Value creation within social network sites
A multiple case study
R. Stomp
2
Value creation within social network sites
A multiple case study
Groningen, January 2011
University of Groningen
Faculty of Economics & Business Msc Business Administration Strategy and Innovation
Name: Reinier Stomp Student number: 1466526
3
Abstract
The main research question of this paper is Which factors determine the success of social
network sites such as Facebook, Myspace, and Hyves?. This paper suggests that these sites’
4
Preface
‘We lived in farms, then we lived in cities, and now we're going to live on the internet.’
(Quote from the movie The Social Network)
From making friends and watching each other’s summer holiday pictures to sharing the latest political news and harvesting crops on Farmville: social network sites offer a previously unprecedented channel for people to interact with each other from anywhere in the world at every moment of the day.
I highly enjoyed investigating this phenomenon which is gradually becoming part of everybody’s daily life. From a social point of view social network sites present form a whole new way in which people interact with each other. From a business point of view they present a unique example of turning the opportunities created by the internet into highly valuable businesses. This gave me a strong motivation to study this interesting new industry.
I would like to thank Thijs Broekhuizen and Pedro de Faria for their supervision and inspiring reflections on my earlier versions of this thesis which highly contributed to its final result. Finally, I want to thank my family for their continuous support during my study.
Groningen, January 2011,
5
Table of content
Chapter / Title Page
Paragraph 1 Introduction 6 1.1 Problem Definition 8 1.2 Relevance 9 1.3 Case introduction 10 2 Theoretical framework 14
2.1 Battle for dominant design 14
2.2 Value creation 16
2.3 Relationships between social network sites 17 and third party developers
2.4 Value appropriation 21 2.5 Conceptual framework 23 3 Research methodology 25 3.1 Data collection 25 3.2 Measurement 26 4 Cases 31 4.1 Facebook vs. Myspace 31
4.2 Facebook vs. Hyves in the Netherlands 47
4.3 Case results 55
5 Conclusions 58
5.1 Results 58
5.2 Implications 61
5.3 Limitations and future research 62
6 References 65
Appendix I Social network sites in the Netherlands 69
Appendix II Valuation of Facebook 70
Appendix III Social network sites in Dutch newspapers 72 Appendix IV Social network sites in international newspapers 73 Appendix V Worldwide visitors of Facebook and Myspace 74
6
1. Introduction
Looking back from 2036, we can see that the internet did indeed changed everything. The world began an historic transition from industrial capitalism to a new kind of economy based on networks, new principles and new ways of creating value.1
The emergence of the internet has pushed the development of numerous new technologies and companies. Services as Amazon, Google, Craiglist, E-Bay, Wikipedia and others have completely changed the way in which people interact with each other. While some of these services changed current ways of doing things, others have emerged as completely new industries. One of these is the industry of social network sites, sites on which users create profiles, maintain friendships and even play games with each other.
Over the last few years, Facebook has become the dominant social network site in the world2. Especially since the reach of its main competitor, Myspace, began to decline in 2008, Facebook has grown to more than 400 million users worldwide3,4 and currently says that it has more than 500 million active users. Half of them even log in to the site every single day5.
There are only a few regional markets left in which Facebook has not become the most dominant social network site6. However, it was only a three years ago that there were just a few regions in which Facebook was dominant7. This puts forth the question of what happened in the past years that caused Facebook to reach such spectacular growth rates.
The industry of social network sites hardly existed ten years ago. The market emerged and increased rapidly over the past years. Despite launching later than sites such as Friendster and Myspace, Facebook managed to become the most dominant social network site. An important element for these sites are the increasing returns of more users; as more users join they interact more and thus the site becomes more valuable for every individual user, also known as network effects. Another interesting feature of Facebook is that it uses an open architecture
7 through which users and other companies can create their own applications within the network. This feature was already introduced back in 2007 when Myspace was still the dominant social network site. This open system created a whole new trend of social gaming or casual gaming as it is also being called within the social network sites. Applications like Mafia Wars and Farmville have millions of daily users8,9. These applications were not developed or owned by Facebook itself but by an external company called Zynga. This relatively young company (founded in 200710) has managed create numerous of Facebook-applications which attract millions of users every day. Also, it has managed to attract hundreds of millions of funds from investors to finance its tremendous growth11. While this is the most known application developer, numerous of other companies and individuals create applications for Facebook.
In 2008, Myspace also introduced an application page in which users and other companies could share their applications12,13. By that time, Facebook had already matched Myspace’s monthly user numbers14. What were the reasons behind these developments and which factors determined the value of both Myspace and Facebook? Currently, little is known about the drivers of success of these social network sites.
Next to the dominance of Facebook and Myspace there are some local markets in which other social network sites have a dominant position15. One of these markets is the Netherlands, in which Hyves has since long been the most used social network site. Hyves differs from Facebook and Myspace not only in that it is a local actor but also by its organizational form. Users of Hyves can upload photos, their travel stories and other content but cannot freely create applications such as in Facebook. However in the spring of 2010, Hyves also launched a game-application within its network in a partnership with Spil-games16. This partnership however is very different from the one between Zynga and Facebook since in this case there is a very close collaboration between the two companies and Spil-games is the exclusive
8 provider of games within Hyves. Although both strategies add value to the experience of the end user it needs to be discovered which organizational form is the most effective.
It is unclear how the different organizational forms of these social network sites have affected their growth and value. There is no specific knowledge available whether such aspects have had a large impact on user numbers or that other factors had a larger influence. Such factors include direct network effects, the media attention the social network sites have received as well as their technological capabilities and objective quality.
1.1. Problem definition
The first aim of this thesis is to give an in-depth insight of how a battle for dominant design developed between Myspace and Facebook on a worldwide level. Second, this thesis will investigate the competition between Facebook and Hyves within the Netherlands. This thesis will investigate the creation of the value by the various actors in these cases and will give a detailed description of the business models of the various cases. Finally, it will give a detailed description of how the actors captured value from their activities.
The main research question for this paper is ‘Which factors determine the success of social
network sites such as Facebook, Myspace, and Hyves?’. To answer this question a number of
sub questions is used, such as ‘How are MySpace, Facebook and Hyves organized?’, ‘Which
factors determine the value creation of the social network sites?’, ´How does the organizational form regarding third party suppliers of the various social network sites affect their growth and profitability?´. These questions will help me to answer the main research
question and to create a detailed insight of this research area.
9 Within this paper, the terms open and closed refer to the collaboration forms of the social media with external parties. An open model implies that the medium is open to everyone while in a closed model the social network site has a contract with a limited or only one external party. This difference will be further explained in the theoretical framework.
This problem definition also implies the specific focus of this paper. Whereas it is possible to focus on different actors in these relationships, such as consumers or application-developers, this paper is mainly focused on the social network site itself, thus Facebook, Myspace and Hyves, and less on the nature of the third parties that supply applications for these social network sites. These cases will create an insight in how open business models can create value and how a business model can be built around them in various ways.
1.2. Relevance
Over the last years, social network sites have become a very dominant aspect of people’s daily internet routine. Hundreds of millions of users a day visit Facebook17 and also a local player such as Hyves has a reach of around two million users a day18. Also, Facebook overtook Google as the most visited website in the world in 201019,20. Morgan Stanley research even stated that social network sites currently attract more users than e-mail21. Because of this huge reach, many aspects of these social network sites affect the daily lives of hundreds of millions of people. So research in this area has a significant value in the light of social relevance. Also, because creators of applications are among the first actors other than the social network sites themselves to create significant value in this area, they form an interesting case and can perhaps form an example for other businesses.
Social network sites are a young phenomenon and thus a relatively unexplored area when it comes to publications. Especially within the research area of value creation and value capturing, social network sites are an almost uninvestigated research area. This follows the
10 trend described by Shugan (2004) that academic research tends to focus on past research topics while it might be more interesting to focus research on the future in order to identify more enduring trends. More specifically, research on social network sites is limited by a lack of experimental or longitudinal studies and there is a need for large-scale quantitative and qualitative research on this topic (Boyd & Ellison, 2007). Papers on this topic are thus of tremendous academic relevance as they can form the base of a currently underdeveloped research area. Although there have been a few scientific publications about social network sites22, few of them look at the specific reasons for their growth, which will be the focus of this paper.
From a business point of view, social network sites present a tremendous opportunity for current as well as new ventures. Knowledge about this industry is still scarce and this paper can thus create useful insights regarding their dynamics. For current owners of social network sites and managers in this industry, this paper will create insights in how they can create value and attract users. Furthermore this paper will describe the effects of media attention and other issues such as the availability of complementary services on the value of a social network site. Managers with knowledge about these issues can use them to stimulate the success of their site.
This research will also create useful insights for application providers as it will describe their importance to the success of the social network sites. Additionally, the insights from this industry might also be relevant to other industries. Similar dynamics may occur in other industries where direct and indirect network effects are present. One such example might be smart phones where the phones also operate as a platform for application providers.
1.3. Case Introduction
This paper is focused on three specific social network sites, Myspace, Facebook and Hyves. In this paragraph Myspace and Facebook will be introduced first and after that Hyves will be described with a short introduction.
22
11 Myspace and Facebook were both founded in the United States and basically let people create a profile and share their thoughts and activities with their friends through their online profile. Myspace was launched in 2003 and quickly became one of the most visited websites in the United States and many other countries23. By 2006, many people believed Myspace to become a very dominant factor in people’s daily internet routines24
. By 2007 MySpace was by far the most visited social network site. However, during 2007 its audience began to stabilize and Facebook overtook its position as the world most used social network site in April 200825.
Facebook’s user numbers more than doubled between September 2007 and September 2008. This was in spite of the presence of Myspace as the dominant social network site at that moment. As shown in figure 1, by 2010 Facebook has completely dominated Myspace in terms of users and the time users spend on the website26.
Figure 1: Number of monthly unique visitors of Myspace and Facebook (Created by author, based on data in appendix V)
23 http://www.web-strategist.com/blog/2008/01/09/social-network-stats-facebook-myspace-reunion-jan-2008/ 24 http://mashable.com/2006/03/27/if-you-dont-get-myspace-youre-a-lametard/
25
For a complete overview of the history of Myspace, see the book ‘Stealing Myspace: The Battle to Control the Most Popular Website in America’ by Julia Angwin. Random House, March 2009. ISBN 1400066948.
26 For a complete overview of the history of Facebook, see the book ‘The Facebook Effect: The Inside Story of
12 Clearly, Facebook is currently much larger than Myspace in total user numbers. The positions of these social network sites have completely changed within just a few years. This paper will focus on the reasons of this shift in popularity between Myspace and Facebook. Furthermore the business models of these sites will be assessed as well as the value appropriating tactics of both companies.
The Netherlands
One of the few markets in which Facebook did not yet become the most used social network site by 2010 is the Netherlands, where it is second to Hyves. Hyves was founded in 2004 and currently has around 10 million accounts. Interestingly, while Hyves still seems to be the most dominant social network site in the Netherlands but Facebook is gaining terrain. Figure 2 demonstrates this with user numbers from 2009 & 2010.
Figure 2: Daily unique visitors of Hyves and Facebook in the Netherlands (Created by author, based on data in appendix I)
One important element which has to be mentioned here is that when we focus on the openness of the social networks to other parties its system is very distinctive from that of Facebook. In
13 a schematic way the relationship with outside application developers of Hyves and Facebook various social network sites can be drawn as follows in figure 3.
Figure 3: The closed system of Hyves and the open system of Facebook. Source: Author.
Both the table and the pictures above clearly give an impression of the main difference between the various social network sites with regard to their openness to external application developers. The implications of these differences will be further elaborated on in the fourth chapter of this paper.
14
2. Theoretical Framework
This chapter will give an overview of the theoretical foundations for this case study. It will give an overview of existing theories in the field of organizational forms and battles for dominant design as well as their relationships with social network sites. In the last paragraph of this chapter these theories will be combined to a solid research design for the case study part of this paper.
2.1. Battle for dominant design
Battles for dominant design are a common phenomenon with regard to emerging technologies. Think for instance about the HD-DVD versus Blu-Ray battle (Perry, 2008) and the video game console-war (Schilling, 2003).
Soh (2010) studied network effects regarding the battle for a dominant design for a standard to let computers interact with each other on a network which we now know as the Local Area Network (LAN). He found that network effects can be a key opportunity for a firm as to acquire complementary products and resources and in the end, win the battle for a dominant design. According to Schilling (2003), other factors that hugely influence the outcome of such a battle are product quality and the installed base of a product.
For a new entrant to win a battle for dominant design Schilling (2003) presents three strategies to create value as well as three similar strategies for incumbents to protect themselves against new entrants:
1. Create/prevent a technological gap with your objective quality
2. Build/protect an installed base and availability of complementary goods 3. Shape perceptions and expectations through signaling
15 needs to leverage its reputation and create an impression that its installed base is, or will soon be, very large.
Important to mention here is that Schilling (2003) states that her concepts should apply equally well to many other industries demonstrating network externalities, a concept which will be explained in the next paragraph. Another interesting point with regard to this paper is that in both these models the availability of complementary goods is an important strategy. Teece (1986) describes multiple forms of both specialized as generic complementary assets. Within this paper the focus is on complementary services in the form of applications which run on a social network site. An important question about this aspect is how new entrants can actually acquire the availability of complementary goods.
Schilling (2002) states that companies may wish to sponsor the development of complementary goods to jump-start their own virtuous growth circles. In this perspective the difference between open and closed business models pops up as an important aspect; can new entrants use an open business model to acquire the availability of complementary goods and thus strengthen their value creation and their position in the battle for a dominant design? Any new business needs a particular business model which describes how it creates and delivers value as well as the way the business captures this value (Teece, 2010).
2.2. Value creation
16 An interesting element of this definition is the role of consumers. In order to create value a company has to deliver an amount of value or novelty to its consumers. This aspect is also included in the definition of a business model by Teece (2010) which was mentioned above. A key element of social network sites is that people use them to interact with each other and that they have little value to a user whose friends do not also use the same site. This puts social media in the same category as telephones and fax machines, which are also products which become more valuable as more people use them. Within academic literature this phenomenon was introduced by Katz & Srapiro (1985) as ‘network externalities’.
Network externalities are present when the perceived value of an innovation depends on the amount of consumers who have adopted it (Katz & Srapiro, 1986). In addition, research has identified various product categories (for example software, financial services and consumers electronics) in which the market performance of an innovation is not only a function of the innovation’s installed base but also of the availability of complementary products. For instance the value of a game console is depending on the availability of games to play on it (Song et al., 2009). This has effect has been defined as the ‘indirect network effect’ (Stremersch et al., 2007).
These direct and indirect network effects are relevant for the success of social network sites because of their influence on their value. Regarding to direct network effects, social network sites are only valuable, and become more valuable, if one’s peers also join them. In addition, social network sites can be seen as platforms in which applications act as complementary services (Cusumano, 2010). These complementary services are formed by applications which can create indirect network effects and thus enhance the value of the social network site.
So, in short the value of a social network site depends on three main dynamics: 1. The installed base of the site which can create direct network effects.
2. The availability and usage of applications which can create indirect network effects. 3. The objective quality of the website determined by its features, and technological
17 2.3. Relationship between social network sites and third party developers
The previous paragraph stressed the importance of the availability of complementary products. Within traditional closed business models, companies rely primarily or even solely on their own knowledge and R&D-departments to stimulate innovation and create new products . In this situation innovation only comes from within the company and innovation projects are primarily focused on matching the companies’ current strategies. Figure 4 forms a representation of this situation.
Chesbrough (2003) describes that although the closed innovation proved its value in the 20th century, its value began to erode by the approach of the current millennium. He describes a situation where companies form an open innovation model in which knowledge and R&D from outside the firm is used as a source for innovation. As described in figure 5, the boundaries of a firm become a semi-permeable membrane through which innovations can easily move between the firm and its surroundings.
Figure 4: Closed innovation model (Chesbrough, 2003)
18 As can be seen in these pictures an important difference between the two innovation models is that in the closed model, the company pushes its own ideas to the market, whereas in the open model the company adopts more ideas from outside the company.
A detailed typology of how organizations can form relationships with other companies is further given by Powell (1990), who defines market, hierarchy and network forms of organization. He defines market relationships as price-based relationships in which actors are independent of each other and in which there is a low mutual commitment. Within a hierarchy, there is a formal relationship between two parties which can be bureaucratic and inflexible. Finally, within a network relationship, parties are interdependent and focused on mutual benefits and complementary strengths. Table 1 describes all factors Powell describes in his typology of these various relationships.
Market Hierarchy Network
Normative basis Contract – property rights
Employment relation Complementary strengths
Means of communication
Prices Routines Relational
Methods of conflict resolution Haggling – resort to courts for enforcement Administrative fiat - supervision Norm of reciprocity – reputational concerns
Degree of flexibility High Low Medium
Amount of
commitment among the parties
Low Medium to high Medium to high
Tone or climate Precision and/or suspicion
Formal, bureaucratic Open-ended, mutual benefits
Actor preferences or choices
Independent Dependent Interdependent
Table 1. - Forms of Economic Organization, adapted from Powell (1990).
19 disadvantage that the company can solely build on its current knowledge and research capabilities without adapting to emerging knowledge and developments outside the company. The open business model presents a solution for this; it gives to opportunity to adapt quickly to outside knowledge and to develop projects in cooperation with the market. Because of their flexible nature, network forms of organization are well suited to adapt to changing environments. Hierarchical forms of organization, which are far less flexible, have more difficulties adapting to a changing environment. Within the field of social network sites these characterizations take form in the hierarchical system of Hyves which has one sole game supplier and the open system of Facebook which matches the network form of organization.
An addition to these models is further explained by the models of technology push and need pull (Burgelman & Sayles, 1986). They argue that in a technology push-model, companies mainly focus on building on their current R&D-knowledge. However, this creates the risk of getting locked-in into one technical solution. On the other hand need-pull models have the risk of focusing too much on current customer needs that will only demand minor advantages rather than breakthrough innovations.
20 Figure 6: Punctuated equilibrium with cumulative network externality and uncertainty.
(Adopted from Hwang & Oh, 2009)
This figure demonstrates that in a growing market a shift in utility can have a strong effect on market share, even in the case of a first mover advantage by one of the competitors. As a result a growing market with network effects present can be highly dynamic and unpredictable. The authors also argue that their model might be applied to various online services which have network effects present.
The next subject of this paper will be how companies can actually capture value from the value they created. With regard to the emerging industries in this paper it is especially interesting how a dominant design pops out of a situation of competition between various actors.
2.4. Value appropriation
21 2008). On the other hand however, social network sites can use consumers for value creation through the process of co-creation of value (Marandi et al, 2010). This is achieved when consumers are willing to upload content, create applications and external linkages which enhance the value of the social network site.
The fact that users are rarely willing to pay for online content makes it hard for social media to let consumers pay for their service directly (Teece, 2010). One solution to this is to create revenue by letting other companies advertise on the pages of the social network site. However, such advertisements in the form of banners and pop-ups have become ubiquitous annoyances for users (McCoy et al., 2007) and there is a reluctance among advertisers to allocate large budgets to advertising on social network sites (McCormick, 2006). A third way to make money, which is linked to the theory of indirect network effects, is to let other companies pay to provide complementary services to the social network site. In a more schematic way this is shown in table 2.
Who Potential revenue base Downsides
Social Network Site
User payments & donations
Consumers rarely willing to pay for content
Advertisements Creates potential annoyance to users Payments by / profit
sharing with application providers
Creates an entrance barrier for application providers
Application suppliers
User payments Consumers rarely pay for content Advertisements Creates potential annoyance to users Selling virtual goods Small revenues per sale, not all users are
willing to pay for such goods
Third party offers Creates annoyance and attracts questionable partners
Table 2: Potential revenue bases for both social network sites as their application suppliers.
To attract other companies to invest in creating applications, it has to be possible for these companies to create revenues through their applications on the social network site. So apart from the value appropriation by the social network site itself, it also has to consider how they can set up a system in which the suppliers of the applications can also appropriate value.
22 (Zeisser, 2010). In this case consumers pay money for assets they can use within their applications or gift to other people to do so, examples of these will be given in the case study chapter. A final, more questionable27, way to create revenues is to offer consumers rewards to subscribe to an offer from another company made within the game. In turn, these other companies pay a fee to the company that owns the application to be visible to its users.
This paragraph described how companies can appropriate value from the value they create for consumers. Together with the insights about battles for dominant design, value creation and open and closed business models these insights present a solid theoretical foundation for the case study section of this paper. The next paragraph will explain how these theories form the bases for the research model for the case study.
2.5. Conceptual framework
The relationships from the theoretical framework form the basis for analyzing the cases in this paper. Together they form the conceptual model for this paper which is drawn in figure 7.
23 Figure 7: Conceptual Framework
The openness of the platform refers to the organizational form the social network has with its application providers. It is expected that how more open the system is, the wider the availability of complementary products is (H1a). This availability of complementary products leads to indirect network effects (H1b). The objective quality refers to the pure technological features for the end user ranging from photo sharing to its download speed. A high objective quality can lead to a positive technological gap with a social network sites’ competitors, and thus stimulate value creation, whereas a low objective quality leads to a negative technological gap (H2). The value created by a social network site is depended on this technological hap (H3a) and further depending on the level of indirect (H3b), and direct network effects (H3c). 2 3a 1a 5b 3b 3c 1b 7c 7a 4 7b 6b 6a 5a Technological gap Installed base (at time T+1) Availability of complementary products Signaling Stimulation of perception and expectations Value creation Direct network effects Indirect network effects
Winning the battle for dominant design
24 The value created by a social network site is expected to affect the number of users that join such a site (H4) because they derive social value from it. In turn, the number of users of a site affects the potential for value appropriation by the social network site (H5a) because a greater number of active users will for attract more advertisers. The competitor with the most users in the industry will also be the winner of the battle for dominant design, so the number of users also affects this aspect (H5b).
25
3.
Research methodology
This paper is aimed at understanding the dynamics behind value creation and the outcome of the battle for dominant design within social network sites. As described in the first chapter of this paper this subject is most suitable for explorative research. Since a comparison will be made of the dynamics around the value creation of various social network sites, a comparative case study is most favorable research method. By using this method this paper aims to create an in-depth overview of this topic and to create a useful base for further research in this area.
Although there are more social network sites present around the world, this research is narrowed down to a few specific cases which are clearly distinctive in their openness to external actors. This creates the opportunity to gain an in-depth insight in how this affects value creation by external actors.
3.1. Data collection
Sites as Facebook, Myspace and Hyves themselves form a very rich source of information on how they work, how their applications work, and how external companies can develop applications and external linkages. However, it is important to evaluate such information in a critical way which is necessary to identify its relevance for an academic paper. Information such as press statements and other expressions from the application developers will also be used and evaluated in the same critical way.
There are a number of sites which offer more neutral data and information in this area. For example, Alexa.com offers in-depth page view and visitor numbers for these sites. Other sources for these aspects are market research agencies such as ComScore and Nielsen which offer market data on social network sites. In addition information from other sites and news sources will be used. However all of these will be critically evaluated and where possible data triangulation will be used to secure the quality of this paper.
26 these relationships form the core of the issues which are relevant to the case study. When investigating their role within the various cases three issues are important; how they will be measured, how reliable this measurement is and how valid this is. The next part of this paper will first describe how the various aspects can be measured.
3.2. Measurement
Number of users
The most direct way in which this factor can be measured is the number of users a site has. However there are multiple concerns with this approach. First of all, the sites themselves will more likely to overestimate their reports on users numbers. Secondly, the sites do not offer comprehensive historical data on those numbers. Thirdly, not all members of a site might actually be active users and thus do not fit with the term ‘users’. Research has shown that many seldom used and abandoned accounts exist (Rejaie et al., 2010).
So in order to create a comprehensive image of the number of users of the various social network sites, additional measurements have to be made next to the user numbers from the social network sites themselves. There are other websites that offer data on how much traffic a website attracts. For example some sources, such as Alexa.com and ComScore, offer data such as the total number of monthly unique visitors a website attracts.
A way to define a user is as someone who visits a site on frequent bases, for instance at least once a month. Using this definition it is possible to use the number of monthly unique visitors as an indication of the number of users. Although visitors also include people who only visit the site and do not necessarily have an account this presents a pretty good indication of the number of users of the site since most people who visit a social network site will also login and use it. The number of daily unique visitors will be used as an additional measure of intensive users of the social network sites.
Installed base
27 greater number than the number of users. The measurement for this aspect is thus the total number of accounts, which data is provided by the social network sites. As this number presents more than people than only the active users it will also include latent users who do have an account on a social network site but who seldom use it.
Availability of complementary products
As this paper is focused on online services instead of tangible products the definition of complementary products does not directly apply to this paper. Teece (1986) describes multiple forms of both specialized as generic complementary assets which are needed to successfully market an innovation. The specific aspect that is meant here is complementary services. With regard to online services, complementary services occur when multiple services are offered through the same portal. Think for instance of a search engine which allows the user to search for a website trough both Google, Yahoo and Ask.com at the same time (Hwang & Oh, 2009). Such services are defined as being complementary because combined they offer more features or more content which can provide larger utility than the sum of utility from each separate service (Hwang & Oh, 2009).
The current academic literature lacks a classification of complementary services within social network sites. However, following their nature and the definitions of online complementary services mentioned here it is most logical to define them as the services that are provided through the portal of the social network site. These services mainly take form in applications that allow users to share photos, play games and other forms of interaction with each other.
The question when such applications became available can easily be answered relying on press statements and newspaper coverage. However, the exact number of available applications is hard to measure.
Openness of the platform
28
Signaling and perceptions & expectations
Some market research companies have sophisticated techniques to measure the amount of ‘Buzz’ a company creates on the internet28
, however such methods are not freely accessible. Another way to indicate how much signals of a website are being picked up by others is to take a look at their coverage within the media. The LexisNexis database offers a way to investigate this by looking up how much a specific name or word, such as Facebook or Hyves, is mentioned in selected newspapers during a specific time period. This method will be used in this paper to get an indication of how many signals the various social network sites send into the world. Both an analysis of the attention of Facebook and Myspace in international newspapers as a specific analysis of the attention of Hyves and Facebook in Dutch newspapers will be made.
Assuming that perceptions and expectations affect people’s actions it is possible to get an indication of these perceptions and expectations about the various websites have changed over time. Google Insights for Search29 offers the function to look up how much a specific word or company name is searched for on a weekly basis from 2004 to now. Using people’s search behavior as an indication of their perceptions and expectations this source can provide an insight into this aspect.
Objective quality
Since social network sites are a relatively new research area, there are no clear definitions of indicators of their quality. So to measure this aspect, context-specific indicators defined by the author will be used to explore this subject. These indicators will evolve around the most used functions by consumers, since these offer the best quality of a social network site to the user. One of these is the download speed of a site which is one of the means of its objective quality (Huizingh, 2000).
Technological gap
The technological gap relates to the difference of technological functionality between two products or services. A product or service has a technological advantage when its technological capabilities are perceived to be better than its competitors. However both actual
28
29 and perceived technological capability is hard to measure and to assess over an extended period of time for an online service. This paper will mainly focus on the technological functionality of the social network sites from a consumer’s perspective. For each social network site a number of key functionalities for its users and the time at which they were made available will be examined. This creates the opportunity to create an overview of how the functionality of these sites varied over time and how this affected their user numbers.
Direct & indirect network effects
Direct network effects occur when network externalities are in place and users attract other users into using a product or service. Although previous research showed that it is not an exact measurement of this aspect (Tellis et al., 2009) we can say that direct network effects exist when more people join a site and it becomes more valuable. Therefore this issue is not directly measurable but is merely being indicated by the increase or decrease in value that result from this effect.
Value creation
From a business point of view, value is frequently defined as the financial or economic value a service or a product has for a company or user. This last perspective will also be taken within this paper, thus the value a social network site has in the eyes of its users. As people click through many websites per day, a ‘valuable website’ for the user is a website where he spends more time on. Earlier research found that the time on site for a commercial website is a good indicator of its revenues (Lin, et al., 2010). So the average time per day per unique user spent on a specific social network will be taken as an indicator for its value.
Another indication of the value created by a company which this study looks at is its share price. The price that is paid for a company’s shares reflects its total market value and can thus also serve as an indication of the total value created by a company.
Winning the battle for dominant design
31
4. Cases
The first paragraph of this chapter will be about Facebook and Myspace. It will mainly contain data on a worldwide scale and does not include numbers about Hyves since this is only a local actor in the Netherlands. The second paragraph will focus on the data within the Netherlands and compare data on Hyves and Facebook in the Netherlands largely excluding Myspace since it has never attracted significant user numbers in this country.
4.1. Facebook vs. Myspace
In July 2003, Myspace was founded in July 2003 and was first made available to employees of eUniverse (the company of its founder Brad Greenspan). A key advantage within its early stage is that within eUniverse they had all resources they needed to run the site, from server capacity and bandwidth to human resources and some financial funding30,31. The site initially attracted users through the existing subscribers of eUniverse. After this, it quickly grew due to users attracting other people into using the site. Its potential was recognized by Rupert Murdoch whose News Corp company bought eUniverse and Myspace for $580 million in 200532,33.
The total user numbers of Myspace quickly from zero in 2003 to 4,9 million in 2004 and 26,7 million in 200534. At this time it was by far the most popular social network site. However in the years following 2005 Myspace was quickly overtaken by Facebook and currently shows declining numbers in various important indicators, such as its overall reach, number of page views and the time users spent on the site35.
Facebook was founded in 200436,37 by three Harvard students who wanted to create a sort of an online yearbook for their school. The site expanded rapidly when it first opened up to other
32 universities and later in 2005 also to U.S. high schools38. Albeit starting as an invite-only site, Facebook attracted a lot of users and ultimately overtook MySpace as the largest social network site in April 200839.
Today Facebook is the largest and most used social network sites on the planet, there are only a few countries left with another social network site with a larger market share40. Interestingly, its daily reach is still growing and users are estimated to spend an average of more than 30 minutes per day on the website41.
Myspace is a daughter of News Corporation42, a public-traded company with Rupert Murdoch as its chairman and CEO43. In its early years Myspace primarily offered its users profile pages which they were able to customize using HTML-code. Next to this it has always had a strong connection to music. One of its key characteristics was that it led upcoming artists not only make a profile but also upload and share their songs.
Facebook on the other hand is still a privately held company which shares are mostly held by its founders, employees and some private investors. It is being expected that the company will hold an initial public offering around 201244,45,46. However there is increasing pressure to go public due to a ruling of the U.S. Securities and Exchange Commision (SEC)47 that private companies may not have more than 499 individual shareholders48,49,50,51.
A key characteristic of Facebook is that the level of customization users can apply to appearance of their profile and other pages is significantly lower than on other social network
33 sites. For instance, on Myspace users can use HTML-code to alter their profile whereas on Facebook profiles are quite static in their appearance. However Facebook does set itself apart from other sites by the way users can post content. Users can easily post photos, links to other sites and even create applications.
Technological capabilities and objective quality
As formulated in the previous chapter, the technological performance of the social network sites will be explored from the consumer’s perspective. The first move Facebook made from creating more functions other than just static profiles was its programming of a photo album function on which users could upload their photos and tag friends in their pictures. This feature was already launched in October 2005 when Facebook was only available in universities and high schools52,53. A year later the site made a big make-over to people’s home pages by adding a news feed which showed each person’s friends status updates and new photos54. Another major improvement Facebook made was its mobile platform which opened in October 200755. Finally in April 2008, Facebook launched an onsite chat / instant messaging service56,57,58. Although some other functions have come gone, these functions are the ones that are mostly being used by consumers. Currently Facebook average page load time of 2.01 seconds is begin ranked by Alexa as rather slow with 64% of other websites loading faster59.
Currently, Myspace offers very similar capabilities to Facebook. However the moment at which these functions were introduced on its network was very different. Myspace’s photo albums were launched in March 200760,61. Shortly after that in December 2007 it also launched a news feed to people’s home pages similar to the one of Facebook62,63
. In 2008 it continued improving its service by adding a developer platform in February64 and a mobile
34 platform (which piloted from September 200765,66) in March67. Although it had a chat service, which operated apart from its network, it launched an on-site instant messaging service in May 200968. Finally, Myspace’s average load time 3.15 seconds is ranked by Alexa as very slow with 83% of other websites loading faster69. Table 3 presents an overview of the different moments of implementation of these functionalities and the performance of the social network sites.
Functionality Myspace Facebook
Photo albums and tagging
March 2007 October 2005
Status/news feeds December 2007 September 2006
Chat May 2009 April 2008
Developer platform February 2008 May 2007
Mobile platform March 2008 October 2007
Avg. page load time (seconds)
3.15 2.01
Table 3: Technological functionalities of Myspace and Facebook.
Openness of the platform and availability of complementary products
Applications on Facebook can be created by virtually anyone with the slightest bit of programming skills. Already in 2006, Facebook began creating a developer platform which went live in May 2007 with a number of 85 applications functioning on the site70,71,72. Since then the number of applications has exponentially grown to around half a million by 201073.
The openness for anyone to create an application on Facebook seems to explain a huge supply of applications. There is no selection mechanism that allows or denies access to certain applications. Additionally, Facebook does not charge money for creating an application or for instance the number of users of an application. In essence, both creating and using
35 applications in open to anyone at no cost. Given the number of applications available, it seems plausible to conclude that this openness has led to a wide availability of complementary services.
Similar to Facebook, Myspace has a developer platform which enables developers to design applications which can run within the platform of Myspace74. However this platform was launched in 200875,76, two years after Facebook launched theirs. The total number of Myspace applications is not clear but is expected to be much lower than that of Facebook.
Installed base
After a start-up phase in which membership was limited to students of specific universities and schools Facebook’s user numbers have grown tremendously after it opened up the possibility for anyone to join the site. According to the data it published on its own website its total number of accounts has developed as shown in figure 8.
Figure 8: Number of accounts on Facebook & Myspace
(Created by author, sources: Facebook.com & Myspace.com77)
This figure presents the persistent growth of the number of accounts for Facebook. Since 2007 Myspace has stopped giving information on its membership numbers. Currently it only mentions that it has ‘more than 100 million users worldwide’78
36 growth of its installed base up until March 2007. Since then it seems to have pretty much been stabilized.
Number of users
Where the installed base is indicated by the number of accounts, the number of active users is more interesting in the light of network effects. Figure 9 shows the number of monthly unique visitors from April 2007 till October 2010 for both Facebook and MySpace. It indicates that Facebook’s active user numbers increased from 40 million in April 2007 to around 250 million by the end of 2009 and 630 million by October 2010.
Figure 9: Facebook and Myspace monthly unique visitors 2007-2010 (Created by author, for source data see appendix V)
Interestingly, Facebook’s monthly unique visitor numbers have developed almost perfectly among an exponential pattern and is slowing down a bit in the past months, presumably ending in an s-curve as the market becomes saturated. Figure 10 presents data on this aspect.
37 Figure 10: Exponential curve analysis of Facebook’s monthly unique visitors
Myspace’s user number quickly grew in its first years but began to stabilize around 2007. Their user numbers have been stable around 100 million users over this time period. Based on the same data by ComScore it is reported that this number has dropped to 95 million by September 201079. However, these are monthly users, data on the number of daily users (the ones who use the site very actively) might present a less optimistic picture.
Unfortunately data on the early growth stage of Myspace is not available. Thus, statistical analysis can only be applied at the period from 2007 to 2010 at which its user numbers had already been stabilized. For this period is mainly of interest whether the number of users in a given period affects the number of users in the following period. To analyze this issue an analysis of the auto-correlation of a data set can be made. The autocorrelation indicates that a variable is correlated with itself.
38 Figure 11 represents the analysis of the autocorrelations of the number of monthly unique visitors of Myspace from the dataset in Appendix V. Within autocorrelation analysis any bar crossing the confidence limit in the chart is perceived as being significant at the 95% confidence interval (Gupta, 1999).
Figure 11: Autocorrelation of the monthly unique visitors function of Myspace (Created by author. Analysis of data in appendix V)
Autocorrelations
Series:Monthly unique visitors
Lag Autocorrelation Std. Errora
Box-Ljung Statistic
Value df Sig.b
1 ,679 ,147 21,213 1 ,000
2 ,526 ,146 34,288 2 ,000
3 ,585 ,144 50,857 3 ,000
a. The underlying process assumed is independence (white noise).
b. Based on the asymptotic chi-square approximation.
39 The results in figure 11 and table 4 indicate a significant autocorrelation for a lag of up to three periods. For a lag of 4 periods the correlation was not significant anymore. Thus at a specific point in time the site’s number of monthly unique visitors is highly depended on the number of monthly unique visitors in the preceding three months.
Value created
Facebook reaches almost 40% of all internet users on a daily bases80. Also, as indicated in figure 12 the average time people spent on the site increased from around 22 minutes a day at the end of 2008 to around 32 minutes at the end of 2010. So its value to users is very high and has stabilized over the past year.
Figure 12: Time on Site of Facebook & Myspace (Source: alexa.com81,82)
Although Facebook is not a publicly traded company it is also likely that its share value has increased remarkably over the past years. Using the investments in shares of Facebook over the past few years as an indication of its value, an estimation of its total value can be made. Appendix II gives a table of transactions of Facebook shares and their implicated valuation of the company. It makes clear that Facebook has grown from a 5 to 10 million dollar company into a company worth tens of billions of dollars today. A real valuation of its shares can only be made when the company goes IPO but at the end of 2010 the second-market for Facebook shares indicated a total value of more than 50 Billion dollar83,84.
40
Value creation & user numbers
Both Facebook social value as share value have grown exponentially over the past years. The theoretical framework of this paper suggests a relationship between a social network site’s value and user numbers. Table 5 presents the analysis of this relationship. The data shows a significant correlation of .889 (p<.01)85.
Table 5: Monthly unique visitors & Time on site for Facebook (Sources; alexa.com & appendix V)
Figure 13: Time on site and monthly unique visitors of Facebook
85
The Pearson Correlation is obtained by dividing the covariance of two sets of variables by the product of their standard deviations. A Pearson value of 1 indicates a perfectly positive correlation, a value of -1 a perfectly negative and a value of 0 indicates an absence of correlation. The significance of this value is depending on the sample size of the sets of variables.
0 100000 200000 300000 400000 500000 600000 700000 Q1 2009 Q2 2009 Q3 2009 Q4 2009 Q1 2010 Q2 2010 Q3 2010 N um be r of m on thl y un iqu e v is tors ( x 1 0 0 0 ) 0 5 10 15 20 25 30 35 Ti m e on S it e
Number of monthly unique visitors
Time on site
Time period Number of monthly unique
41 When we look at the web traffic statistics of Myspace we can see that the time spent on its site per visitor has dropped significantly over the past years. According to the data from Alexa.com highlighted in figure 12 above the average daily time on site per user was more than 20 minutes at the beginning of 2009 and dropped to 6 minutes by the end of 2010; indicating a decline of 70% in one and a half year.
Myspace share value is harder to estimate since it has been a daughter of News Corp for a while now. In 2005 News Corp paid $580 million to buy Myspace86. Over the past two years it has regularly been reported that its value since then has declined heavily and its current value is much lower87,88,89. Additionally, it is being reported that the company is currently in the process of laying off a large part of its personnel90,91 while Facebook is currently rumored to hire a larger office building to house its growing Silicon Valley staff numbers92.
42 Time period Number of monthly
unique visitors Time on site Q1 2009 124000 22 Q2 2009 110204* 19 Q3 2009 110204* 18 Q4 2009 101887 16 Q1 2010 109972 14 Q2 2010 109327 12 Q3 2010 94658 8
Table 6: Monthly unique visitors and time on site for Myspace (numbers marked with a * are estimated based on linear estimation in SPSS)
Figure 14: Monthly unique visitors and time on site of Myspace
Signaling and perceptions & expectations
43 When signaling would have an effect on the number of users of Facebook, then the rise in user numbers should be preceded by a rise in media attention. In this definition the amount of attention is an indicator for the amount of signals a website gets in these media. Put in one graph the development of the amount of users (data from table 5) and the amount of media attention Facebook received throughout the years (see appendix IV) the picture in figure 15 comes out.
Figure 15: Media attention, search statistics and unique visitors of Facebook (Created by author, based on LexisNexis, Google Search Analytics and ComScore)
Figure 15 highlights that Facebook’s user numbers and media attention have both developed along an exponential growth pattern. The peak in news coverage in Q3 2008 can be explained by the fact that news coverage peaked because Facebook overtook Myspace in worldwide web traffic the months before. According to this graph, the number of newspaper articles on Facebook has shown a slightly higher growth rate than Facebook’s user base growth during the year 2007. However, overall the statistics show a very similar pattern. The results indicate almost perfect correlations for all three relationships, indicating a very strong relationship at the 0.01 confidence level between both the variables signaling and perceptions as with the number of users of Facebook.
0 100000 200000 300000 400000 500000 600000 700000 0 20 40 60 80 100 120 Q 1 2 0 0 5 Q 3 2 0 0 5 Q 1 2 0 0 6 Q 3 2 0 0 6 Q 1 2 0 0 7 Q 3 2 0 0 7 Q 1 2 0 0 8 Q 3 2 0 0 8 Q 1 2 0 0 9 Q 3 2 0 0 9 Q 1 2 0 1 0 Q 3 2 0 1 0 M o n tly u n iqu e v isi to rs (x10 00 ) Ind ex ( fo r se ar ch & n ew s cov er age)
News coverage index
Search index
44 0 20000 40000 60000 80000 100000 120000 140000 0 20 40 60 80 100 120 Q 1 2 0 0 5 Q 3 2 0 0 5 Q 1 2 0 0 6 Q 3 2 0 0 6 Q 1 2 0 0 7 Q 3 2 0 0 7 Q 1 2 0 0 8 Q 3 2 0 0 8 Q 1 2 0 0 9 Q 3 2 0 0 9 Q 1 2 0 1 0 Q 3 2 0 1 0 M ont hl y uni que v isitors ( x1 00 0) Index ( S ea rc h & N ew s cov er age )
News coverage index
Search index
Montly unique visitors (x1000, quarterly median)
Monthly unique
visitors Search Index News_Coverage Pearson Correlation ,980** ,983**
Sig. (2-tailed) ,000 ,000
N 14 23
Monthly unique visitors Pearson Correlation ,997**
Sig. (2-tailed) ,000
N 14
**. Correlation is significant at the 0.01 level (2-tailed).
Table 7: Correlation of usage indicators of Facebook
For Facebook both its search index, newspaper coverage as user statistics developed by a very similar pattern over the past few years. The same kind of data on Myspace can also be found in Appendixes IV to VI. Picture 16 presents a short overview of this data and reveals that similar to Facebook the news coverage and search statistics developed among a similar pattern. However, it is also visible that the decline in user traffic followed the decline in news coverage and search statistics quite a few months later.
Figure 16: News coverage, search index and unique visitors of Myspace (Created by author, based on data in appendix IV,V,VI)
45 show a correlation of .974(p<0.01). Also, looking at figure 16, it is visible that some time after these factors have began to decline, the monthly unique visitors also began to decline (due to unavailable data this relationship has not been quantitatively assessed) . This suggests a lagged impact of these factors on the actual use of the social network site when in a declining stage. One of the reasons why there is a lag in this relationship is the autocorrelation of the user numbers mentioned before which showed that the number of users at a specific time is highly depended on the number of users in the previous periods.
Value appropriation
The main way in which Facebook creates revenue is by selling advertisements on its site. Already in 2006 it signed a partnership with Microsoft which provided it with banner advertisements and sponsored links93. Furthermore it improved its attraction to advertiser by making it possible to target specific audiences for their advertisements94. Where advertisements on for instance Google are aimed at search profiles of users, advertisements on Facebook can be aimed at very specific audiences, due to the personal information users upload on Facebook95.
Facebook’s attempts to create revenue from personalized advertisements have also been subject to debate over the past years. In November 2007 it launched a feature called Beacon which shared user data with outside companies96. Already a month after its launched Facebook admitted that it did a bad job with its release97 and its advertisement partners begin to turn down the service98. Eventually Facebook even put up a $9.5million settlement fund to compensate people who used the feature99, 100.
Since Facebook is still a privately held company it does not publish any financial revenue statements. However estimates of its revenue in 2008 vary from $230 million101 to $300
46 million102. For 2009 this increased to estimates of $800 million103 and for 2010 rumors are projecting revenues of $1 billion104 up to $2 billion105,106 ,107.
Similar to Facebook, Myspace sells ads to other companies as its major source of revenue. Unfortunately News Corp does not publish separate financial statements on Myspace, so its revenues and possible profits can only be estimated. Back in 2005 its revenue was estimated to be somewhere between $24108 million to $30 million109. This increased heavily to estimated revenues of $200 million in 2006110. This further increased to a figure of about $500 million in 2007111. However, similar to its user numbers Myspace’s revenues are also believed to have fallen since 2008. In that year its revenues were at its highest with an estimated value of about $582 million which dropped 15% to $495 million in 2009112. In that year it was forced to close many of its recently opened offices and lay off a large number of its staff because of its falling revenues113,114,115.
So although Myspace was valued on $580 million at its acquisition by News Corp in 2005 it has seen a steady decline in its revenues since 2008. Since these revenues do not say anything about its actual profitability it is hard to guess whether Myspace has really been able to appropriate enough value from its user base.
Battle for dominant design
The previous pages already showed that Facebook has overtaken Myspace as the largest worldwide social network site. However, the emergence of a dominant design is only defined
47 as a design’s competitors accept its dominance. An indication of such an acceptance of the domination of Facebook by Myspace occurred in November 2010 when Myspace opened up its site to users with only a Facebook profile116,117. A move which was explained by industry experts as a recognition of Facebook’s market dominance by Myspace118
.
4.2. Facebook vs. Hyves within the Netherlands
Facebook is by far the most dominant social network site in the world. However, there are a few countries left in which other social network sites have a larger national market share. These include China, in which Facebook is banned119 and Brasil, in which Orkut from Google is more dominant120. In the western world there is only one country left in which Facebook is not the dominant player, which is the Netherlands where Hyves has a larger market share.121
Hyves was launched in 2004 and is positioned as a social network site for the Dutch market. It was launched by a few internet-entrepreneurs who had previously launched other successful sites such as iex.nl122. The site quickly grew and became one of the most popular websites in the Netherlands by 2007123. In 2008 it received an investment by Joop van den Ende which was estimated to be worth tens of millions of Euros124,125. By 2009 it was still the most popular social network site in the Netherlands and it did not believe to have serious competition left126.
However, a bit similar to Myspace, some of its key user statistics are showing a decline of its use over the past two years, see the graphs in figure 17. Currently it has around 10 million accounts127 of which an estimated 93% originate from the Netherlands128. Within the
48 Netherlands it is still the sixth most visited website, however, according to some sources Facebook has passed it and is currently the fourth129.
Figure 17: Hyves’ daily reach and Time on Site statistics (source: Alexa.com).
Technological capabilities and objective quality
Currently Hyves has very similar functionalities as Facebook and Myspace. Ever since its launch it has constantly added new features to its site, among the first were its news feed (called Wiewatwaar, Dutch for Whowhatwhere) and photo albums, which both were already launched back in 2005130. Hyves was also relatively quick with adding a mobile platform in February 2006131. Later that year it also launched an on-site instant messaging function132, which was one and a half year earlier than Facebook did. In 2008 it also launched a developer platform which enabled other parties to create applications for its network133. Finally an aspect at which Hyves performs significantly better than its competitors is its load time. According to Alexa, this is fast with only 1.25 seconds, a rate which is faster than 61% of other sites134. This information is compared with Facebook in table 8.
128 http://www.alexa.com/siteinfo/hyves.nl
129 http://www.alexa.com/topsites/countries/NL 130
‘Vriendennetwerk op de Smartphone’, De Telegraaf, 22 april 2005.
131 ‘Hyves wordt mobiel’, De Telegraaf, February 24, 2006. 132 http://www.emerce.nl/nieuws.jsp?id=1701962
133
