The role of opinion leaders in the adoption and diffusion process of Web 2.0 concepts

(1)

The role of opinion leaders in the

adoption and diffusion process of Web

2.0 concepts

A look at the Facebook case

Sjoerd Gerbrand Huitema S.G.Huitema@student.rug.nl

MSc BA Master Thesis Small Business & Entrepreneurship

September 2011

Supervisors dr. C.K. Streb ir. H. Zhou

University of Groningen Faculty of Economics and Business

(2)

2

SUMMARY

The new generation of internet concepts, better known as Web 2.0 concepts, which arise in the presence are significantly different from the early Web 1.0 concepts. Web 2.0 is known for their strong network effects. For Web 2.0 companies this implies that attracting as many users as fast as possible is critical for creating value for the same users; this results in a fierce competition for users in the sector. The aim of this study is to provide more insight into the role opinion leaders play in this process of attracting users by researching the adoption and diffusion process of Web 2.0 concepts. Therefore, the research question in this study is: What is the role of the opinion leader in the adaption and diffusion process of Web 2.0 concepts? And how does this influence the adaption behavior of the targeted mass?

A literature analysis identified several roles an opinion leader can play in the adoption and diffusion process. Based this analysis a conceptual model showed up on how these roles can affect the adoption behavior of the targeted mass. The model assumes that the media and the opinion leader directly influence the adoption behavior of the targeted mass. Another factor which influences this, is the product positioning. The meaning of this factor is that the Web 2.0 concept should be positioned in such way that it adds for both opinion leader and targeted mass (or non-leader) a significant value. It is however expected that this value proposition can be different for the two groups. To validate the model, opinion leaders and non leaders were interviewed on the subject of adoption of Web 2.0 concepts. A specific focus in these interviews was the adoption of Facebook. Facebook is according to the literature a good example of Web 2.0 concept and it is also widely adopted and diffused. The opinion leaders were selected students from the University of Groningen. Out of a database of about 200 students who showed opinion leader behavior, a selection was made after observing their Facebook behavior. The 6 selected students which were willing to cooperate were double checked for being an opinion leader by using the designation approach to identify opinion leaders. For this self-designation approach a sample group of 100 students from the University of Groningen was used. It was confirmed that all the 6 selected students were indeed opinion leaders for Web 2.0 concepts. The selection of non-leaders was done by choosing randomly 10 students from the University of Groningen. Also for this group the self-designation approach was used to make sure the leaders are non-leaders with respect to Web 2.0 concepts. 8 of 10 students were indeed non-non-leaders for Web 2.0 concepts. The interviewdata of the other 2 students was not used. Despite the small group of interviewees, the interviewdata showed congruent answers for both groups.

(3)

3

INTRODUCTION

Why and how certain concepts come to be adopted widely, while other, equally plausible, alternatives languish, is a question that has troubled researchers for some time (Munir & Philips, 2005). Although practitioners can provide a wide array of anecdotes, systematic research on this issue has been rather sparse (Cavusoglu, Hu, Li & Ma , 2010). Answering this question from the perspective of Web 2.0 start-ups is interesting and important since it are network products, these concepts only have value for users if the concept is adopted widely (Shuen, 2008). User benefits will be realized only when other people begin to adopt the concept too (Sundsøy, Bjelland, Canright, Engø-Monsen & Ling, 2010). Because of such a lack of user benefits at the early stage, under-adoption is natural for network products and services (Rohlfs, 2001). Meaning that those concepts will fail when they are unable to attract sufficient users to create value (Lee & O’Connor, 2003). An example of a typical Web 2.0 concept that coped with this issue extremely well is Facebook, this social network site create user benefits only because so many people are using it; Facebook without users does not make sense; you cannot search for friends and contact them, meaning that it has no value for the user because stay in touch with your friends is the aim of the site (Shuen, 2008).

The term Web 2.0 refers to the next generation of internet applications which are more social than previous (Web 1.0) applications and whereby the user is not only the consumer but also the contributor of data (Kim, Yue, Hall, Perkins & Gates, 2009). Web 2.0 makes the web more than a collection of stand-alone websites; it makes the web an interactive platform for online applications (Constantinides & Fountain, 2008). Due to the rise of the internet and the ongoing digitization of information many business opportunities arise in this sector (Laursen, 2004). The start-ups which arise in the Web 2.0 sector are different from typical bricks and mortar start-ups: Web 2.0 start-ups are known for their very short time to market, the low sunk costs (“it’s just a computer code”) (Afuah & Tucci 2000, Stam 2008) and the presence of strong network externality effects (Shuen, 2008). Due to these characteristics the barriers to entry the market are low, but achieving success is hard since a lot of entrepreneurs start a Web 2.0 company. As a result there is a fierce competition among those firms for users (Shuen, 2008). This competition is even fiercer because Web 2.0 concepts are network products – the value for the user increases as more people are using it – which makes the user more valuable compared to non-network products. Lee & O'Connor (2003) found that for these network products the size and development speed of the installed base plays a critical role in a product's long-term performance and that building up network effects at the early stage is the key to the success of product diffusion.

(6)

6

enjoying general academic acceptance (Constantinides et al., 2008). So far the academic research on Web 2.0 is mainly focusing on defining Web 2.0 (Kim et al., 2009) and how it can be applied in business for marketing and other purposes (Constantinides et al., 2008, Greenhow, Robelia & Hughes, 2009).

The traditionally literature on the adoption and diffusion of innovations has examined innovations that diffused fully, meaning that every potential adopter adopted them (Rogers, 1995). According to Rogers (1995) the focus has not been on explaining why these innovations diffused fully, but rather on explaining the rate at which these innovations diffused or the order in which they were adopted. Attewell (1992) classified the diffusion studies into two groups, the first group correspondents with the traditional view and is called the macro diffusion-view. This group deals with mathematical characterization of the rate and the pattern of adoption of an innovation among potential adopters. The second group, the so-called adopter studies, focuses on the identification of factors that facilitate or hinder the adoption of the innovation. This group corresponds with the more recent focus of researchers why innovations diffuse (Wejnert, 2002). Within these studies the concept of the ‘Opinion Leader’ is playing a key role (Eck, Jager & Leeflang, 2011). These opinion leaders tend to influence others to change their attitudes and behaviors and are therefore capable to start the diffusion of a product. According to Roch (2005) the idea that a small group of influential opinion leaders may accelerate or block the adoption of a product is widely accepted. However, recent work of Watts & Dodds (2007) states that there is only a limited effect of opinion leaders on the diffusion process. We still expect however that those opinion leaders can play a crucial role for the diffusion of typical Web 2.0 products (network products with network externalities); they can start the diffusion and influence other people to adapt the Web 2.0 concept (Shuen, 2008, Eck et al., 2011, Gladwell, 2002). In order to get a more clear perspective on the role of the opinion leader within the diffusion and adaption of Web 2.0 concepts the following research question will be answered in this thesis:

What is the role of the opinion leader in the adaption and diffusion process of Web 2.0 concepts? And how does this influence the adaption behavior of the targeted mass?

To answer this question properly, subquestions will be used. These are the following:

Do opinion leaders directly influence the targeted mass in their adoption behavior? If yes, how does this influence look like?

Is the targeted mass directly influenced by the media? Or happens this via the opinion leader? Is focusing solely on the opinion leader, with respect to the positioning of the product, sufficient for a proper diffusion? Or should also be focused on the targeted mass?

(7)

7

(8)

8

LITERATURE REVIEW

Defining Web 2.0

To understand the nature of Web 2.0 concepts and underline the major importance of the user for Web 2.0 concepts, this chapter starts by defining Web 2.0. Web 2.0 involves web development and design to facilitate interactive sharing, user-based design, and collaboration on the internet. Web 2.0 companies allow users to do much more than just retrieve information on the platform (Lee, Kim, Noh & Lee, 2010). Time and the continual growth of web technologies, new web applications/services, and ideas have determined that there is a significant difference between the key features of the Web today, compared to the key features of the Web several years ago. Web 2.0 is a massive topic with a large number of elements that interact with it (Kim et al., 2009).

Differences with web 1.0

Unlike users of Web 1.0 platforms, Web 2.0 allows users to own and exercise control over the data in platforms. According to Kim et al. (2009) Web 2.0 companies may have architecture of participation that encourages users to add value to the application as they use it. They summarized the most striking differences between Web 1.0 and Web 2.0 in a table which is shown below. Their work is based on studies of Musser & O'Reilly (2006), Murugesan (2007) and O'Reilly (2005)

Dimension Web 1.0 Web 2.0

Business/ Application

Ad-revenue based model is one of many Web business models.

Relevance, ubiquity and the indirect benefits of network externality ensures Ad-revenue based model to become increasingly important. (e.g., Google AdSense, AdWords, Facebook Ads) Publishing using top down

approach (e.g., Britannica Online)

Participation from bottom up approach (e.g., Wikipedia, blogging, MySpace, YouTube, Flickr, Facebook)

Advertiser initiated advertisements (i.e., one-way, read-only; e.g., DoubleClick)

User oriented and enhanced advertisements (i.e., two-way & dynamic); relevant, easy to share, access, and consume: user experience, information, and knowledge [e.g., Google AdSense, Facebook Ads)

Social/user Small crowd interactions and networks

Massively connected social interactions, social networks, social computing, community empowerment (e.g., MySpace, Facebook)

Limited collaboration Collaboration; easy to participate, create, and update content (e.g., Wikipedia, blogging, MySpace, Facebook)

(9)

9

Difficulties with defining web 2.0

Despite of these differences mentioned in the table the difference between Web 2.0 and Web 1.0 are not crystal clear (O'Reilly, 2005); defining Web 2.0 is difficult for several reasons. First, there are no brand new and revolutionary technologies that make Web 2.0 applications available. AJAX is considered a major technology for Web 2.0 applications such as Google Maps, Facebook, and Flickr. However, it is important to note that AJAX is not a new technique. Because of this reason, the boundary between Web 2.0 and the previous Web 1.0 is blurred, which makes it difficult to define Web 2.0 clearly. (O’Reilly, 2007, Kim et al. 2009) Second, there is a varying and diffused understanding of Web 2.0. Many people claim Web 2.0 is a buzzword, while others say it is a tangible structure determining how applications and services are to interact. Clearly, Web 2.0 is not a tangible object that was marketed as a product, nor is it a structure that was developed in the planning room. (Kim et al. 2009) Another difficulty with defining Web 2.0 is that it is a massive topic with a large number of methodologies and components that interact with it. The number of Web 2.0 technologies, concepts, and applications involved is too complex to have a boundary for clearly understanding the concept of Web 2.0. (Musser & O'Reilly, 2006; Kim et al., 2009)

Despite the lack of a uniform definition for Web 2.0 there is consensus on the general characteristics of Web 2.0 concepts. According to Best (2006) Web 2.0 has the following unique characteristics: rich user experience, dynamic content and scalability. Other characteristics, such as openness, freedom, and collective intelligence, can be viewed as essential attributes of Web 2.0 (Greenmeier & Gaudin, 2007). However, the most vital factor for all categories of Web 2.0 applications is the user, not only as a consumer but mainly as a content contributor. (Constantinides et al., 2008). For this thesis the following definition of Web 2.0 will be used:

(10)

10

Adopter studies

Why some innovations diffuse and others don’t is the scope in the so-called adopter studies. These studies focus on the identification of factors that facilitate or hinder the adoption of a product or innovation. (Wejner, 2002) A specific scope within this field of study is the social network approach (Delre, 2007). This discipline assumes that innovations spread through social networks at a certain speed and that the speed and completeness of the diffusion depends on several factors, e.g. position and the ties of the actors and innovators in the social network.(Delre, 2007, Eck et al. 2011) This view will be held during this literature review since it is expected that, based on the strong network effects & network externalities of Web 2.0 concepts (Shuen, 2008), it will be the best approach to explain the diffusion of Web 2.0 concepts. Within this discipline the concept of the opinion leader is playing an important role. The concept of opinion leadership is discussed widely in the literature (Burt, 1999; Kelly, 1991). Katz and Lazarsfeld (1955) originally defined opinion leaders as “the individuals who were likely to influence other persons in their immediate environment,” during the decades who followed this definition remains more or less unchanged (Grewal, Mehta and Kardes 2000). The concept is part of the two-step flow model of influence; this theory suggests that a small minority of opinion leaders act as intermediaries between the mass media and the majority of society (Katz & Lazarsfeld, 1955). According to Weimann (1994), over 3,900 studies of influentials, opinion leaders, and personal influence were conducted after Katz and Lazarsfeld’s study. Burt (1999) stated that the concept of the opinion leader had become “a guiding theme for diffusion and marketing research.” Roch (2005) recently concluded that “In business and marketing, the idea that a small group of influential opinion leaders may accelerate or block the adoption of a product is central to a large number of studies.”

(11)

11

Opinion leaders

As said before, the opinion leader is playing a key role in the adopter studies. The opinion leader is an actor who is an active media user and who interprets the meaning of media messages for lower-end media users. The opinion leader is held in high esteem by those who accept his or her opinion since they are perceived as experts in a certain field. By being so they tend to influence others to change their attitudes and behaviors. Opinion leadership was originally conceptualized as a combination of personal and social factors. As Katz (1957) noted, being an opinion leader is related to three attributes: (1) "who one is"; (2) what one knows; and (3) "whom one knows".

‘Who one is’ refers the personification of certain values. Influence is often successfully transmitted because the influencee wants to be as much like the influential as possible. An example where ‘who one is’ counts very heavily is in the fashion business; young unmarried girls are fashion leaders since in this culture youth and youthfulness are supreme values. But also other personal characteristics play a role, Weimann et al. (1994) suggested that personal characteristics that seem to have relevance to the adoption of innovations are self-confidence and independence or “psychological strength” because they would likely modulate the extent to which an actor adopts an innovation without waiting for the security of knowing that others have so acted. Opinion leaders select the most important innovations from the abundance of information, rapidly adopt those innovations, and using their own social networks to create a public agenda that significantly promotes adoption. Conversely, psychologically weak actors depend on the opinions of stronger actors who relay media information. Furthermore, the perceived social status of an actor also decides if he or she is an opinion leader. Actors with high status usually adopt an innovation first and then impose adoption of the innovation on lower status actors. Herbig & Palumbo (1994) found that the high social position of an actor significantly modulates the probability of adoption within homogenous groups, such as when adoption of innovations by high-status firms generates adoption of technological innovations in similar firms.

‘What one knows’ focuses on the competence of the opinion leader. Opinion leaders are considered experts in their field, but this is an informal recognition by friends, relatives, colleagues and acquaintances. Most of them tend to be monomorphous: they are usually experts in one field but rarely in various fields. Since they are considered as experts in a field people in their network will accept and adopt their opinion.

(12)

12

individual actors. The first is network connectedness (e.g., closeness of communication between members). The second concerns characteristics of actors that influence openness to novel information. According to Rosero-Bixby & Casterline (1993) the most significant factor for predicting network connectedness is the network size: the smaller the network, the more connected it is. However, also important are frequency of interactions among members (Katz & Lazarsfeld, 1955) and openness of communication within a network, referring to the level of privacy of shared information (Rogers & Kinkaid, 1981). The adoption rate within interpersonal networks also appears to be modulated by variables that determine openness of an actor to novel information, such as prestige and authority (Burt, 1987), extent of an actor’s social connectedness with others as close friends, advisers or discussion partners, and the relative level of innovation-relevant knowledge (Valente et al., 1995). More about the structure of social networks was found by Granovetter (1973), he conceptualized the architecture of social networks. In Granovetter’s view, a social network consists of two essential elements: (1) cliquish sub-networks and (2) bridges. A cliquish sub-network consists of individuals who are interacting extensively with each other (e.g. family and friends network). According to Granovetter (1973) this sort of sub-network is not very helpful when people look for jobs. His main argument is that information traversing through cliquish subnetworks is more likely to be limited to a few cliques, which tend to share redundant ties. Instead, people get more useful job information from random contacts, or people who are not in extensive relationships. Such connections are called bridges, which serve to connect diverse members from different, or often socially distant, subnetworks. Since opinion leaders are active in different social networks they are good in forming bridges, which is essential for a product to diffuse completely (Delre, 2007).

Opinion leaders, Non-leaders and media

(13)

13

non-purposive. People do not feel they are being tricked into thinking a certain way about something from someone they know. However, the media can be seen as forcing a concept on the public and therefore less influential. While the media can act as a reinforcing agent, opinion leaders have a more changing or determining role in an individual’s opinion or action. Though, McQual (1979) states that the effect of media on the ‘masses’ cannot be answered in broad generalities. According to the author various types of effects impact various types of people at various levels of society, under various conditions.

Product/Market fit

Focusing solely on the people who have to adopt the product is not enough; the product on itself should be in line with what the people want (Yap & Souder, 1994). Gladwell (2002) identified factors which cause social epidemics; he researched the so-called tipping points. Gladwell (2002) called one the factors necessary to start a social epidemic the ‘stickiness factor’: it is not enough to let the right people tell about an action or product; it should also stick to the peoples mind. Moreover, taking it from the Web 2.0 perspective Cooper & Vlaskovits (2010) call this phenomenon the product/market fit; they define it as being in a good market with a product that can satisfy that market. According to Cooper et al. (2010) many start ups fail because they target users for which the concept solves an insignificant problem. If there is no product/market fit the concept will not diffuse (Gatignon & Robertson, 1985), it just does not stick to the peoples mind and the benefits of adapting the concept are too low compared to the investment.

The determination of the customer needs and product specifications are usually made at the development stage of the concept which is early in the project life cycle (Yap & Souder, 1994). However, in the Web 2.0 sector the development of new concepts are going that fast and it is known for a very high uncertainty (Shuen, 2008), thus making it difficult and for developers to track changing customer needs. The inability to track rapid changes occurring in the market while the product is being developed may cause a mismatch between the product’s benefits and the customer needs. This mismatch has been shown to lead to failure (Cooper & Kleinschmidt 1987). According to Aaker & McLoughlin (2007) the Segmentation, Targeting and Positioning approach (STP) can be used to identify customer needs and help to target the right group, this should not only be done at the early stage of development but also through the whole development process.

Segmentation can be described as the act of dividing the market into distinct and meaningful groups of customers by definable consumer characteristics to identify consumers whose buyer motivation and responses to product are the same (Freathy & O’Connell 2000). Segmentation of the market can be useful for resource constrained start-ups since it allows a start-up to use a particular marketing mix and target a smaller group with greater precision by doing so resources can be deployed more effectively and efficiently. (Aaker & McLoughlin, 2007) Next to that, dividing the market into smaller homogenous groups allows companies to provide products that better suit the needs of their customer (Bennion, 1987).

(14)

14

target marketing is to create ambassadors out of satisfied users. These ambassadors are the most common source for generating new users, when they are satisfied they spread the product by word-of-mouth. According to Gatignon & Robertson (1985) adoption is likely to be faster if the marketing strategy is compatible with the segmented targeted. Furthermore, Easingwood & Beard (1992) proved that clearly targeted products diffuse more rapidly than non-targeted products. The target markets are often defined in terms of the groups that adopt at different stages of the product life cycle (Beard & Easingwood, 1996).

Positioning is the act of placing a product within a market landscape (Cooper & Vlaskovits 2010) and it specifies how the product aspires to be perceived relative to its competitors and market (Aaker & McLoughlin, 2007). As said before, wrong positioning could lead to failure of the product. According to Yap & Souder (1994) successfully positioning a new product to its unique benefits and superior need-satisfying capabilities is very hard for small firms which operate in a market which develop at high speed and are known for their uncertainty. When the new product is positioned as an improvement compared to other similar products (of competitors) this was found to foster success (Wind, 1982). They argued that this strategy is used to convince customers that the new product was not radically different from those currently in use, by doing so it appealed to the customers that they are not required to evaluate the from a new frame of reference, thus saving them time and avoiding the risk in using a new product.

Summarizing, the product/market fit is not only important for the adaption of the concept because it targets the right potential users (opinion leaders and/or targeted mass). It also gives focus to the resource constrained start-up and helps to use their resources as efficient as possible.

Conceptual model

Figure 1: Conceptual model

(15)

15

Based on the literature review a conceptual model has been created, it represents the interaction of factors which influence adaption behavior of potential users of web 2.0 concepts in such way that the targeted mass can be reached.

Opinion leaders: The opinion leader is the central point in the model; they are the early adapters of the new concept. They are more then only passive adapters of the new concept; they are also active in promoting the concept in their social network, in this way they influence the media and targeted mass (Katz, 1957, Eck et al. 2011).

Proposition 1: Opinion Leaders influence media coverage by interacting with the media. Proposition 2: Opinion Leaders influences the adoption behavior of the Targeted Mass.

Media coverage: According to the two-step flow communication model concepts flow from the media to opinion leaders and from them to the targeted mass. However, based on the strong personality, their high status and high visibility of the opinion leader it is likely that the media and the opinion leaders will interact with each other; it is not a one way direction of influence (Weimann et. al, 1994). When the opinion leaders adopt a new Web 2.0 concept, it is likely that the media will cover this; “the people that matters are using it so it must be something good”. This gives rumor to the concept and increases the familiarity with the concept which facilitates the adoption behavior of the targeted mass (Wejnert, 2002). So, it is expected that the media directly influences the adoption behavior of the targeted mass; therefore the following is proposed:

Proposition 3: Media coverage influences Opinion Leaders adoption behavior.

Proposition 4: Media coverage directly influences the adoption behavior of the Targeted Mass.

Product positioning: Product positioning refers to the product/market fit whereby targeting and positioning of the concept in the market place key is. The web 2.0 startup should target the potential user for which the added value of their concept is the largest, those users are most likely to adopt. According to Cooper et al. (2010) many startups fail because they target users for which the concept solves an insignificant problem. The benefits of the concept should fit with the needs of the potential user (opinion leader and targeted mass), if there is no product/market fit the concept will not diffuse (Gatignon & Robertson, 1985). Since opinion leaders can have other needs than the targeted mass it is important to position the concept that it fit with both needs. It is expected that because of these potential different needs the role the opinion leaders plays in the adoption process of Web 2.0 concepts is different compared to non-network effect products. In order to clarify this, the following propositions will be researched:

(16)

16

Proposition 6: The Web 2.0 concept is positioned such that it adds significant value for the Targeted Mass (TM).

Targeted Mass: Potential users of the concept whereby the benefit of adapting the concept is large enough to adopt the concept; it has to solve a significant problem (Cooper et al., 2010). These users are non-opinion leaders (NLs) and tend to be influenced by the media and opinion leaders (Katz, 1957, Eck et al. 2011). The targeted mass is necessary for Web 2.0 concepts in order to have enough users to create value for them due to the fact that Web 2.0 concepts are network products (Shuen, 2008).

Increasing familiarity in social network: The adaption behavior of the targeted mass is facilitated by becoming more familiar with the concept. There are a number of factors which reduce the novelty and increase the familiarity with the innovation, among them is when the targeted mass obtains information from opinion leaders and media coverage. (Meyer et al.,1977; Weimann et al, 1994). Proposition 7: The increasing familiarity in the social network due to media coverage facilitates the adoption behavior of the Targeted Mass.

(17)

17

METHODOLOGY

This research design concerns the plan and structure of the study. The research design presents a blueprint for the collection, measurement and analysis of the data (Cooper & Schindler, 2003).

Research strategy

The goal of this research is to get a better insight in how the adoption and diffusion of Web 2.0 concepts goes and what role the opinion leader plays in this process. This is done by researching the adoption process and usage of Facebook by students. The Facebook case is used since it is considered as a good example of a Web 2.0 concept; it has all the essential elements which a Web 2.0 concept should have (Shuen, 2008), with this study a start can be made for researching the adoption behavior of Web 2.0 concepts. The newness and unique characteristics of the Web 2.0 sector are debit on the limited research in this specific field. In order to validate the hypotheses and answer the research questions an explorative study will be performed. In explorative studies researchers search for structures to discover future research tasks (Cooper and Schindler, 2008). They are particular useful when researchers lack a clear idea of the problems they will meet during the research process. Through exploratory studies researchers are able to develop concepts more clearly and help to establish priorities. An often used research strategy for such studies is the case study strategy; it focuses on understanding the dynamics present within single settings (Eisenhardt, 1989). According to Eisenhardt (1989) building theory from case study research is most appropriate in the early stages of research on a topic or to provide freshness in perspective to an already researched topic, this is because theory building does not rely on previous literature or prior empirical evidence. Case studies typically combine data collection methods such as archives, interviews, questionnaires and observations, this is called triangulation. The method of triangulation can be traced back to the study of Campbell and Fiske (1959), they developed the idea of "multiple operationism". They argued that more than one method should be used in the validation process to ensure that the variance reflected is of the trait and not of the method. Thus, the convergence or agreement between two methods enhances the belief that the results are valid and not that they are not a methodological artifact (Jick, 1979) and it increases theoretical generalization of the phenomenon as well (Flick, 2006).

(18)

18

Sample selection

The interview data will be collected by interviewing students from the University of Groningen. This group is chosen since this groups match with the initial target group of Facebook: college students (Shuen, 2008). Besides, the students range in age between 18 and 25 years, people of this age are significantly more often early adopters of Web 2.0 concepts than people of other ages (Shuen, 2008). Another, more practical, reason to focus on this group is that it is easy accessible for the researcher.

(19)

19

Opinion leader selection

The method which will be used to come up with suitable OL interview candidates is as follows:

Step 1: Opinion leaders will be selected out of a database of approximately 200 Dutch students, who are mentoring international students in Groningen, a selection will be made based on observation of their Facebook usage. This group of students is chosen since those people are showing opinion leader behavior: by taking the unpaid mentor job they involve in various social activities and become a central position for the mentoree in their social network (Weimann et al., 2007). Next, they are aware that the mentoree will view them as an expert and are an important source of information and thus influence them (Brosius & Weimann, 1996; Katz, 1957).

Step 2: Though this group shows opinion leader behavior, it is unknown if they are opinion leaders for Web 2.0 concepts. In order to check for this, their Facebook behavior will be observed. The key focus is here the content of their “status updates” and “wall posts” and if it indicates opinion leader behavior for Web 2.0 concepts. For the observation form see APPENDIX V: Opinion Leader Observation Form. Step 3: Based on the observed behavior 10 students will be selected for a semi-structured interview and asked to collaborate.

Step 4: In the first part of the interview the interviewee will be asked to fill in the self-designation (SD) questionnaire, which is needed for step 5: SD analysis. The second part of the interview will be related to gather data with respect to the research questions. Semi-structured interviews will be used for the data collection because of the fact that they are well suited for the exploration of the perceptions and opinions of respondents regarding new and sometimes complex issues and they enable probing for more information and clarification of answers (Barriball & White, 1994). This might be of major importance in order to validate the conceptual model in this stage of the study. The interview questions can be found in APPENDIX VI: Interview Questions. The interview questions aroused after following the method described by Emans (2004) on how to come up with interview questions which answer your research questions. This specific method used for this thesis can be found in APPENDIX I: Interview guide.

Step 5: The SD analysis is an extra check via the self-designating test to ensure that the interviewee is an opinion leader for Web 2.0 concepts. Due to the newness of the research subject and the limited

Step 1: Database of 200 Dutch students

Step 2: Observation for Web 2.0 OL behavior

Step 3: 10 students are selected and approached

Step 4: SD Questionnaire & interview Step 5: SD analysis Interviewee is OL: Data used Interviewee is not OL: Data

(20)

20

academic attention to this subject the importance of having good qualitative data becomes even more important. Therefore the extra check is performed to ensure the quality of the interviewees, which again affects the quality of the results and analysis. See self-designating test. If the SD analysis shows that the interviewee is not an OL, the interview data will not be used for this research.

Non-Leader selection

The interview for the NL will have the same structure as the OL interview, also for the NL 10 people will be interviewed. The selection method for the NL is different compared the selection for OLs, this is because of the different nature of NLs.

Step 1: Randomly 10 students who use Facebook are picked for an interview. This can be done since the majority of the Facebook users are NLs. These students will be randomly picked in the canteen from the university library. Interviewing the NLs gives more insight in the differences in motivations between OL and NL in adopting Web 2.0 concepts and it allows checking the statements of the OL. Step 2: This step is identical to step 4 in the selection procedure for OLs.

Step 3: The SD analysis will be used to check if the interviewee indeed is a NL. If it turns out that the interviewee is not a NL the interview will not be used for the qualitative analysis.

Self-designating test

In order to use the self-designation method well for the determination of OLs and NLs a sample group is needed to compare the outcomes of the SD questionnaire of the interviewees with. The procedure to come up with the sample group is as follows:

Step 1: 100 students between 18 – 25 years will be randomly picked in the canteen of the library of the university.

Step 2: The students are asked to fill in the SD questionnaire which consists out of 6 questions regarding being an opinion leader for Web 2.0 concepts (See APPENDIX IV: SD Questionnaire).

Step 3: The results of the SD questionnaires of all 100 students is the SD sample group. In line with Kings and Step 3: SD analysis

Step 1: 10 students randomly picked

Step 2: SD Questionnaire & interview Interviewee is NL: Data used Interviewee is not NL: Data not used

Step 3: SD Sample Group Step 1: 100 students

randomly picked

(21)

21

Summers (1970), the 29.4% who score highest on the scale are considered as OL. These 100 students are picked in order to have a significant sample group to compare the individual SD questionnaire scores of the interviewees with. More on the analysis can be found under Self-designation analysis.

Data analysis

Analyzing data is the heart of building theory from case studies, but it is the most difficult and challenging part. It is especially difficult since published studies in general only describe the research and data collection methods; the analysis of the data is only incidentally discussed (Eisenhardt, 1989).

Self-designation analysis

The first part of the data analysis will be focused on the self-designating test since this has an effect for the interview data which can be used for the qualitative part. The self designation analysis will consist out of analyzing the self-designating questionnaires. Every answer on the questionnaire accounts for a certain amount of points, based on the answers in the questionnaire the total ‘opinion leader’ score will be calculated. The top 29.4 % of the people of the sample group who score highest on the scale of opinion leader will be considered as opinion leaders, as in line with Kings and Summers (1974). Next, the score of the questionnaires of the interviewees will be compared with scores of the sample group. In order to belong the group of opinion leaders, the OL interviewee must have a minimal score that is at least as high as the score of the lowest opinion leader in the sample group. For the Non-leaders this is the opposite, their score should not be equal or higher as the score of the lowest opinion leader. If it turns out that the score of the interviewees is insufficient to belong to the supposed group, the interview will not be used in this research.

Qualitative analysis

Miles & Huberman (1984) present diverse methods that can be used in different situations and in different stages of the qualitative analysis. The focus in this study is on the analysis during the data collection and on the within-case analysis. A summary sheet will be used during the data collection and analysis, this sheet contains a series of focus and summarizing questions and remarks about a particular field of interests (depending on the interviewee; OL or NL). After a review of the field notes with the interviewee, which are made during the interview, the topics on the summary sheet are filled in to get an overall summary. Afterwards, a detailed write-up is made of the interview, this is done to note interview details which might affect the data analysis. By using this approach it is made sure that all the information, which is exchanged between the interviewer and interviewee, is right and complete as possible (Flick, 2007).

(22)

22

amount of data. By doing this in an early stage the generation of insight is also actively stimulated (Pettigrew, 1990). There is no standard format for within-case analysis (Eisenhardt, 1989), but for this study a checklist matrix will be used since it gives the researcher the opportunity to compare outcomes from different interviews in a clear and structured way. The checklist matrix combines the data from the OL and NL interviews on each interview question.

(23)

23

RESULTS

In this chapter the findings of the data gathering and analysis will be presented. 100 students filled in the self-designation questionnaire, 4 of the questionnaires could not be used since one or more questions were not filled in. For the NL interviews, 10 students were interviewed, after checking for being a NL 2 students turned out not to be a NL based on the self-designation questionnaire. These interviews are not used in the data analysis. Out of the ESN database 141 students were observed on Facebook, after the observation 23 suitable candidates were selected, they were the people that showed OL behavior for Web 2.0 concepts. These students were approached via Facebook if they want to be interviewed. 6 students want to cooperate and were interviewed. Via the self-designation questionnaire it was confirmed that all 6 students were indeed OLs for Web 2.0 concepts. According to the methodology part 10 OLs should be interviewed, this was not possible due to the limited available time for gathering data and the limited number of people who are suitable for an interview. A schematic overview of this process and results of it are given in on the next page.

Opinion Leaders and media

Figure 2: Overview of the proposed and the founded influence of the media and opinion leader on each other. In order to get an impression about how the OLs interact with the media, the OLs were asked about their contacts with the media and how they interact with each other. Out of the 6 interviewed OLs 2 of them had contacts with journalists, one OL has contact on a regular basis with a journalist since the two are friends. They discuss Web 2.0 concepts, however, this is not only done because of the general interest for the subject, but mainly since the OL is working on building his own Web 2.0 concept. The interaction is especially beneficial for the OL in this case since the journalists is well aware of Web 2.0 trends. The OL uses the journalists in this case for gathering market information to use it for its own concept. The other OL interviewee which is in contact with journalists is doing this only since it is beneficial for himself because he launched his own Web 2.0 concept in 2010 and he wanted to write media about it. He got in touch with journalists which specifically write about new Web 2.0 concepts and they covered his startup after their contact.

Looking the other way around something was found about how media influences OLs. 5 out of the 6 OLs stated that they are consuming specific media with respect to Web 2.0 concepts. The main source of media they are using are blogs, especially TechCrunch and Frankwatching (a Dutch blog about online trends) are read a lot. According to the interviewed OLs these blogs do not only inform them but also activate them, as one OL stated: “When TechCrunch is covering a new start-up which appeals to me, which happen every now and then, I immediately go check out their site and try the concept. I’m curious what the concept looks and feels like, if TechCrunch is covering it, it must be something good.”

Opinion leader Media coverage

Proposed: Found:

(24)

24 Figure 3: Schematic overview of the research process and its results

SD sample group (96 students) 100 students randomly picked SD Questionnaire SD analysis 10 students randomly picked SD Questionnaire & interview 8 Interviewees were NL: Data used 2 Interviewees were not NL: Data not used

Database of 141 Dutch students

Observation for Web 2.0 OL behavior

23 students are selected and approached

SD Questionnaire & interview (6 students)

SD analysis

6 Interviewees were OL: Data used

0 Interviewees were not OL: Data

not used

(25)

25

Depending on the concept they keep using it (and tell their friends about it) or reject it, means using the concept only ones. The OL who did not read specific Web 2.0 media heard about Web 2.0 concepts via two of his friends who are very into new internet concepts, when they are meeting they are discussing and showing him cool concepts. He again tells the rest of his network about the concept.

With respect to the Facebook case; all the OLs stated that they started using Facebook after hearing about it in their social network; they saw high status people using it and that is what triggered them to adopt it.

Opinion Leaders and their influence on the Non Leaders

Figure 4: Overview of the proposed and the founded influence of the Opinion Leader on the Non-Leader.

All NLs were told about Facebook by word-of-mouth, after hearing this several times from different people this became the incentive to start using Facebook. Based on the results of the NLs interview data it is not sure if the NLs were influenced by OLs in this process, since the interviewees cannot remember who exactly told them to start using Facebook.

Also OLs were interviewed and asked to their attitude towards non-Facebook users. The answers gathered from the OLs were only partly congruent with each other. It became clear that all the interviewed OLs informs non-Facebook users about the concepts, but convincing the non-user to adopt to Facebook differed per OL. While half of the OLs stated that they tried to push the non-user to adopt “You really should start using Facebook, as a student you cannot not be on Facebook”, other OLs stated that they did not really care about the non-user, one OL even stated the following : “I think it’s good that they are not on Facebook, it doesn’t make sense, it’s just a waste of time”. For Web 2.0 concepts in general, there was consistency about informing non-users about a concept, but activating non-users to adopt depended on the concept. The depending variable for the OLs here was how valuable it is that people in their social network adopt the concept as well.

Remarkable, the NLs themselves also informs non-Facebook users about the concept. Also here the same difference as with the OLs showed up; some NLs are actively trying to push the non-user to adopt, while others do not care about it. When asked to their behavior of Web 2.0 concepts in general there is however consistency among the NLs; they do not inform others about it. The given reason is that they think that they are not the person to inform/activate other users, because of a lack of knowledge and experience.

Opinion leader Non-Leader

Proposed: Found:

(26)

26

Media and their influence on the Non leaders

Figure 5: Overview of the proposed and the founded influence of the Media coverage on the Non-Leader.

As mentioned before, all the NLs were informed about Facebook via their social network and this is also what triggered them to start using it. According to the NLs the media did not influence them in adopting Facebook, it was the social influence that did. However, one NL interviewee stated that it did and said the following: “First, I got a lot of Facebook invitations in my e-mail inbox, I thought those emails were spam and I didn’t do anything with it. At a certain moment however, I saw a news item which was covering the story of Facebook, suddenly the pieces came together and I recognized the name from the ‘spam’ I got. Two days later I created my Facebook account”. In this case the media attention gave credibility to the concept and the invitations, which resulted in adoption. The other NLs stated that the media attention did not influence their adoption behavior.

Product/market fit

Figure 6: Overview of the proposed and the founded product positioning on both the Opinion leader and the Non-leader of the Media coverage on the Non-Leader.

For the both the OLs and NLs Facebook did added significant value to start and keep using Facebook. The OLs see the added value to started using it in the intrinsic value (its features), while the NLs names the extrinsic value (its users) as leading to start using it. Another remarkable finding is that OLs use Facebook in a different way than NLs. The interview data indicates that there are three main uses of Facebook:

1. Interacting: Posting and discussing issues on Facebook with the goal to keep your social network up-to-date and maintain friendships.

2. Branding: Posting and discussing issues on Facebook with the goal to brand yourself in your social network; have a very critical look on what to post on your Facebook-account.

Opinion leader Non-Leader Proposed:

Product positioning

Opinion leader Non-Leader Product positioning Found:

Media coverage Non-Leader Media coverage Non-Leader

(27)

27

3. Watching around: See what other people on Facebook and your network are doing. The goal is to be informed about other peoples’ life.

These three usages have two dimensions for the interviewees; active and passive. To underline the difference in usage between the OLs and NLs the following table is presented:

Active Passive

Interacting NL OL

Branding OL NL

Watching around NL and OL - Table 2: Different usages of Facebook by different groups

(28)

28

DISCUSSION

The previous chapter provided an overview of what is happening during the adoption process of OLs and NLs. The discussion continues by trying to understand why these things are happening. Though, connecting the explanations of findings with the literature should not be underestimated. According to Miles and Huberman (1994) an explanation is not just a comparison of theory and data. Whereby they see theory as facts of a qualitative research, which never speaks for themselves, and data as notions used by the researchers to explain the patterning of the theory. Which means, that the ‘facts’ who are discovered in this research are already the product of many levels of interpretation. Besides, Miles and Huberman (1994) remind that it is important to be careful with ‘explaining’ events. People tend to misperceive and misinterpret data, by paying too much attention to indefinite data and end up with biased results. The approach taken in this research is not only to explain the findings by the literature from the literature review but also have a fresh look at new literature in order to be able to explain the findings from different perspectives and reduce biased explanations. The findings will be explained by discussing the propositions. But first, the selection of OLs and NLs is discussed.

A critical point in this research was the selection of OLs and NLs. According to Mak (2008) the identification of appropriate OLs for practical purposes remains a challenge. The chosen self-designating questionnaire for this study is easy to apply to large groups, but it is not as reliable as the observation method (Solomon, 1999). The observation method in this study is again not fully reliable because it is about observing online behavior, which does not necessarily represents real world behavior (Kim et al, 2009); the combination of both takes away the weaknesses of both methods. By being very selective and checking twice for being an OL, only a limited number of people could be interviewed; 6 OLs and 8 NLs were interviewed. Interviewing more people would be beneficial for the analysis; however, due to time constraints and the small ESN database, which limited the number of OL interviewee candidates, interviewing more people was not possible. It can be argued that the selection of OLs for Web 2.0 concepts out of the ESN database limited the research more than it benefited of using the database. However, based on the approach of being picky who to interview, the reliability of the findings increases; the approach taken resulted in a 100% score with respect to identifying OLs: indicating that the OLs are real OLs. Despite the fact that only a limited number of people were interviewed, no new information showed up in the last interviews, indicating that the information need is saturated.

(29)

29

Lazarsfeld (1955) recognizes this point as well and suggested that the OLs, though more exposed to the media, also more often reported that they sought information and advice from other persons, which are likely to have influence on their adoption behavior. The third proposition proposed that the OLs influence the media. However, the limited visibility of the interviewed OLs to the media caused that influencing them was simply not possible. If OLs would shape the media if they were more visible to the media is unknown. The approach taken in this research is insufficient to answer that question and it might not even be relevant since the OLs were successfully identified: the observation and self-designation questionnaire showed congruent answers in identifying OLs; indicating that the right OL were interviewed and there no sign is of being wrong with this. Opposite, Nisbet & Kotcher (2009) do suggest that there might be influence of the OL on the media but that this influence depends on the nature of the subject , to be sure on this issue interviewing Web 2.0 media/journalists about how and who influences them would give a clarification on this issue. Overall, the interaction between the OLs and the media seems to be one way direction, whereby the media the OLs influence and, as our results indicate, activate to adopt a concept.

(30)

30

entry with complementary products or services (Shurmer 1993). In the Facebook case, this phenomenon becomes visible by offering more and more options to its users. In line with this is the shift in usage of Facebook the different groups (OLs vs NLs). First Facebook was used ‘to build your identity’ and view other profiles (Schulz, 2005), later on, the main point became interacting with friends. The interview results shows a similar phenomenon; the OLs, which were the early adopters, using more the initial ‘core’ functionalities of Facebook (watching around and building a brand), while the NL mainly use the functionalities which were offered later on (interacting with friends).

Summarizing, in the case of Facebook; targeting and positioning at a specific group of potential users led to adoption and created the opportunity to broaden their target group later on and satisfy them with other functionalities. According to Hultink, Griffin, Hart & Robben (1997) this launch strategy is called ‘niche targeting’ and is a common used strategy for launching innovative products. A remarkable difference in literature is found in the work of Lee et al. (2007), they state that unlike innovative products, network effect products succeed by pursuing to launch as broad as possible with the objective of rapid market penetration. This difference might find its origin in the characteristics of the concept of Facebook, according to Turow (2005) gaining trust was one of the essential points for Facebook to gain users; recall that they share personal information on their profile. Launching in a small, trusted community (Harvard University) created trust, while when launching broad from day 1 on, gaining trust would be hard. This implies for Web 2.0 concepts in general that those concepts should start with targeting and positioning at a specific group with one core functionality which appeals to the OLs. The extra features, which are needed to target the mass, can be built later on. Shuen (2008) recommends this strategy as well in her book “Web 2.0: a strategy guide”. This niche target strategy might not be only beneficial for the product diffusion but also for the resources of the start-up because this strategy does not require as much resources at the initial launch. This implies that the role the OL plays in the adoption and diffusion of Facebook is even more important compared to non-network products and the OL should be targeted actively.

(31)

31

use it; they are not in my network and I’m not interested in them”. The common factor of the OLs is however informing the NLs about a concept, this does not directly lead to adoption, but the accumulation of knowledge and familiarity with the concept results in adopting the concept, this is in line with the study of Greve (1998) which states that the rate of adoption of a product increases as its novelty decreases.

Summarizing, the OLs do influence the NLs by informing them about the concept, but there is no consistency about activating the NLs to adopt the concept. According to the NLs, it is not just one person who pushed them to start using it but it is the accumulation of people informing them about Facebook which results at a certain moment in adoption. The literature on opinion leadership agrees on that the OLs inform people in their direct environment and these close peers are likely to accept the advice from the OLs because they are seen as experts in a specific field (Weimann et al., 1994). According to Gelb & Johnson (1995) this word-of-mouth effect is more likely to activate people compared to mass media; which is congruent with the results found for propositions 4 and 7; the mass media did not influence NLs adoption behavior of Web 2.0 concepts.

Shuen (2008) gives an argument on why OLs were eager to inform non-Facebook users about the concept: Facebook benefited from direct network effects; the more a user’s friends use Facebook, the more valuable it is to the user. This suggests that the OLs do not only inform the NLs from their perspective as OL, but also because their own benefits will increase by doing so. Generalized for Web 2.0 concepts, this would imply that when a concept has network effects it stimulates the OLs even more to inform the NLs on the concept.

Proposition 1 Opinion Leaders influence media coverage by interacting with the media.

Not supported

Proposition 2 Opinion Leaders influences the adoption behavior of the Targeted Mass.

Supported

Proposition 3 Media coverage influences Opinion Leaders adoption behavior.

Supported

Proposition 4 Media coverage directly influences the adoption behavior of the Targeted Mass.

Not supported

Proposition 5 The Web 2.0 concept is positioned such that it adds significant value for the Opinion Leader (OL).

Supported

Proposition 6 The Web 2.0 concept is positioned such that it adds significant value for the Targeted Mass (TM).

(32)

32

Proposition 7 The increasing familiarity in the social network due to media coverage facilitate the adoption behavior of the Targeted Mass.

Not supported

Proposition 8 The increasing familiarity in the social network due to opinion leaders facilitates the adoption behavior of the Targeted Mass.

Supported

(33)

33

CONCLUSION

The aim of this study was to provide insight into the role of OLs in the adoption and diffusion process of Web 2.0 concepts. The conceptual model, which was leading in researching the role of the OL, was derived from the two-step flow model of communication. The factor of ‘product positioning’ was added as well to the model, which assumes different product/market fits for the two groups. By interviewing both OLs and NLs on this issue (with a focus on the Facebook-case), the model was checked. Despite the limited number of interviewees, the data showed congruency in the topics which can be explained by the strict and comprehensive method of selecting OLs and NLs for the interviews.

The analyzed data provided information about the main research question on what the role of the OL in the diffusion and adoption process is of Web 2.0 concepts. It can be said that this study found evidence for the Facebook case that OLs did play an important role in the adoption and diffusion process. The choice to target a specific group of OLs was beneficial for the adoption and diffusion process of Facebook; which confirms that the factor of ‘product positioning’ in the conceptual model is justified. The OLs, which were influenced by the media, had an informing role on the adoption behavior of the NLs; by becoming more familiar and more informed about the concept the NLs started to adopt the concept. Although it is clear what the role of the OL was for the adoption and diffusion of Facebook, the results cannot be directly translated for Web 2.0 concepts in general. The variety of characteristics of Web 2.0 concepts is too big to do this based on this single case study. The same counts for the results found with respect to the positioning of the product; for Facebook the diffusion worked out well to focus on a small target group at the initial launch and broaden their target group later on and serving them with other functionalities. Despite the fact that it can not be translated for Web 2.0 concepts in general, this research provides a direction for further research. With respect to the research questions on the role of the media, this study did not show surprising results. As both literature and this research shows, the influence of the media goes from the OL to the NLs, indicating that the NLs are not directly influenced by the media.

Limitations

(34)

34

Implications for further research

(35)

35

REFERENCES

Aaker, D.A. & McLoughlin, D., 2007. Strategic Market Management. John Wiley & Sons, New York.

Abrahamson, E. & Rosenkopf, L., 1997. Social Network Effects on the Extent of Innovation Diffusion: A Computer Simulation. Organization Science 8(3). pp. 289-309.

Afuah, A., & Tucci, C., 2000. Internet Business: Models and Strategies: Text and Cases. McGraw-Hill Higher Education, New York.

Arrington, M., 2005. 85% of college students use facebook. Retrieved from http://techcrunch.com/2005/09/07/85-of-college-students-use-facebook/

Arrington, M., 2006a. Odeo releases Twttr. Retrieved from http://techcrunch.com/2006/07/15/is-twttr-interesting/

Arrington, M., 2006b. A look at eight multi-person SMS services. Retrieved from http://techcrunch.com/2006/09/27/a-look-at-eight-multi-person-sms-services/

Arrington, M., 2009. Twitter Closing New Venture Round At $1 Billion Valuation. Retrieved from http://techcrunch.com/2009/09/16/twitter-closing-new-venture-round-with-1-billion-valuation/ Attewell, P., 1992. Technology diffusion and organizational learning: The case of business computing. Organization Science 3(1). Pp. 1–19.

Barriball, K, & While, A., 1994. Collecting data using a semi-structured interview: a discussion paper. Journal of Advanced Nursing 19(2). pp. 328-335.

Bass, F. M., 1969. A New Product Growth for Model Consumer Durables. Management Science 15. Pp. 215-27.

Beard, C., & Easingwood, C., 1996. New products launch: Marketing action and launch tactics for high-technology products. Industrial Marketing Management 25. Pp. 87–103.

Bennion, M.L., 1987. Segmentation and positioning in a basis industry. Industrial Marketing management 16(1). pp. 9–18.

Brosius, H.B., & Weimann, G., 1996. Who sets the agenda? Agenda setting as a two-step flow. Communication Research 23. Pp. 561–580.

(36)

36

Burt, R., 1999. The Social Capital of Opinion Leaders. The Annals of the American Academy of Political and Social Science 566. Pp. 37-54.

Campbell, D.T. & D.W. Fiske, 1959. Convergent and discriminant validation by the multitraitmultimethod matrix. Psychological Bulletin 56. Pp. 81-105.

Campbell, T., 1999. Back in Focus. Sales and Marketing Management 151 (2). Pp. 56-61.

Cavusoglu, H., Hu, N., Li, Y. & Ma, D., 2010. Information Technology Diffusion with Influentials, Imitators, and Opponents: Model and Preliminary Evidence. Journal of Management Information Systems 27(2). pp 305-334.

Carlsson, B., 2004. The Digital Economy: what is new and what is not? Structural Change and Economic Dynamics 15. pp. 245-64.

Childers, T.L., 1986. Assessment of the Psychometric Properties of an Opinion Leadership Sca1e. Journal of Marketing Research 2 (3). Pp. 184-188

Choi, H., Kim, S., & Lee, J., 2010. Role of network structure and network effects in diffusion of innovations. Industrial Marketing Management 39. pp 170–177.

Coleman, J.S., Katz, E. & Menzel, H., 1966. Medical Innovations: A Diffusion Study. New York: Bobbs-Merrill, New York.

Constantinides, E., & Fountain, S. J., 2008. Web 2.0: Conceptual foundations and marketing issues. Direct, Data and Digital Marketing Practice 9(3). Pp. 231-244.

Cooper, R.G. & Edgett, S.J., 1996. Critical Success Factors for New Financial Services. Marketing Management 5 (3). Pp. 26-37.

Childers, T.L., 1986. Assessment of the Psychometric Properties of an Opinion Leadership Sca1e. Journal of Marketing Research 2 (3). Pp. 184-188

Cooper, R.G. & Kleinschmidt, E.J., 1987. What makes a new product a winner, success factors at the project level. R&D Management 17. pp 175-189 (1987).

Cooper, D.R., & Schindler, P.S., 2003. Business Research Methods (8th ed). New York: McGraw-Hill Education.

Cooper, B. & Vlaskovits, P., 2010. Entrepreneur’s Guide to Customer Development. Cooper & Vlaskovits, Menlo Park.

Crunchbase, 2011. Facebook Crunchbase profile. Retrieved on 25-07-2011 from http://www.crunchbase.com/company/facebook.

The role of opinion leaders in the adoption and diffusion process of Web 2.0 concepts