e-consumers is on Facebook and Twitter. Therefore, the model of Lu & Fan (2014) will be extended, by adding webcare interactions on social media platforms Facebook and Twitter as a factor of influence on social presence, more specifically the perception of others and social presence of interaction with sellers, and also the direct influence of interactions on Facebook and Twitter on e-trust and purchase intention will be investigated.

From a managerial perspective this study is interesting, as it could help marketing managers to explain the importance of social media, especially Facebook and Twitter, and it could help marketing managers to determine how to use these social media platforms. By understanding the influence of webcare interactions on Facebook and Twitter on social presence, e-trust and purchase intention marketing managers could develop new tactics to

(21)

21

increase purchase intentions. Also this study will give more insights in the differences in social presence between Facebook and Twitter and the importance of the two types of webcare interactions, proactive and reactive. This will help marketing managers in choosing which of these two platforms they should use in their marketing strategy.

2.7. Conceptual model

Following the literature review and research question, the conceptual model that will be tested in this study has been developed. ‘Interactions between online sellers and e-consumers on Facebook and Twitter’ consist of the two types of webcare interactions, proactive and reactive webcare, which are independent variables in this research. Social presence dimensions

‘perceptions of others’ and ‘social presence of interaction with sellers’ are dependent variables when testing the influence of social media interactions on social presence and mediators when testing the influence of social media interactions on e-trust and purchase intention.

The influence of the first dimension of social presence, social presence of the web, will be studied by testing if social media buttons on the e-commerce website influence the dimension of social presence. In this case, the Facebook and Twitter buttons on the website are independent variables and social presence of the web is the depending variable when testing the influence of the social media buttons on social presence of the web and the mediating variable when testing the influence of social media on e-trust and purchase intention.

(22)

Model 1.

Conceptual model.

2.8. Hypotheses

To make the research question testable, some hypotheses have been determined that were tested in this study.

Hypothesis 1

First of all, the relation between Facebook and Twitter buttons on the website and the first dimension of social presence, social presence of the web, were tested. As Facebook and Twitter buttons can be seen as social functionalities of the website, and Lu & Fan (2014) showed that social functionalities on the website increase social presence of the web, a positive relation between these variables is expected.

Hypothesis 1a: Social presence of the web is higher when a website has a Facebook button on the contact page compared to a website without a social media button.

Hypothesis 1b: Social presence of the web is higher when a website has a Twitter button on the contact page compared to a website without a social media button.

(23)

23

Hypothese 1c: Social presence of the web is higher when a website has a Facebook and a Twitter button on the website compared to a website without social media buttons or only one of them (Facebook/Twitter).

Hypothesis 2, 3 & 4

As webcare is comparable with the job of an employee in a physical store, i.e. helping a client out with questions or complaints. It is expected that both types of webcare, reactive and proactive, have a positive influence on social presence. As well perception of others as social presence of interaction with the seller are expected to be positively influenced by webcare, as there is interaction with the seller, and also other consumers can take part in this interaction as the interaction is publicly visible. Although a positive relationship between proactive and reactive webcare and social presence is expected, a higher perception of social presence is expected for proactive than for reactive webcare. This is because other consumers will be more likely to react on this type of interaction, as the message send out by the consumer was initially directed to these other consumers, and the seller reacts on it unexpectedly. This causes a higher perception of others.

Hypothesis 2a: Perception of others is higher when the webshop uses proactive webcare interactions on Facebook compared to when the webshop doesn't have webcare interactions on Facebook.

Hypothesis 2b: Perception of others is higher when the webshop uses proactive webcare interactions on Twitter compared to when the webshop doesn't have webcare interactions on Twitter.

Hypothesis 2c: Perception of others is higher when the webshop uses proactive webcare interactions on Facebook & Twitter compared to when the webshop doesn't have proactive webcare interactions, or only on one of these social platforms (Facebook/Twitter).

(24)

Hypothesis 2d: Perception of others is higher when the webshop uses reactive webcare interactions on Facebook compared to when the webshop doesn't have webcare interactions on Facebook.

Hypothesis 2e: Perception of others is higher when the webshop uses reactive webcare interactions on Twitter compared to when the webshop doesn't have webcare interactions on Twitter.

Hypothesis 2f: Perception of others is higher when the webshop uses reactive webcare interactions on Facebook & Twitter compared to when the webshop doesn't have reactive webcare interactions, or only on one of these social platforms (Facebook/Twitter).

Hypothesis 3a: Social presence of interaction with the seller is higher when the webshop uses proactive webcare interactions on Facebook compared to when the webshop doesn't have webcare interactions on Facebook.

Hypothesis 3b: Social presence of interaction with the seller is higher when the webshop uses proactive webcare interactions on Twitter compared to when the webshop doesn't have webcare interactions on Twitter.

Hypothesis 3c: Social presence of interaction with the seller is higher when the webshop uses proactive webcare interactions on Facebook & Twitter compared to when the webshop doesn't have proactive webcare interactions, or only on one of these social platforms (Facebook/Twitter).

Hypothesis 3d: Social presence of interaction with the seller is higher when the webshop uses reactive webcare interactions on Facebook compared to when the webshop doesn't have webcare interactions on Facebook.

(25)

25

Hypothesis 3e: Social presence of interaction with the seller is higher when the webshop uses reactive webcare interactions on Twitter compared to when the webshop doesn't have

webcare interactions on Twitter.

Hypothesis 3f: Social presence of interaction with the seller is higher when the webshop uses reactive webcare interactions on Facebook & Twitter compared to when the webshop doesn't have reactive webcare interactions, or only on one of these social platforms

(Facebook/Twitter).

Hypothesis 4: A higher perception of social presence for proactive webcare is expected than for reactive webcare.

Hypothesis 5 & 6

Gefen & Straub (2004) describe social presence as an important factor for e-trust, as trust relates to other people and is nourished through interactions with them. Therefore, a positive relation between social presence and e-trust is expected. Trust is seen as an important

determinant for purchase intention. As trust is a prerequisite for social behavior, especially regarding important decisions, like spending money, it’s expected to influence purchase behavior and it’s proven to influence purchase intention (Gefen, 2000; Lu & Fan, 2014).

Hypothesis 5: Social presence has a positive influence on e-trust. Hypothesis 6: E-trust has a positive influence on purchase intention. Hypothesis 7

According to the classification of social media websites by social presence/ media richness and self-presentation/ self-disclosure by Kaplan & Haenlein (2010), a higher perception for social presence is expected for Facebook than for Twitter. This is because Facebook has a slightly higher media richness as Twitter has.

Hypothesis 7: A higher perception of social presence for Facebook is expected than for Twitter.

(26)

3. Methods

3.1. Research design

To give a conclusive answer, quantitative surveys were conducted among 4 groups of at least 40 people, to investigate their perception of social media buttons on the website and

interaction with online sellers on Facebook and Twitter and the influence on social presence, e-trust and purchase intention. The minimum of at least 40 respondents per group was determined to be able to generalize the results. The survey was introduced by a short

description of the fictional webshop called ‘E-shop’. Participants were randomly assigned to 1 of the 4 groups. The division of the groups is as follows:

• Group 1: E-shop.com has no Facebook and no Twitter. • Group 2: E-shop.com only has Facebook

• Group 3: E-shop.com only has Twitter

• Group 4: E-shop.com has Facebook and Twitter

3.2. Sample

Respondents were randomly selected e-commerce consumers, it was just required that the respondents ever bought something online and that they’re familiar with social media platforms Facebook and Twitter. By familiarity with Facebook and/or Twitter it was meant that respondents had an account on these social media platforms that they regularly use or that they have good knowledge about what functionalities this social media platform offer.

Respondents were selected through convenience sampling. This means, a link to the online survey was shared via social media and e-mail.

3.3. Measures

Table 2 gives an overview of the survey questions Lu & Fan (2014) used to measure the three dimensions of social presence. Based on these questions, new survey questions testing the influence of proactive webcare, reactive webcare and engagement interactions on Facebook

(27)

27

and Twitter on social presence, e-trust and purchase behavior were developed. The following subparagraphs will explain the measurement of the variables.

Table 2.

(28)

3.3.1. Social presence of the web

This study measured whether Facebook and Twitter buttons on a webshop have a positive influence on social presence of the web. An image of ‘E-shop’ without Facebook or Twitter button was presented to respondents of group 1, the same image with a Facebook button in it was presented to respondents of group 2, the same image with a Twitter button in it was presented to respondents of group 3 and the image with as well a Facebook button as a Twitter button in it was presented to respondents of group 4. After the image was shown participants were asked to fill if they felt a sense of human contact of the seller on the website and if they felt there is a sense of personal treatment on the website. The more human contact and personal treatment the participants felt, the higher the perception of social presence of the web was suspected.

3.3.2. Perception of others

This study measured whether reactive and proactive webcare on Facebook and Twitter have a positive influence on perception of others. Therefore, an introduction of a situation in which ‘E-shop’ has no Facebook or Twitter page was presented to group 1, so no reactive or proactive webcare. An introduction and example of reactive and proactive webcare on Facebook was presented to group 2. An introduction and example of reactive and proactive webcare on Twitter was presented to group 3 and group 4 got an introduction and example of reactive and proactive webcare on Facebook and Twitter. First, participants read the

introduction and the example of the webcare interaction. Than, the participants were asked to fill in if they could also sense other consumers who provide information about the seller and if they felt they could sense others who provide information about the product. The more participants could sense other consumers, the higher the perception of others.

3.3.3. Social presence of the seller

Social presence of the seller was measured in the same way as perception of others was measured. Namely, the influence of reactive and proactive webcare on Facebook and Twitter

(29)

29

on social presence of the seller was measured. The introduction and examples that were shown to the 4 groups, are the same as described in the subparagraph above. After reading the introduction and perceiving the example the participants were asked to fill in if they felt that they could make sense of the attitude of the seller and if they could imagine how the seller may look like. The more sense of the attitude of the seller participants can make and the more they can imagine how the seller may look, the higher the perception of social presence of the seller.

3.3.4. E-trust

In all situations the trust in the seller was measured. After reading the introductions and seeing the examples, participants were asked to fill in if they felt that promises made by the seller were likely to be reliable and if they do not doubt about the honesty of the seller. The more they felt that promises made by the seller were likely to be reliable and the less they doubt about the honesty of the seller, the higher the e-trust.

3.3.5. Purchase intention

Just like e-trust, purchase intention was measured for all groups in all described situations. This means that after reading the introductions and seeing the examples participants were asked if they were likely to buy a product from the seller and if they considered buying a product from the seller in the future. The more participants were likely to buy a product and considered buying a product from the seller in the future, the higher the purchase intention.

3.4. Survey design

Respondents were first asked if they ever bought something online and if they are familiar with Facebook and Twitter, which resulted in nominal variables. Only after answering both of these questions positively, participants were included in the survey.

(30)

After randomized assignment to one of the four survey groups, participants were given a short introduction of the fictional webshop ‘E-shop.com’ and afterwards they were asked to fill in the survey. The survey for group 1 (no Facebook and no Twitter) consisted of 40 questions and the surveys for group 2, 3 and 4 consist of 60 questions with a five-point Liker scale, ranging from strongly disagree to strongly agree. Completing the survey will take approximately 10 minutes (surveymonkey.com, consulted on October 20th 2015).

The last questions were about gender (nominal variable) and age (ratio variable). These questions were included because there could be a gender difference in the perception of social presence and also age could play an important role in the perception of social presence and e-trust. This is, as people who grew up with e-commerce platforms are expected to have a higher perception of e-trust compared to people who only used to shop in physical shops.

4. Data analysis

4.1. Descriptive data of sample

In total 442 participants opened the survey, 260 cases were valid, which means participants filled something in. Of which a total of 182 (41,2 %) valid cases with no missing items were included in this study. In total 101 males and 81 females filled in the questionnaire. A Pearson chi-square test on gender and groups revealed a non-significant (χ2 (3, n = 182) = 2.513, p = .473) result, so groups did not differ on gender. An independent sample t-test on the groups showed that groups did not differ on age (t(65)= .821, p = .42).

The constructs were checked on realiability using the Cronbach’s Alpha. Table 4 shows that the Cronbach’s Alpha per construct is for almost every construct > 0.7, which means these constructs are reliable. There is one construct that is a little below 0.7, which is e-trust (integrity). However, because this construct is 0.69, which is almost 0.7, I still expect this construct to be reliable.

(31)

31 Table 3.

Group distribution.

Group Frequency Percent Valid percent Cumulative

percent 1 47 10.6 18.1 18.1 2 116 26.2 44.6 62.7 3 52 11.8 20.0 82.7 4 45 10.2 17.3 100.0 Total 260 58.8 100.0 No missing 999 182 41.2 Total 442 100.0 Table 4. Reliability of constructs. Cronbach’s Alpha N of items

Social presence of the web 0.88 5

Perception of others 0.91 4

Social presence of interaction with the seller 0.98 4

E-trust (benevolence) 0.88 4

E-trust (integrity) 0.69 4

E-trust (competence) 0.74 4

Purchase intention 0.92 3

4.2. Hypothesis testing

Next step in the data analysis process is testing the hypothesis. Model 2 shows the relation between the variables of the conceptual model and the hypothesis tested in this study. This paragraph will describe the steps taken per hypothesis. Participants were asked to rate statements on a five point Likert scale ranging from strongly disagree to strongly agree. The

(32)

influence of Facebook and/or Twitter buttons on the webshop on social presence of the web was measured with the following statements.

I feel there is a sense of human contact of the seller on the website. I feel there is a sense of personal contact with the seller on the website. I feel there is a sense of sociability in the website.

I feel there is a sense of human warmth in the website. I feel there is a sense of human sensitivity on the website.

The influence of proactive and reactive webcare on Facebook and/or Twitter on perception of others was measured with the following statements:

I feel I can sense others who feel interest with the product.

I feel I can sense others who provide information about the seller. I feel I can sense others who provide information about the product. I feel I can sense others who have browsed Eshop.com.

The influence of proactive and reactive webcare on Facebook and/or Twitter on social presence of interaction with the seller was measured with the following statements:

I feel I can make sense of the attitude of the seller by interacting with him/her. I can imagine how the seller may look like by interacting with him/her.

I feel there is a sense of human touch to communicate with the seller. I feel communication with the seller is warm.

The influence of social presence on trust was measured with help of the 3 dimensions of e-trust; benevolence, integrity en competence. Benevolence was measured with the following statements:

I believe the seller would act in my best interest.

I feel that if I required help, the seller would do his/her best to help me. I feel that seller is interested in my wellbeing.

(33)

33 Integrity was measured with the following statements:

I feel that promises made by the seller are likely to be reliable. I do not doubt the honesty of the seller.

I feel that the seller is sincere and genuine.

I expect that the seller will keep promises he/she makes. Compentence was measured with the following statements:

I feel that the seller is competent and effective. I feel that the seller performs his/her role very well. I feel that the seller knows about the product.

I feel that the seller knows how to provide excellent service. Purchase intention was measured with the following statements:

I am very likely to buy a product from the seller.

I consider buying a product from the seller in the future. I intend to buy the product from the seller.

Model 2.

(34)

Table 4.

Descriptive data overview.

Social presence of the web

Perception of others Social presence of interaction with

the seller

Proactive webcare Reactive webcare Proactive webcare Reactive webcare

M SD n M SD n M SD n M SD n M SD n Group1 32.82 7.41 38 11.88 3.19 43 11.88 3.19 43 47.73 7.33 15 47.73 7.33 15 Group2 40.26 7.99 82 14.99 2.68 87 42.61 6.31 54 41.21 7.23 71 45.47 6.71 15 Group3 39.89 8.04 38 12.13 3.75 40 36.79 8.14 29 39.95 4.61 21 41.52 6.06 36 Group4 40.91 7.78 33 15.00 2.25 35 42.06 6.16 33 14.31 2.14 39 13.58 2.87 34 Hypothesis 1

An independent sample t-test between group 1 and 2 showed a significant (t(120) = -,873, p = <.001) lower mean (M = 32.82; n = 38; SD = 7.41) on social presence of the web was found for the contact page without social media buttons compared to the contact page with a

Facebook button (M = 40.26; n=82; SD= 7.99). An independent sample t-test between group 1 and 3 showed that the contact page without social media buttons scored significant (t(72) =

-3.99, <.001) lower (M= 32.82; n= 38; SD = 7.41) on social presence of the web than the

contact page with a Twitter button (M = 39,89; n=38; SD = 8,04). In summary, both a Facebook button and a Twitter button on the contact page result in higher social presence of the web than having no social media button on the contact page. Therefore, both hypothesis 1a and hypothesis 1b are supported.

As shown with an independent sample t-test, social presence of the web is

significantly (t(69) = -4.49, <.001) higher on a website with a Facebook and Twitter button (M= 40.91; n=33; SD= 7.78), compared to a website without social media buttons (M= 32.82; n=38; SD= 7.41). However, no significant difference between a website with Facebook and Twitter button (M= 40.91; n=33; SD= 7.78), and a website with only a

Facebook button (M = 40.26; n=82; SD= 7.99) or a website with only a Twitter button (M = 39,89; n=38; SD = 8,04). Therefore, means hypothesis 1c is partly supported.

(35)

35 Hypothesis 2

The perception of others is significant (t (128) = -5.83, p= <.001) higher for proactive Facebook interactions (M= 14.99; n=87; SD = 2.68) than for no Facebook webcare interactions (M= 11.88; n=43; SD= 3.19). Therefore, hypothesis 2a is supported. An

independent sample t-test between group 1 and 3 showed that perception of others was not significant (t (81) = -3.16, p= .753) higher for proactive Twitter webcare interactions (M= 12.13; n=40; SD=3.75) than for no Twitter webcare interactions (M=11.88; n=43; SD= 3.19). Therefore, hypothesis 2b is rejected. The results also showed that perception of others

is significantly (t (76) = -4.87, p = <.001) higher with proactive webcare interactions on Facebook and Twitter (M = 15.00; n = 35; SD = 2.25) than no webcare interactions (M = 11.884; n = 43; SD = 3.194), or only proactive webcareinteractions on Twitter

(M= 12.13;

n=40; SD=3.75). However, the perception of others is not significantly higher when E-shop

uses proactive webcare interactions Facebook and Twitter compared to just on Facebook

(M=

14.99; n=87; SD = 2.68). Therefore, hypothesis 3c is partly supported.

An independent sample t-test showed a higher perception of others for reactive Facebook webcare interactions (M= 42.61; n= 54; SD 6.31) was found compared to no Facebook webcare interactions (M= 11.88; n=43; SD= 3.19), so hypothesis 2d is supported. Perception of others was also higher for reactive webcare interactions on Twitter (M = 36.79; n = 29; SD 8.14) webcare than no webcare interactions on Twitter (M=11.88; n=43; SD= 3.19).

Therefore, also hypothesis 2e is supported. Perception of others is also significantly higher when E-shop uses reactive webcare on Facebook and Twitter (M = 42.06; n = 33; SD =

6.164) compared to when E-shop does not use any webcare interactions (M=11.88; n=43;

SD= 3.19), or only on Twitter (M = 36.79; n = 29; SD 8.14). However, the perception of others is not significantly higher when E-shop uses reactive webcare interactions Facebook and Twitter compared to just on Facebook (M= 42.61; n= 54; SD 6.31) Therefore, hypothesis 2f is partly supported.

Hypothesis 3

Proactive webcare on Facebook (M = 41.21; n=71;SD = 7.23) scored significantly lower (Mdif = 6.52, CI = 1.81-11.24, p = .003) on social presence of interaction with the seller than no webcare. Therefore, hypothesis 3a is rejected.

(36)

Moreover, proactive webcare interactions on Twitter (M = 39.95; n= 21; SD = 4.61) scored significant (Mdif = 7.78, CI = 2.30-13.26, p = .003) lower on social presence of interaction with the sellers than having no webcare (M= 47.73;n=15; SD=7.33). Therefore, also hypothesis 3b is rejected. The results also showed that social presence of interaction with the seller is significantly (t (75) = -5.86, p = <.001) higher when the webshop uses proactive webcare interactions on Facebook and Twitter (M = 14.31; n = 39 , SD = 2.14,), compared to when the webshop does not use any webcare interactions (M = 10.47; n = 38; SD = 3.46). However, the combination of using proactive webcare on Facebook and Twitter compared to only on Facebook or only on Twitter, does not cause higher social presence of interaction with the seller. Therefore, hypothesis 3c is partly supported.

In contrast, after conducting a one-way ANOVA and Bonferroni correction for

multiple testing it was shown that reactive webcare interactions on Facebook (M =45. 47; n= 15; SD = 6.71) and on Twitter (M = 41.52; n=36; SD = 6.06) score higher on social presence of interaction with the seller than having no webcare interactions on Facebook or Twitter. Therefore, hypothesis 3d and 3e are supported. The results also showed that social presence of interaction with the seller scores significantly (t (69.63) = -4.17, p = <.001) higher when the webshop uses reactive webcare interactions on Facebook and Twitter (M = 13.59; n = 34

SD = 2.87

) compared to when the webshop does not use any webcare interactions. However,

the combination of using reactive webcare on Facebook and Twitter compared to only on Facebook or only on Twitter, does not cause higher social presence of interaction with the seller. Therefore, hypothesis 3f is partly supported.

Hypothesis 4

Reactive webcare (M = 38.26; n = 53; SD 8.78) did not differ significantly (t (134) = -.536, p = .593) from proactive webcare on perception of others (M = 39,10; n = 83; SD 8,87).

(37)

37 Hypothesis 5

Social presence correlates significantly (p = <.001) with e-trust explaining R2_{=.357604. In}

other words, the data tells us that social presence explains 35.7% of the variance. Therefore, hypothesis 5 is supported.

Hypothesis 6

E-trust correlates significantly (p = <.001) with purchase intention explaining R2=.4651 of variance. In other words, e-trust explains 46,5% of the variance in purchase intention. Therefore, hypothesis 6 is supported.

Hypothesis 7

Although Facebook had a higher social presence (M = 40.26; n= 84; SD 7.99) than Twitter (M = 39.89; n = 38; SD 8.04), no statistical (t (120) = .235, p = .815) difference was observed. Therefore, hypothesis 7 is rejected.

Table 5.

Overview hypothesis testing.

Hypothesis 1

a. Social presence of the web is higher when a website has a Facebook button on the contact page compared to a website without a social media button.

Accepted b. Social presence of the web is higher when a website has a Twitter button on the contact

page compared to a website without a social media button.

Accepted c. Social presence of the web is higher when a website has a Facebook and a Twitter button

on the website compared to a website without social media buttons or only one of them (Facebook/Twitter).

Partly supported

Hypothesis 2

a. Perception of others is higher when the webshop uses proactive webcare interactions on Facebook compared to when the webshop doesn't have webcare interactions on Facebook.

Accepted b. Perception of others is higher when the webshop uses proactive webcare interactions on

Twitter compared to when the webshop doesn't have webcare interactions on Twitter.

Rejected c. Perception of others is higher when the webshop uses proactive webcare interactions on

Partly supported d. Perception of others is higher when the webshop uses reactive webcare interactions on

Facebook compared to when the webshop doesn't have webcare interactions on Facebook.

Accepted e. Perception of others is higher when the webshop uses reactive webcare interactions on

Accepted f. Perception of others is higher when the webshop uses reactive webcare interactions on

Facebook & Twitter compared to when the webshop doesn't have reactive webcare

Partly supported

(38)

interactions, or only on one of these social platforms (Facebook/Twitter).

Hypothesis 3

a. Social presence of interaction with the seller is higher when the webshop uses proactive webcare interactions on Facebook compared to when the webshop doesn't have webcare interactions on Facebook.

Rejected b. Social presence of interaction with the seller is higher when the webshop uses proactive

webcare interactions on Twitter compared to when the webshop doesn't have webcare interactions on Twitter.

Rejected c. Social presence of interaction with the seller is higher when the webshop uses proactive

webcare interactions on Facebook & Twitter compared to when the webshop doesn't have proactive webcare interactions, or only on one of these social platforms (Facebook/Twitter).

Partly supported d. Social presence of interaction with the seller is higher when the webshop uses reactive

webcare interactions on Facebook compared to when the webshop doesn't have webcare interactions on Facebook.

Accepted e. Social presence of interaction with the seller is higher when the webshop uses reactive

webcare interactions on Twitter compared to when the webshop doesn't have webcare interactions on Twitter.

Accepted f. Social presence of interaction with the seller is higher when the webshop uses reactive

webcare interactions on Facebook & Twitter compared to when the webshop doesn't have reactive webcare interactions, or only on one of these social platforms (Facebook/Twitter).

Partly supported

Hypothesis 4

A higher perception of social presence for proactive webcare is expected than for reactive webcare.

Rejected

Hypothesis 5

Social presence has a positive influence on e-trust. Accepted

Hypothesis 6

E-trust has a positive influence on purchase intention. Accepted

Hypothesis 7

A higher perception of social presence for Facebook is expected than for Twitter. Rejected

5. Discussion

5.1. Discussion of findings

The results supported a large part of the hypothesis that followed from the theory, however there were also some surprising results, that will be discussed in this paragraph.

First of all, following the theory, proactive webcare on Facebook and Twitter was expected to have a positive influence on the perception of social presence, even more than reactive webcare. However, proactive webcare scored significantly lower on social presence

(39)

39

of interaction with the seller than having no webcare on both Facebook and Twitter. Thereby, proactive webcare on Twitter didn’t score higher on perception of others than having no webcare. For Facebook, this result might be influenced by the uneven groups with N= 15 for no webcare and N= 71 for proactive webcare on Facebook. However, for Twitter there are no uneven groups to declare this difference. Also, it was expected that proactive webcare had a more positive influence on social presence compared to reactive webcare, but the results showed it is the other way around. This means, customers apparently appreciate reactive webcare more than proactive webcare to increase their perception of social presence.

This could be explained by the following theory; customers can perceive the monitoring of networks by companies as an intrusion into their privacy, even when this monitoring is on an online network that is publicly visible (Bekkers et al., 2013). For proactive webcare, the webshop is monitoring messages and responds on some of them, without the customer asking for a response from the webshop. In contrast to reactive webcare, where the customer asks for a response from the webshop. The feeling of intrusion of their privacy could be the reason customers don’t experience proactive webcare as a positive influence on perception of social presence.

However, proactive webcare did have a positive effect on perception of others on Facebook. This could possibly be declared by the media richness theory. As the theory described, media richness stands for the intimacy, immediacy and the amount of information a social media website allows to be transmitted (Kaplan & Haenlein, 2010). The theory also described that Twitter has a little less media richness compared to Facebook. This could explain the fact that proactive webcare doesn’t increase the perception of social presence on Twitter, while it does increase one dimension of social presence, the perception of others, on Facebook, as proactive webcare is more intimate and immediate compared to reactive webcare.

Second, it is strange to see that having proactive webcare on Twitter doesn’t cause a higher perception of the second and third dimension of social presence, perception of others and social presence of interaction with the seller, than having no webcare and having proactive webcare on Facebook doesn’t cause a higher perception of the third dimension of social presence. But having proactive webcare on both of these social media platforms does increase the second and third dimension of social presence. It could be that when reading the

The influence of Facebook and Twitter on the perception of social presence, e-trust and purchase intention in e-commerce

Master Thesis