Master Thesis
The effect of social media message characteristics on
branding of Dutch Monopolists
A case study at NV Nederlandse Gasunie
University of Groningen Faculty of Economics & Business
Master Thesis Business Administration – Strategy & Innovation
Name student: Steven van der Wal Student number: 1784307
Preface
The thesis that you are about to read is my last product as a master student in Business Administration, specialization Strategy & Innovation. I could say that developing this thesis has been the most time consuming and intensive, but also one of the most interesting, assignments of the last four year at the University of Groningen. By conducting my thesis in the form of a case study, I had the opportunity to gain experience in conducting research in an empirical setting, which is a valuable addition to the courses I had during my Bachelor and Master. By conducting research in a real business setting, results could be a direct contribution to the firm’s decisions.
Social media is still a new phenomenon. Little academic literature seemed to be available. At moments, it was hard to define how social effectiveness can be measured in a valuable way. Hours of searching for literature and reliable websites about social media measurement has taken place. Therefore, I could state that I am content that this study revealed some findings that are a contribution to understanding of use of social media within Dutch Monopolists.
First of all, I would like to thank the Gasunie for providing me an internship at an inspiring company. I would especially like to thank Mrs. Drs. Jet Havinga for her supervision during the entire project. I appreciated how much time she took to review and discuss several versions of my thesis.
At University, I would like to thank Mr. Dr. Rene van der Eijk for his supervision during this thesis. I appreciated his clear communication and valuable feedback on my thesis. I would also like to thank Mr. Dr. Florian Noseleit for reviewing my thesis as a second supervisor.
And last, but not least, I would like to thank my family, friends and girlfriend for supporting me in all sorts of ways throughout this study.
Groningen, August, 24th 2012
Abstract
Use of social media by businesses is constantly increasing. However, little is still known about its actual effectiveness. This study aims to identify ways in which social media use by Dutch monopolists, and in particular those with a high degree of potential resistance against its building activities, contribute to branding of the firm. This study focused mostly on Twitter. 500 Tweets of similar firms have been analyzed on several characteristics, while sentiment of 1,228 Tweets has been measured. Different methods to identify Twitter messages (Kwon & Sung, 2011; Lin & Pena, 2011; Hoffman & Fodor, 2010) have been found and tested in benchmark research. In addition, key-influencer identification seemed to be of importance when it comes to branding of the firm. Finally, in-depth interviews revealed more insights in use of social media by relevant firms in practice.
The study provides theoretical and practical insights in use of different types of messages that are expected to have a positive effect on sentiment towards firms, as a measure for branding. It revealed positive relationships between Company related Tweets, redirecting Tweets, Tweets that reflect a firm’s opinion, Tweets that offer a suggestion, and Tweets with verbs in imperative form, and sentiment. One negative relationship, between product/activity Tweets and sentiment has been found. Moreover, a combination of both quantitative and qualitative is proven to be necessary to properly identify key-influencers. These influencers seemed to have common characteristics. Moreover, degrees of interactivity for the different firms have been assessed.
Executive summary
Firms that are using or considering to use social media should be aware of the effects of the style they adopt in their communication. Little is still known about the effects of social media behavior on actual branding of the firm. Relevant previous studies mainly focused on ways in which messages could be characterized. However, which specific types contribute to branding of the firm? This study provides, therefore, guidelines for adequate use of Tweets that seemed to have a positive effect on branding. As a result, practioners of social media in the relevant types of firms are able to make the right decisions in deploying their messages when branding is an objective. In addition, deployment of messages that could have a negative effect on branding can be prevented.
Literature research revealed several ways to characterize online messages. Lin & Pena (2011) distinguished Task-instrumental and (positive and negative) Socio-emotional message styles. In addition, several forms of brand anthropomorphism (Kwon & Sung, 2011) have been presented as other potential measures for branding/attitude towards firms. The last 100 Tweets of the firms Noordzuidlijn, Tennet, ProRail, VanAnaarBeter and Gasunie have been analyzed on each characteristic provided by the authors mentioned above. Results pointed out that in some cases, large differences occurred. For example, Noordzuidlijn structurally uses human representatives in its Tweets, while other firms communicate less ‘personal’. In addition, sentiment of Tweets of the audience about the different benchmark firms are manually analyzed. Two different moments of measurement have been chosen, in order to decrease influence of certain events that could influence the sentiment about firms largely. A total of 1,228 Tweets has been analyzed on sentiment. In addition, the different categories of messages have been compared with sentiment. Results indicated several indications of positive relationships and one negative relationship.
Practioners should use the following types of Tweets if branding is the objective of the communication strategy; Company related Tweets, redirecting Tweets, Tweets that reflect a firm’s opinion, Tweets that offer a suggestion, and Tweets with verbs in imperative form, and sentiment. One negative relationship, between product/activity Tweets and sentiment. They should, however, be cautious with using product/activity Tweets, as results pointed out that there is a possible negative relationship with sentiment towards the firm.
Identification pointed out that most influencers have common factors. In their Tweets, they seemed to be;
- Focused rather than Broad - Consistent rather than casual - Listening rather than participating - Creating rather than sharing
Table of content
1. Introduction 7
1.1. Research objective and research questions 11
1.2. Relevance of the topic 11
1.3. Scope of the research 11
1.4. Overview of the paper 12
2. Theoretical framework 13
2.1. Definitions social media 13
2.2. Relevant factors for social media effectiveness 15
2.3. Summary 25
3. Conceptual model 27
4. Methodology 29
4.1. Choice for a case study 29
4.2. Literature research 33
4.3. Benchmark research – data generation 33
4.4. Qualitative interviews 36
4.5. Quality of the research 38
4.6. Summary 40
5. Results 41
5.1. Key-influencers 41
5.2. Online Benchmark research 43
5.3. Qualitative interviews 58 5.4. Summary 69 6. Discussion 70 6.1. Interactivity 70 6.2. Key Influencers 70 6.3. Message styles 72 6.4. Brand anthropomorphism 73 6.5. Information type 73 7. Conclusions 74
7.1. Research objective and research questions 74
7.2. Theoretical implications 75
7.4.Limitations of the research and future research 75
8. Findings and recommendations for NV Nederlandse Gasunie 77
8.1. Sentiment 77
8.2. Online interactivity 77
8.3. Key influencers 78
8.4. Overview of social media considerations made by the benchmark
Firms 79
References 82
Appendix I: Overview of key-influencer identification 84
Appendix II: Interview Noordzuidlijn 87
Appendix III: Interview Gasterra 93
Appendix IV: Interview Netbeheer Nederland 97
Appendix V: Interview VanAnaarBeter 102
Appendix VI: Interview Nacap 107
Appendix VII: Klout Influence Matrix 108
1. Introduction
The existence of the world wide web (WWW) has created new possibilities for communicating, collaborating, and sharing content (Enders et al., 2008). Today, most people cannot imagine a life without the internet. The rise of the so-called ‘Web 2.0’ has created more and more possibilities to communicate, collaborate and share content. Tim O’Reilly, a prominent IT-expert, author and media entrepreneur was the first to introduce the term ‘Web 2.0’; ”the business revolution in the computer industry caused by the move of the internet to understand the rules for success of that new platform” (Enders et al., 2008). People are making use of the internet in a different way than before; they often make more use of the internet and have the desire to share information about themselves and interact with others through the web. According to Kolbitsch and Maurer (2006), this is caused by three major developments. First, technological advance (for example, RSS feeds) has led to new interactive opportunities. Second, the number of people worldwide that have access to the internet through a broad band connection increased. Third, peoples’ internet behavior and mindset perceive the internet as a medium to share information and to identify themselves rather than a platform to derive content from portals.
The world wide web has increasingly won popularity since its introduction. Today’s businesses are making use of the internet on a large scale. It is often their main connection to their stakeholders. One of the most important reasons why those companies make use of the internet is the fact that it offers the possibility to communicate with (potential) customers. Moreover, the internet can be used as a quick sales channel.
As mentioned, social media may enhance brand awareness. One of the most prevalent advantages of the use of social media in branding is the ‘continued dialogue’ (Yan, 2011), which means that when users of social media ‘Like’ a certain message or firm, this is visible for all their contacts (for instance on Facebook) or by ‘re-tweeting’ a message from or about a firm to their followers (for Twitter), a certain firm can be constantly ‘in the picture’.
In order to determine the effectiveness of the use of social media, measures for the content of online interpersonal messages can be used. Lin and Pena (2011), for example, discussed two different message orientations; Task/instrumental and Socioemotional messages. Task/instrumental messages provide, among other things, the possibility to exchange information, ask for opinions and give suggestions. Socioemotional messages are more about sharing feelings about certain topics with others, showing solidarity and agree or disagree. In their case study, Lin and Pena (2011) identified different types of messages that television networks employed by using social media as Twitter to maintain and attract stakeholders (mostly viewers). Results indicated that television networks mostly made use of social media for informing and giving suggestions to viewers (task/instrumental), while Twitter users seemed to be the most influenced by a network’s socio-emotional messages.
1.1. Research objective and research questions
The previous has led to the following goal of this thesis:
”To complement existing literature on the effectiveness of the use of social media in terms of branding, among monopolistic firms by conducting a case study at N.V. Nederlandse Gasunie”. The goal of this thesis results in the following research question:
”How can firms in monopolistic industries communicate via and monitor social media to form positive attitudes among the general public towards the firm?”
The research question will be divided into the following sub questions:
How can messages of firms on social network sites be categorized and analyzed?
Which factors enhance branding of a firm within the relevant industry? ”Social media measurement is still quite immature, just as web-analytics was back in the mid-1990s.
Which criteria could be used to measure effectiveness of the social network site communication?
In order to answer these questions, a case study at N.V. Nederlandse Gasunie was conducted. Gasunie is a European gas infrastructure company that provides the transport of natural gas and so-called ‘green gas’ in the Netherlands and the Northern part of Germany. The organization provides gas transport to shippers (such as energy companies) through its subsidiaries Gas Transport Services (GTS) and Gasunie Deutschland. To make sure there is sufficient transport capacity in the future Gasunie also enlarges its gas infrastructure by placing gas tubes throughout the Netherlands and in the Northern part of Germany. The Gasunie transport system connects different gas systems internationally. Dutch policy is that the Gasunie system will function as the so-called Gas Roundabout of Europe. Therefore, the Gasunie system is internationally connected to gas fields in the Eastern of Europe and Norway in order to be able to facilitate its gas markets, but also the United Kingdom, Belgium and so forth. Finally, Gasunie has gas storage facilities in order to be able to meet demand at all times (www.gasunie.nl).
Figure 1.1. Online searches(standardized) in ‘Gasunie’ over the years (www.google.com/insights)
As can be seen, there is a minor downward trend over the years. The most striking peaks in the year 2011 are explained by several prominent appearances in the media. For example, at the point B, Gasunie announced that it cut off 100 to 150 jobs, and at point E, Gasunie announced that Gasunie and the firm Vopak started to start investigating new ways of distributing gas. Searches for Gasunie in digital media are highly influenced by press releases.
The following figure depicts searches for the Word ‘Gasunie’ and related words on Google, spread over the different provinces in the Netherlands.
Figure 1.2. Online searches for ‘Gasunie’ between 2007 and 2012 (www.google.com/insights)
employees) is located in Groningen and searches also involved searches for job openings. The organization is one of the largest employers in the North of the Netherlands, so Gasunie is one of the major organizations to check if there are job openings. However, the gas infrastructure is located throughout the whole country, so many citizens have to do with the organization.
1.2. Relevance of the topic
As mentioned, little is known about measures that can be used in order to identify the effectiveness of social media. More specific, the relationship between social media and effectiveness on corporate branding for firms has hardly been discussed in academic literature. So, the results of this research can be considered as complementary to existing literature. By applying theories that already exist to a specific business context, the existing theories could be deepened and new theories may be composed.
The results of this study will be especially beneficial for Gasunie itself. Based on the results of the master thesis, strategic recommendations on the use of social media will be given. As an indication to what extent different social media can be used and which types of messages are effective in realizing more brand awareness for monopolistic firms, results of the thesis may be guiding in the ways Gasunie may integrate social media within its corporate communication strategy.
Moreover, since benchmark research is one of the ways in which will be attempted to provide information about specific ways of coping with social media for monopolistic firms, the results could be beneficial for other firms with similar characteristics, (such as firms that need support from society for doing their businesses and that have a high need for awareness of their brand for the general public). Based on the results of this paper, these firms could make decisions about the deployment of social media in a responsible way.
1.3. Scope of the research
The research has been conducted during a case study within N.V. Nederlandse Gasunie. Benchmark research has been conducted via online data generation and in-depth qualitative interviews. In addition, online key-influencers for different firms have been identified. These firms have common factors; Dutch Monopolist, relatively low degree of product/service awareness for the general public, low to moderate appeal of the product/service to the general public, high need for brand awareness/acceptation, high degree of potential resistance against the business’ building activities, and no clear Business to Business or Business to Consumer market. Since strategies for the use of social media for business-to-consumer markets have other purposes, for example short-term purchases and updates about new products, these firms cannot make use of the results of this thesis. In addition, the high need for acceptance of the firms’ building activities could be regarded as a clear and specific boundary. All firms in the benchmark research meet the requirements mentioned, results of the thesis are mostly applicable for firms that operate in similar conditions.
1.4. Overview of the paper
2. Theoretical framework
In this section, several relevant theoretical concepts will be discussed. First of all, relevant definitions of social media and social network sites will be given. Then, attempts will be done to identify the factors that are relevant to measure social media effectiveness. Thereafter, the different factors that are found as indicators for social media effectiveness will be further elaborated.
2.1. Definitions social media
Social media: According to Lietsala and Sirkkunnen (2008), the term ‘social media’ can be described
as ”web services that receive most of the content from their users or that aggregate the content from other sites as feeds. The sites build on social networks and on the creativity of the participants of one or more communities”. Social network sites can be defined as web-based services that offer the possibility for individuals to (1) construct a unique (semi)public profile within a bounded system, (2) to construct a list of users with whom they are connected and (3) to view the activities of the connections in their list (Boyd, 2008). For the sake of this research, both definitions of social media and social network sites could be treated as interchangeable. Examples of widely used social network sites nowadays are Facebook, Twitter, LinkedIn, and Youtube. Internet is ‘alive’. This also means that the use of certain social network sites ‘die’, like at Myspace and Hyves, while new social networks like GooglePlus and Pinterest are upcoming. It is hard to predict their future roles as their success highly depends of the willingness of the mass to adopt the new social network site.
Facebook: A social networking community originally for college students, then expanded to include
high school students and since September 2006 available for everyone; founded by Harvard student Mark Zuckerberg and friends in February 2004; in October 2007, Microsoft bought a 1.6% share of Facebook for about $240 million (Bottles and Sherlock, 2011)
LinkedIn: A Business oriented social network of more than 75 million professionals from around the
world; founded by Reid Hofman and former employees of Paypal and Socialnet.com; launched in May 2003 (Bottles and Sherlock, 2011).
Twitter: A social networking and micro-blogging service (maximum of 140 characters in messages)
where updates are displayed on the users’ profile page and instantly delivered to other users (‘followers’) who have signed up to receive them; founded by Jack Dorsey and Biz Stone; launched in October 2006 (Bottles and Sherlock, 2011).
Youtube: A video-sharing service founded in February 2005 by former Paypal employees Chad Hurley,
Reach of the different websites:
The following table shows the estimated percentage of global internet users who visit the four different websites over the last six months (Alexa.com).
Figure 2.1. Daily reach of Twitter.com, Linkedin.com, Facebook.com and Youtube.com (Alexa.com)
As can be seen in the figure, Facebook and Youtube have by far the most global internet users. However, the different features of the different websites have to be kept in mind. For example, Facebook shows updates of users, but also provides a chat function and online games, while Twitter is more focused on exchanging statements in the form of micro blogs. The following figure depicts how often the users of the websites have visited website on a daily basis over the last six months.
2.2. Relevant factors for social media effectiveness
The use of social media has been increasing significantly over the last years, for both business and private use (Michaelidou et al., 2011). In a study under 1000 B2B Small and Medium Enterprises in the United Kingdom, Michaelidou et al. (2011) revealed that more than 25 percent of the SME’s uses social media to achieve their marketing objectives. In addition, none of the firms in the sample planned to decrease the budget for social media, while 44 percent of the firms was planning to increase the budget. Another research, (Murdough, 2009) underlined the importance of having a social media strategy and presented the main phases for the social media measurement process. As this thesis seeks to find the relevant characteristics of social media messages that influence attitude of social media users, the ‘Definition’- phase will have the most attention.
Figure 2.3. Social media measurement process (Murdough, 2009)
The Definition- phase emphasizes on the creation of an outline of how social network sites could be leveraged to reach and interact with the target audience. Monitoring of social media could be seen as a quantification of the impact of the consumer dialogue. Although social media activities can have influence on different points the relationship between a firm and its target audience, but Murdough (2009) argues that there are three pillars that have to be taken into account to cope with the complexity of social media performance;
Reach: Monitors the quantity of mentions and quality of authors. Not every author has as much as
Discussions: Is about understanding the chatter topics and sentiment composition. It is of great
importance to identify which topics are discussed and then which sentiment the messages have. Then, firms have to consider how they should respond to those messages, including tone and sentiment. Over time, firms can add value to discussions through conversations, which enables them to influence the topics and perceptions that are being discussed.
Outcomes: Diagnoses behavioral engagement indicators of intent of the audience. Outcome
measures could take the form of the factors that are relevant to the brand. Examples of these measures are site traffic and purchase intent (for B2C firms).
To conclude, there seem to be different aspects of measuring social media effectiveness. As mentioned, firms have to consider which topics are relevant for the audience and themselves. Moreover, key-influencers have to be identified based on the number of their mentions and the degree of influence they have. Furthermore, tone and sentiment of a firm’s messages should be adapted to what the audience appreciates. With regard to the outcomes, outcome measures have to take the form that are relevant for the specific firm. As this thesis is written within the context of Gasunie and the firm is focusing on brand awareness and appreciation of the firm (‘draagvlakcommunicatie’), the public opinion could be an indicator of the outcomes aspect. Sentiment measurement of the messages of the audience could be regarded as a measure of the public opinion.
Thus, tone and style of messages, topics, interactivity with the audience, the power of key-influencers and sentiment of tweets as an indicator of public opinion seem to be important indicators of social media effectiveness in the relevant context. In the following sections, the different aspects, which are divided in branding. information types and styles, and key-influencers, will theoretically further elaborated.
2.2.1. Branding – attitude towards firms
As branding is closely related to user’s attitude towards firms, branding and factors that enhance branding will be addressed.
Definition branding; Branding is defined as ‘a name, term, sign, symbol or design, or a combination of
Social media and branding; According to Yan (2011), any social media platform should serve the
organization in both and internal and external way, By making use of social media, the line between ‘them’ and ‘us’ could be less strict. The social media strategy should meet several requirements. At the outset of the process, the organization should:
a) Build a sense of citizenship with the organization;
b) Encourage the acceptance and communication of brand values; c) Encourage the audience to engage in dialogue and promote the brand. Strategically, the organization should:
d) Enhance the dialogue in such a way that it helps the organization with finding and maintaining a competitive advantage;
e) Inform the vision behind the brand and build differentiation for it;
f) Monitor whether the brand is properly communicated and understood by the relevant audience.
As a consequence, the organization should be able to: g) Build positive brand associations;
h) Build the perceived quality of the brand;
i) Build greater awareness of the brand to audiences that is not yet reached.
Hoffman & Fodor (2010) investigated to what extent it is possible to measure the ‘Return on Investment’’ (ROI) of social media. They understated the pressures and desires to quantify the return generated by investing in social media, but believe that most marketers are doing this in the wrong way. Traditionally, businesses emphasize on their marketing (and social media) investments and calculating returns in terms of customers response. However, when it comes to social media, this is not applicable.
In their opinion, measures should be more focused on the consideration of consumers to use social media and then measure social media investments customers make as they engage with the marketer’s brands. A company should think about what marketing objectives messages might satisfy, why its customers would visit a blog of social network site, and what behavior they would show if they are on social media. This should be linked to the social media strategy. Whether certain social media measures are appropriate depends on the objectives. These could be divided in Brand Awareness, Brand Engagement and Word-of-Mouth. Each objective has its own fitting measures. These are depicted in the following table (Hoffman & Fodor, 2010):
Social Media Application
Brand Awareness Brand Engagement Word of Mouth
Microblogging (e.g. Twitter)
Number of tweets about the brand Valence of tweets (+/-) Number of followers Number of followers Number of @replies Number of re-tweets Social Networks (e.g. Facebook and LinkedIn) Number of members/fans Number of installs of applications Number of impressions Number of bookmarks Number of reviews and ratings and valence (+/-)
Number of comments Number of active users Number of ‘Likes’ on ‘friends feeds’ Number of user generated items (photo’s, threats, replies) Usage metrics of applications/widgets Impressions to interaction ratio Rate of activity (how often members personalize profiles, bios, links, etc.)
Number of references in mock-ups or derived work
Number of times republished in other social media and offline
Number of ‘Likes’
Table 2.1. Effective social media measures per marketing objective (Hoffman & Fodor, 2010)
As interaction is expected to increase the number of followers, it will be tested whether this is the case.
Proposition 1: The more interaction with other users, the more followers a firm has 2.2.2. Key influencers
Due to the rise of social network sites, new key influencers on branding of firms occurred. ”Nobodies” could transform into ”Somebodies” by demanding attention for their messages. ”Ordinary” consumers are gaining a large share of the voice within the market because of the power of the internet and technologies that bring together people with similar interests (Booth & Matic, 2011). Social media influencers (SMI’s) ‘represent a new type of independent third party endorser who shape audience attitudes through blogs, tweets, and the use of other social media’ (Freberg et al., 2010, p.91). Several other reasons for identifying key influencers on an organization are (Falls, 2011);
- They make blogger or media outreach for the organization or clients more efficient
- They help spread the word about an organization’s news or programs to more people and faster
- They help organizations build relationships with the online voices that have the largest impact on the relevant industry
However, identifying the most influential individuals is challenging. Falls (2011) states that there is actually no perfect system to measure how much influence individuals have on companies. When obtaining all available data, one has to consider all data and identify the pieces that are most important for an individual organization. Falls (2011) further describes that both quantitative and
qualitative data have to be considered to obtain the most reliable results. There are dozen of ways
not considered as reliable. For example, Fall’s own website, socialmediaexplorer.com had been entered in several websites, like Quantcast.com, Compete.com and GoogleAnalytics.com. All sites showed, for example, different numbers of unique visitors per month. He therefore argues that no site could be trusted for hundred percent. He further mentions Klout.com as a potential relevant site, as is has been improving its method since its start-up. Klout presents individuals’ influence on a 100 point scale, which is based on their online activities on different social media. At first it focused on Twitter, but currently it makes also use of social media behavior on different social network sites as Facebook, LinkedIn, Foursquare, and Google+ (www.klout.com). One of the drawbacks mentioned by Falls (2011) is that the degree of influence only applies to social media and not in real life. Currently, if someone has a small followers base, this could mean that he or she is not influential. However, if this person is an influential journalist in real life, its tweets could be more worth and influential than those of someone that just has many friends as followers. As Vanderveer (2012) notes, Klout determines how many people a Twitter user influences, how much he or she influences them, and the power of the network someone has. So, however some drawbacks are mentioned around Klout, it could function as an indicator for one’s influence on social media.
Figure 2.4. Klout’s Influence Matrix (www.klout.com)
As can be seen in the matrix, influence is divided in 16 different categories of social media users, using the variables Sharing versus Creating, Broad versus Focused, Casual versus Consistent, and Listening versus Participating. Klout locates every user in the matrix in order to explain which behavior is expected from the users. By being aware of the types of influencers on a firm, firms can use these expected behavior in their social media strategy.
However, making only use of quantitative data about key influencers is not sufficient to judge their true power. As Falls (2011) states, identifying bloggers’ power also depends on qualitative judgment of their social media behavior. By actually reading messages that have been posted by the influencer on a frequent basis, firms can obtain much additional information about the relevance and power of someone’s online messages. Questions that could be asked in order to judge the power of an influencer are, for example (Falls, 2011);
- Is he or she respectful towards firms in messages?
- Has he or she actually written about my product, firm or industry so far? - If this was the case; was this in a fair way?
- Would my audience enjoy reading messages from this person and why? - How many comments does the average message of this person have?
When quantitative data derived from, for example Klout, is combined with qualitative data gained from own interpretation of influencer behavior, firms could be able to obtain an adequate image of the extent to which someone is influential on the public opinion. Fall (2011) emphasizes much on the fact that firms also have to use common sense when it comes to interpretation of social media behavior. It could sometimes be more beneficial to personally decide whether, for example, someone could be an opportunity or threat to the firm by communicating openly about a firm. The combination of these two seemed to be the most appropriate.
As literature pointed out that both qualitative (in this case; manual interpretation of key-influencers) and quantitative analysis (in this casel the results of the Klout scores) of key-influencers is needed, it can be tested whether this is the case in practice.
Proposition 2: Qualitative analysis of influencers contributes to quantitative analysis of
key-influencers.
2.2.3. Message styles and information types
This section will discuss several ways to analyze online messages.
As mentioned, Lin and Pena (2011) presented a difference between Task/Instrumental and Socio-emotional (positive or negative) message orientations (in this thesis: style) when it comes to deploying messages via social media to stakeholders. Both orientations have several categories. These are, together with their aim and an example, presented in the following table;
Message orientations Aim of the message Example
Task/instrumental:
Ask for opinion Evaluation, analysis, expression of feeling
What do you think of the problem?
Ask for suggestion Direction, possible ways of action What can the actor do in the given situation?
Ask for information Information, repetition, confirmation
Gives opinion Evaluation, analysis, expressing feeling, wish
The show was really amazing last night
Gives suggestion Direction, implying autonomy for the other
Watch the show tonight
Gives information Information, repeats, clarifies, confirms
The preview of tonight’s episode is now available online
Positive Socio-emotional:
Shows solidarity Raises other’s status, gives help, reward
Thanks so much for liking the show
Show tension release Jokes, laughs, shows satisfaction Wow, that was funny
Agree Shows passive acceptance,
understands concurs, complies
Yeah, I agree with you
Negative Socio-emotional:
Disagree Shows passive rejection, formality withholds help
I told you that’s not allowed in here
Shows tension Ask for help, withdraws out of the field
I am not happy without it
Shows antagonism Deflates other’s status, defends or asserts self
Why don’t you just shut up?
Table 2.2. Task/Instrumental and Socio-emotional messages in social media (Lin & Pena, 2011)
Dimensions Operationalization
Brand
anthropomorphism:
Human
representatives
Pictures of marketers, celebrity endorsers, or their information (name, contact information), as well as signature of marketers on tweets Personal pronouns First-person (e.g., "I," "my," "me," "myself," "we," "us," "our")
Second-person (e.g., "you," "your," "yours," "yourself") Verbs in imperative
forms
Asking for Relationship
Follow, stay tuned, become a fan on other social media, sign up/register, join an event, support the company (e.g., vote for us)
Asking for Feedback & Contact Customer Service
Let us know what you think, submit your suggestions, e-mail us, send us a direct message, text us, contact customer service
Redirecting to Other Media
Check out the links, watch new ads, learn about something Sales
Use/purchase products, preorder, download, enter sweepstakes or contest Electronic word of mouth
Tweet, share tweets, re-tweet Nonverbal cues Abbreviations (e.g., LOL, OMG)
Emoticons (e.g., :), :( )
Repeated punctuation (e.g., !!!, ...) Capitalization (e.g., FOLLOW US, )
Sound mimicking (e.g., Ahhhhhh, ooopsy)
Information Types:
Product/activity Price, quality, performance, components, availability, special offers, taste, packaging, nutrition, new ideas, download
Company Job-related, sales-related, sponsorship/affiliation, investor relations, education
Source Independent research, company-sponsored research, executives, celebrities, media
Redirecting 800 number, URL, mail-in/e-mail address Simple brandmention Brand name
The previous has led to the first proposition;
Proposition 3: The higher the number of number of positive message styles, the higher the chance
that sentiment towards the firm is positive
As the different forms of brand anthropomorphism are considered as enhance branding, so it could be expected that an increase in messages with these characteristics also increase sentiment.
Proposition 4: The higher the number of Tweets that show brand anthropomorphism, the more
positive sentiment towards the firm
In addition, with regard to the information types, it could be expected that monopolistic firms that have a high degree of potential resistance against the business’ building activities highly value the product/activity information type. However, it could also be expected that sentiment about these building activities (that fall under product/activity) will receive negative attention. Therefore, the second proposition will address this aspect:
Proposition 5: The higher the number of product/activity messages, the higher the chance that
sentiment towards the firm is negative.
2.3. Summary
This chapter focused on the main theoretical concepts with regard to social media effectiveness and its context. It started off with an overview of the relevant definitions to increase understanding of the phenomenon. Then, attention has been paid to the increasing recognition of social media and different stages for developing an appropriate strategy.
3. Conceptual model
Based on the theoretical framework, some factors seemed important to be taken into account when it comes to identifying factors that could enhance the general attitude towards monopolistic firms. Firstly, Kwon & Sung (2011) identified five different information types, based on the recent changes within the media environment. These types are useful to research which effects using the different types have. Secondly, certain social media metrics are presented as useful means to measure effectiveness of social media efforts in terms of branding (Hoffman & Fodor, 2011). Thirdly, Task Instrumental and Socio-emotional message styles are seen as a framework to study computer-related interactions (Lin & Pena, 2011). In addition, some forms of brand anthropomorphism could have stronger effects on attitude/branding than others and are therefore worth to investigate. Finally, as several researchers underline the power of key-influencers on the general opinion about firms (Booth & Matic, 2011, Freberg et al., 2010), identification of common characteristics of key-influencers in the relevant industry seems desirable. The different relevant variables that are expected to be influential on the attitude towards the firm will now be conceptualized.
Figure 3.1. Conceptual model of the research
Based on the literature review, it could be stated that information type, degree of online activity, message style and key influencer’s characteristics could have an influence on the sentiment of an attitude towards the firm. They could be considered as indicators for people’s attitude towards the firm. It is expected that the attitude the general public has towards a firm will depend on the
information types that are used by the firm in social media. A firm could, for example, choose to
focus its messages on providing information about the activities, providing links to interesting websites, or to just mention the brand name in short messages. As can be read in the literature review, social media messages are categorized in several information types (See Kwon & Sung, 2011). The number of all tweets, re-tweets, and replies on Twitter could be regarded as an indication for the online activity. The degree of online interactivity of Gasunie will be measured and benchmarked with other companies that have similar relevant characteristics (see methodology section). Afterwards, it will be attempted to relate the degree of interactivity to the attitude towards the firm. The degree of online interactivity will based on data like generosity of updates, Following, Re-Tweeting and referencing (‘@username’) behavior. Finally, the benchmark research will also possibly reveal differences in message styles. Message styles could be characterized by making use of the different brand anthropomorphism dimensions of Kwon & Sung (2011) and Task/Instrumental and Socio-emotional characteristics of messages presented by Lin & Pena (2011). Differences in message styles could influence the attitude of the general public towards a firm. Monitoring the effectiveness of social media behavior, as mentioned at the research questions, will reveal to what extent it is possible to define measurements for effective application of information types, degree of online interactivity and message styles and will depend on the results of the literature review and benchmark research.
The different information types could have different effects on the attitude towards a firm. As Hoffman & Fodor (2010) state, brand awareness of the brand can, for example be measured by the number of tweets about the brand or the number of followers. This could be seen as an indication for a positive or negative attitude towards the firm. However, data like number of followers and frequency of mentions do not say anything about the sentiment of the attitude. Therefore, sentiment will be measured based on own interpretation by judging whether the tweets could be regarded as positive, neutral, or negative. Practically, Twitter reveals tweets from the last 7 days on a searched topic. The number of tweets about different topics (benchmarked firms) could differ. This has to be taken into account.
4. Methodology
In this section, the ways in which data will be searched and analyzed is discussed. Literature research, benchmark research and qualitative interviews has been conducted in order to answer the research questions.
Dul and Hak (2008) discuss the decision between theory-oriented and practice-oriented research. They describe theory-oriented research as ”research that is aimed at contributing to the development of theory” (p.30). Practice-oriented research, in addition, is ”research that is aimed at contributing to the knowledge of specific practioners responsible for a specific practice” (p.30). As this study is aimed at contributing to the existing theory about social media use by Dutch monopolists with no clear B2B or B2C industry and a high need for acceptance of their activities, it could be stated that this research has characteristics of both types. They refer to the famous quote; ”Nothing is quite so practical as a good theory”. This also seems to apply to this study, as there is little theory available about social media use in this specific industry and the results of this study could partially close this knowledge gap, and results could be considered as practically useful for Gasunie in its corporate communication strategy. However, theory-oriented research has different manners in which theory is developed. Dul and Hak (2008) distinguish three different types of activity that contribute to theory development; exploration, theory-building research and theory-testing research. By applying the exploration type, which stands for the collection and evaluation of relevant information and practice in order to assess how exactly research could best contribute to theory development, propositions about theoretical concepts were developed.
find evidence for the propositions. The danger of these ‘single-shot’ studies is that wrong conclusions about the concepts will be drawn. The time scope of this thesis does not allow to repeat the study completely. However, multiple moments in time have been chosen to collect data of Tweets about the different benchmark firms. As a result, the data could be regarded as a more reliable reflection of the population. In order to further increase generalizability, replicating the study has been mentioned (Dul & Hak, 2008). As just mentioned, the time frame of the thesis does not enable to repeat the entire study and future replication could be beneficial.
Fig. 4.1. Steps taken in flowchart for deciding on the type of theory-oriented research (Dul & Hak, 2008)
and Twitter make it possible to automatically send posted messages Twitter to the Facebook profile and vice versa.
For this research, the focus will mostly be on Twitter. Firstly, literature and internet search pointed out that there are much more possibilities to analyze Twitter behavior than for Facebook. Second, as Tagtmeier (2010) stated, it is important to assess the usefulness of social network sites. Currently, Gasunie has only 42 followers on Facebook. Also the firms to be benchmarked showed few followers and show little to no activity on their Facebook page. Gasunie’s Twitter account has over 1,000 followers and this number is increasing slowly, while some others have over 2,000 followers. Obviously, using Twitter is more suitable to analyze online behavior and attitude amongst users. 4.1. Choice for a case study
According to Yin (2009, p.18), a case study is ”an empirical inquiry that investigates a contemporary phenomenon in depth and within its real-life context, especially when the boundaries between phenomenon and context are not clearly evident”. In addition, ”The case study inquiry copes with the technically distinctive situation in which there will be many more variables of interest than data points, and as one result relies on multiple sources of evident, with data needing to converge in a triangulating fashion, and as another result benefits from the prior development of theoretical propositions to guide data collection and analysis” (Yin, 2009, p.18). The research at Gasunie is in line with these descriptions. As there are many variables that determine the public attitude towards firms, several sources of information have to be used. Theoretical propositions will be derived from academic literature, qualitative in-depth interviews will be conducted and social media behavior of different comparable firms will be analyzed by generating primary data about their behavior and the attitude of the online audience towards these firms will be measured. In addition, the results of these aspects will be combined with the Gasunie’s communication strategy. As a result, the insights gained from the literature research, benchmark research and qualitative interviews give an indication of the aspects that have to be incorporated at in the social media/communication strategy of Gasunie to become more effective.
Method Form of the research question Requires control of behavioral events? Focuses on contemporary events?
Experiment How, why? Yes Yes
Survey Who, what, where,
how many, how much?
No Yes
Archival analysis Who, what, where, how many, how much
No Yes/no
History How, why No No
Case study How, why? No Yes
Table 4.1. Relevant situations for different research methods (Yin, 2009)
Since the research question is ”How can firms in monopolistic industries communicate via and monitor social media to form positive attitudes among the general public towards the firm?” and the sub questions are ”How can messages of firms on social network sites be categorized and analyzed?”, ”Which factors enhance branding of a firm within the relevant industry?” and ”Which criteria could be used to measure effectiveness of the social network site communication?”, the ‘how’ and ‘why’ questions seem to be relevant. As later on in this chapter will be explained, data on social media use will be generated. In this respect, the ‘how many/how much’ question could also be appropriate for the research. As this research is combined with an internship at Gasunie and the communication strategies behind the social media behavior (generated data) is at least of importance, archival analysis does not entirely contribute to understanding these strategies. The fact that social media is a relatively new phenomenon and the research is combined with an internship, the focus can be on contemporary events. In addition, social media behavior and the attitude towards firms can be largely influenced by recent events. Control of behavior could be relevant when it is relevant to influence behavior directly and the researcher is able of doing so (for example, in an laboratory setting). Then, variables could be isolated and tested, and it is presumed that the laboratory environment is able to control all the remaining variables beyond the variables of interest. However, due to the fact that all variables that influence social media behavior and attitude could be relevant, it is not appropriate to isolate certain variables or events. Social media behavior and attitude is always influenced by the frame of reference of the users and events that took place. So, there is no requirement for control over behavioral events.
4.2. Literature research
In order to be able to give sufficient answers to the different research questions, extensively literature research had to be conducted. As can be expected, academic literature on social media is becoming increasingly available. Since its introduction in about 2006, more and more authors did research to the new phenomena. Many articles concern explanations about the meaning of and possibilities that several types of social media offer. Unfortunately, there is still relatively little academic literature on the measurement of social media and its effectiveness. In addition, the relation between the deployment of social media and its effect on corporate branding has hardly been discussed.
4.3. Benchmark research – data generation
In order to be able to analyze social media use and the effect on corporate branding, organizations with similar characteristics to Gasunie will be analyzed. Benchmark research will reveal in which ways these companies cope with the opportunities (and threats) offered by social media, and to which extent this contributes to brand awareness with the general public. Important factors of Gasunie, relevant for comparing it to other organizations, are:
- Dutch monopolist
- Relatively low degree of product/service awareness for the general public - Low to moderate appeal of the product/service to the general public - High need for brand awareness/acceptation
- High degree of potential resistance against the business’ building activities - No clear Business to Business or Business to Consumer market
- Prorail (an organization that is responsible for the maintenance and expansion of the Dutch railway infrastructure)
- VanAnaarBeter (an information provider of the Ministry of Infrastructure and Environment - Verkeer en Waterstaat- about infrastructure projects in order to stimulate safety and accessibility)
- Noord/Zuidlijn Amsterdam (a large scale construction project that builds a new subway line under the historical centre of the city of Amsterdam)
4.3.1. Measures used in benchmark research
Based on the measures for social media’s effectiveness obtained from the literature research, the activity on social media of the in this thesis included firms has been be compared. In addition, for each firm, it will be attempted to relate the activity on social media to their corporate branding. The benefits of this aspect could be that corporate branding will be traced to certain characteristics of messages in the social media that are used. The last 100 tweets of each firm, for example, could be analyzed to find differences in the characteristics derived from the literature review and could then be compared with the interactivity measures like the number of followers, re-tweets and replies in order to find a relationship between certain message characteristics and brand recognition or a positive attitude towards the firm.
Now, each variable that is measured will be discussed in relation to the theories and possible data sources used, the method used to obtain the data, and how the results will be presented in the thesis.
Variable Theories and other sources used
Method Result
Information type
Kwon & Sung (2011), Twitter.com
Based on the different information types by Kwon & Sung, a sample of tweets will be manually
identified/scored for each benchmarked firm
An overview of the different proportions of used information types for each benchmarked firm
Degree of online activity
Twitalyzer.com Data to indicate online activity for each benchmarked firm will be generated from Twitalyzer.com
Generosity of updates, Following, Re-tweeting and referencing
(‘@username’) behavior
Message style Kwon & Sung (2011), Lin & Pena (2011), Twitter.com
Based on the characteristics of Brand anthropomorphism of Kwon & Sung (2011) and the different message
orientations presented by Lin & Pena (2011), a sample of tweets will be manually identified.
characteristics for each firm.
Key Influencer’s characteristics
Booth & Matic (2011), Tweetreach.com, Topsy, com, Klout.com
Based on data about key influencers (most contributing users), derived from Tweetreach.com, the most important influencers will be identified.
Based on Klout, key influencers on the benchmarked firm can be obtained and categorized in a Klout Influence matrix.
Tweetreach and Topsy.com indicate, among other things, the top contributors to a certain firm.
An identification of characteristics of the key influencers for each benchmarked firm in an influence matrix.
Attitude towards the firm
Hoffman & Fodor (2010),
Twitalyzer.com, Twitter.com,
Based on data retrieved from Twitalyzer, number of followers, number of re-tweets by others, Potential reach, effective reach will be identified. Moreover, based on own interpretation, the sentiment of messages about each firm will be analyzed.
Number of followers, number of re-tweets by others, Potential reach, effective reach.
Sentiment of the
messages about the firms will be presented in positive, neutral, or negative character.
Table 4.2. . Measures, theories used, method used and results of method used during thesis
4.3.2. Explanation of sources for measures
of time. In addition, Twitter.com will be used to identify the sentiment of tweets about the different firms to be able to indicate the attitude towards firms.
Researchers have been attempting to analyze the sentiment of tweets (Savage, 2011). Often, word analysis is used in order to base the sentiment of tweets on the use of clear positive or negative words. For many of these words, the emotional value can be predetermined in order to indicate a certain ‘mood’ of a tweet. However, this way of analyzing the sentiment of tweets has major limitations. For example, the word ‘Sick’ could be argued as negative. However, in certain ‘slang’ – street language- this word even could be considered as very positive. In addition, it is almost impossible to include all words that are indicative for a tweet to be positive, neutral or negative (Savage, 2011). In addition, many tools that could be used to identify the sentiment of tweets, are freely available, but are only applicable for English written tweets. Freshnetworks (Freshnetworks.com), for example, provides an overview of 30 free tools for finding social media influencers and sentiments, but all are in English and are therefore not appropriate for analyzing the sentiments of tweets within this research. One of their presented websites; feeltiptop.com, claims to provide an overview of tweets about certain topics or users, also in Dutch. However, their method of identifying tweets as positive or negative does not seem to work. For example, when searching for ‘Prorail’, the tweet ” Weer bijna een treinbotsing vandaag, gaat lekker op het spoor. En dan nog “zelden“ zeggen. #prorail ” is regarded as a neutral message and the tweet ” vertraging met de trein 010-020; reden: glad spoor. KOM OP ZEG, het is geen herfst of winter, zeg dan niets #NS #prorail #klagen ” is regarded as positive. Therefore, the most efficient way to determine the sentiment of tweets about the involved organizations in manually.
According to Internetmonitoring (intermonitoring.nl), positive, negative and neutral tweets are described as follows;
- Sentiments are considered as positive when a message contains satisfaction about a certain event. Re-tweets are also considered as positive, as the twitter user shares a certain opinion or message to support it or show solidarity with it and therefore could be regarded as positive.
- Negative sentiments are given to tweets that show aversion or that cause a negative experience.
- Neutral sentiments are given to messages that do not contain an opinion. 4.4. Qualitative interviews
the main aspects of the organizations’ strategies are and the way they bring these into practice, several experts in social media have been interviewed. Another motivation for conducting additional interviews is that literature about online influencers points out that identification of key-influencers should, amongst other things, based on own interpretation and common sense. This process has not been found in the literature. Finally, as social media are still new, especially to the relevant industry, recent experiences from firms that are considered as active in social media could be highly valuable in addition to the data-analysis.
By making use of the network of some Gasunie employees, relevant employees of comparable organizations were asked to share their insights and experiences in an interview on the use of social media to increase acceptance/branding of their organization. Based on contacts obtained during the internship, five different companies have been chosen to arrange face-to-face, semi-structured interviews. Unfortunately, Tennet and ProRail could not participate in an interview. Noordzuidlijn and VanAnaarBeter did offer the opportunity to ask questions in person. By adding Gasterra, Netbeheer Nederland and Nacap, several firms that are relevant to Gasunie’s industry are included. Unfortunately, the interview at Nacap revealed that policy on social media is not been used any more, as the department that was responsible for social media usage had been cut off. Therefore, only the results of interviews at Noordzuidlijn, Gasterra, Netbeheer Nederland and VanAnaarBeter will be discussed.
Firm name and location
Interviewee name(s)
Occupation Date Duration
Noordzuidlijn Amsterdam
F. Elink -Schuurman
Consultant Corporate Communications and new media April 27th 2012 1 hr. 30 m. Gasterra Groningen I. Weel M. Kanon Senior Communication Advisor Communication Advisor May 16th 2012 60 m. Netbeheer Nederland Arnhem
M. Boelhouwer Senior Communication Advisor May 23th 2012 1 hr. 15 m. VanAnaarBeter Utrecht L. van Dijk C. Zeren J. Fokkema
Online Channel Manager Advisor Online Marketing Online Public Relations
May 30th 2012 60 m.
Veendam I. ter Beek Tender Coordinator
Table. 4.3. Participants in qualitative interviews
All interviews have been based on the following pre-determined structure: 1. Introduction of interviewer and interviewee(s)
2. Explanation on the purpose and structure of the interview 3. Discussion on report of interview and publicity of final thesis 4. Explanation on the structure of the interview
5. Possibility to ask preliminary questions 6. Discussion of 14 questions
7. Closing the interview
The research questions have been based on the literature research. The interviews have been divided in five different sections. Firstly, the general context of the use of social media by the firm is discussed. The other sections are based on the literature review. Secondly, categorization of social media messages and interactivity is addressed. Questions are mostly focused on information types, sentiment, re-tweeting and following. Thirdly, identification of, and coping with key-influencers will be given attention. Fourthly, branding of firms via social media is addressed, which covers awareness of stakeholders’ needs and sentiment, and the relationship between social media and attitude towards firms. Finally, manners in which the firms monitor the results of their social media efforts and overall success factors for effective social media use will be elaborated. The entire work-outs of the interviews can be found in the appendices.
4.5. Quality of the research
Yin (2009) presented different ways to determine the quality of research designs. Firstly, construct validity is about whether the correct operational measures for the concepts being studies are identified. Important, here, is the use of multiple sources of evidence. This is the case in this thesis, since documentation, archival records and interviews are used in order to increase the construct validity. An important factor to take into account is the factor that authors are biased, which will influence the choice for sources. Another important factor is to have key informants that review draft versions of the case study. At Gasunie, Mrs. Havinga, reviewed draft versions on a regular basis. In addition, several co-workers at Gasunie are asked to give their opinion about the correctness of certain parts of the thesis. At University, mister van der Eijk (first supervisor) and mister Broekhuizen (second supervisor) will review the quality of the thesis.
Secondly, internal validity is about searching for a causal relationship, where certain conditions are expected to lead to other conditions. It is important to consider if inferences are correct. In addition, it has to be checked whether all rival explanations and possibilities have been considered. During the thesis, these factors were constantly kept in mind in order to increase internal validity of explanation building.
relatively high. However, the results derived from the online benchmark research could differ when conducting a new research, since especially attitude towards certain firms is very dependent of relevant events that took place and influenced the public opinion among social media users. For each firm, the last 100 tweets will be considered. In addition, the attitude of the Twitter users towards each firm will be identified, based on sentiment of tweets of the last seven days. The results of both are very time dependent, as for example information types and attitude towards certain firm is influenced by events at the period in which the tweets will be identified. To increase the reliability of the results of the attitude towards firms, tweets about the firms of the last 7 days will be measured on two different moments. As these tweets are only judged on its sentiment (positive, neutral, or negative), this is not highly time-consuming, but increases reliability of the results significantly. 4.6. Summary
5. Results
In this section, the results of the different research methods will be presented. Firstly, an identification of the key-influencers for the different firms (Noordzuidlijn, Tennet, ProRail, VanAnaarBeter, and Gasunie) is given. Secondly, the identification of different information types and styles by generating data from Tweets on Twitter.com is discussed. Thirdly, findings based on the qualitative interviews at several firms that are active in the use of social media will reveal in which way different firms apply their social media to effectively communicate with their stakeholders and how they monitor the effectiveness of these efforts.
5.1. Key- influencers
As Falls (2011) mentioned, the identification of the power of key-influencers is a complex endeavor and needs to consist of both qualitative and qualitative aspects. Qualitative aspects are about judging the power of key-influencers based on own interpretation and common sense. Although this aspect could be seen as time-consuming, it will contribute to the quality of identification of key-influencers significantly.
In the following paragraph, the focus will be on the results on quantitative identification of influencers. For each firm (Gasunie, Tennet, Prorail, VanAnaarBeter, and Noord/zuidlijn), the key-influencers, based on the number of contributions, generated by Tweetreach (Tweetreach.com) are presented. Then, the key-influencers are entered in Klout (Klout.com), which is able to generate their Klout-score, indicating their online influence. Klout provides the possibility to characterize a certain user based on his/her network, number of mentions, frequency of postings and range of topics discussed. In addition, each key-influencer will be placed on the Klout-influence matrix (see 2.2.2)so the main characteristics of the key-influencers of each firm can be identified. Tables that provide an overview of these characteristics can be found in the appendices.
Legend: Prorail: Tennet: VanAnaarBeter: Noordzuidlijn: Gasunie: Sharing Creating Li sten in g Pa rti ci p ati n
g Curator Broadcaster Tastemaker Celebrity Bro
ad Fo cu se d
Syndicator Feeder Thought Leader Pundit
Dabbler Conversationalist Socializer Networker
Observer Explorer Activist Specialist
Casual Consistent
