This study assesses international corporate blogging practices and their effects on the popularity of corporate blogs in terms of comments received and incoming links. Building on the blogging practices framework by Schmidt (2007b), a theoretical model is developed incorporating cause–and–effects relationships for blog characteristics and their impact, as well as international differences. The five Hypotheses are tested using a sample of 20 German, 10 Russian and 77 U.S. corporate blogs. The results of stepwise regression analyses confirmed most hypotheses regarding effects of blog diversity, blog authenticity, blog usability, blog sophistication and networking efforts. The theoretical and practical implications of this are discussed.
1. Corporate blogging practices and effects: An introduction
2. Theoretical background: From blogging practices to blog acceptance
4. Analysis and results
5. Discussion and conclusion
1. Corporate blogging practices and effects: An introduction
The evolution of the Internet has had significant effects on corporate strategy considerations, as online communication has become a key element of companies’ communication strategies (Porter, 2001). Social media, such as blogs and social network sites, are becoming increasingly popular among Internet users and account for a quickly increasing share of the time individuals spend online (Nielsen, 2010). Social media is adding to the already growing range and diversity of information and communication channels. In such a complex communication environment, companies’ ability to communicate effectively with stakeholders has become a major competitive advantage (Argenti, et al., 2005). Individuals’ growing use of social media applications puts pressure on companies to implement a social media mix, seeking an opportunity to establish trust through direct communication to different stakeholders (Kotler, et al., 2010). Although social media allow firms to directly interact, i.e., with potential customers, at lower transaction cost and higher efficiency compared to traditional communication media (Kotler, et al., 2010), companies still feel uncomfortable engaging in social media owing to the lack of control over the communication process (Kaplan and Haenlein, 2010). This situation is aggravated by companies’ short–sightedness. Expecting quick successes in the fast developing social media environment, companies quickly become disappointed as reader numbers and reader feedback remain at low levels during early stages of their corporate blogs, and often decide to abandon corporate blogging activities (König, 2011). However, the formation of social relationships (which is at the heart of success in social media) takes time and thus companies need to adapt their expectations. A key to shaping expectations and appropriate strategies is for companies to understand the effect of the communication channels’ design characteristics and to anticipate users’ responses to new corporate communication offerings such as corporate blogs. With an increasing number of corporate blogging activities, it becomes possible to compare blogging practices and statistically analyze larger sets of corporate blogs. This form of benchmarking (rather than looking at individual best practices and how they achieved results within a specific communication situation) should enable researchers to provide more practical guidelines for corporate bloggers and allow them to change blogging practices in order to better reach desired communication goals. Therefore, this study aims at analyzing a large set of existing corporate blogs in order to determine factors which have an effect on how well corporate blogs are accepted by their audience. Furthermore, this study will assess differences in international blogging practices by comparing corporate blogs from three countries, i.e., Germany, Russia and the U.S. This will allow researchers to better identify and understand the effects of differing institutional frameworks on corporate blogging. 20 German, 10 Russian and 77 U.S. corporate blogs will be analyzed in order to identify key determinants of corporate blog popularity and compare blogging practices across the three subsets. Using this research companies should be able to shape more realistic expectations regarding corporate blogging activities and outcomes and enable them to eventually better tailor their activities to their blogging goals.
2. Theoretical background: From blogging practices to blog acceptance
Schmidt (2007b) has identified three dimensions of blogging activities that guide a blogger’s publication practices and therefore can be used as a ‘general model to analyze and compare different uses of the blog format’ . According to this framework (see Figure 1), a blogging episode is largely determined by sets of rules, code and relations, which form an interdependent and dynamic system. Transferred to the corporate blog context, it can be used to assess both corporate blogging practices and their effects on the audience.
Figure 1: An analytical model of blogging practices. Source: Schmidt (2007b, p. 1,419).
Blogging rules are the generalizable routines and procedures that govern the act of blogging. In the corporate blogging context they refer to the initial establishment of the corporate blog as well as the standard routines and processes connected to running the blog. These blogging rules comprise adequacy and procedural rules, that can be differentiated as either governing the decision to choose the corporate blog format as adequate for a specific corporate communication situation out of the large range of available media (adequacy rules), or as governing the actual usage of a corporate blog after the decision for this format has been made (procedural rules) (Schmidt, 2007b).
In the corporate blog context, adequacy rules first of all relate to the decision to establish a corporate blog (Schmidt, 2007b). This decision is usually tied to an assessment of the communication situation including the identified stakeholders and their characteristics, the company’s resources and capabilities, as well as the specifics of the available communication media (Smith, 2005).
Procedural rules in the corporate blogging context refer to the routines and procedures that govern the day–to–day operation of the corporate blog. In this context, Schmidt (2007b) differentiates three types of procedural rules, i.e., selection, publication and networking rules. Taken the establishment of a corporate blog as given, these rules can be adapted to better address certain target group needs. In addition, procedural rules for corporate blogs need to be differentiated by the involved actor, as they can be tracked to and conducted by either the company as the strategic initiator (usually the communications or marketing department as its responsible strategic department) or its individual employees as authors. Table 1 summarizes these different types of procedural rules and outlines the role of the company and the blogging employee.
Table 1: Overview of procedural rules. Type of procedural rule Employee/author’s role Company’s role Context Selection rules Reading, searching, considering company strategy Recruiting authors, determining overall purpose and strategy Company sets overall strategy and decides on goals and broad agenda; blogging employee assumes responsibility for detailed information search and selection Publication rules Writing, presenting Narrow agenda setting, determine blog policy and processes Company determines the detailed publication process and determines the specific topics to be covered; blogging employee produces the content and determines how it is presented Networking rules Representing, networking, connecting Determine relationship strategy and further usage of interaction results Company sets the specific goals for relationship building activities; blogging employee implements this strategy, represents the company towards the target group and is responsible for using feedback and forwarding inquiries
The second structural dimension of blogging practices is the code, i.e., the ‘blogging software and its underlying architecture’ . This is usually completely determined by the company (typically the corporate communication department in coordination with the IT department [Schabel, 2008]), deciding about the type of software used, the basic functionality as well as the connection to other communication activities of the company (e.g., the corporate Web site).
Last but not least, a blogging activity involves relations, including hyper–textual as well as social relations. In the corporate communication context one can furthermore identify internal relations, i.e., the relations between the corporate blog and other departments or employees of the company (König, 2011). Again, one can distinguish relations at the corporate and the individual level. In this context, the company needs to identify strategic decisions and incentives for the blogging employees to connect to the blog’s audience, i.e., establishing social relations. This goes beyond the networking rules discussed above by going beyond the single blogging episode and even the blog context. For the company, this also relates to the integration of blogging with other corporate communication activities, which can extend or contribute to the relationships with its target groups. Furthermore, both the company and the individual author are responsible for establishing hyper–textual relations through the use of hyperlinks and trackbacks. This enables a deeper integration with the blogosphere and to connect to other bloggers.
The above outlined aspects of blogging practices are in a range of ways interrelated. Procedural rules and adequacy rules affect the form and extent of the established relations by influencing ‘the size and composition of hyper–textual and social networks emerging from ongoing blog–based interactions’ . Therefore, how information is selected, processed and presented in a corporate blog eventually determines the type and size of the attracted audience and how the provided information is used by readers and other bloggers. Not only do rules determine the use of the software, but the software also limits rules or enables them to emerge. A company can only publish video or audio content if the software platform is capable of handling multimedia formats. Finally, code and relations influence each other as technical networking features influence how relations are established, e.g., by enabling comments, provide trackback technology or offer other more advanced networking function.
Consequently, this comprehensive framework of blogging practices can be used to structure to analyze practices and effects of corporate blogging. This study will therefore focus both on the individual elements as well as on the interrelationships. It will be assessing whether the outlined relationships, particularly the effects of rules and code on the evolving network (i.e., the size and activity of the audience or in other words popularity) also holds for the corporate blog context.
Several aspects and elements of private and corporate blogging practice have been assessed in theory and practice. However, most of the presented theory limits itself to theoretical evaluations or descriptive statistics, without specifically linking such blog characteristics to the performance of a weblog, which is a necessary basis for companies to evaluate the usage of corporate blogs as communication tools and to initiate corrective action. In order to analyze the determinants of corporate blog acceptance by Internet users, it is important to first define a set of blog characteristics. Due to the fact that the study has a very explorative character, it is crucial to observe a high number of characteristics. Fleck, et al. (2007) define the three dimensions authors, readers and format as important blog characteristics. In addition, Herring, et al. (2007) underline the importance of the posted texts itself as well as of general blog features, such as design elements and media usage. As a generalization, six main factors characterize a blog and can serve as a framework to analyze corporate blogs, i.e., general blog characteristics, general post characteristics, content characteristics, author characteristics, company characteristics, and reader characteristics. To illustrate the six sets of characteristics, the following table provides examples of more detailed characteristics:
Table 2: Overall weblog characteristics. Post characteristics Blog characteristics Post frequency Blog age Post variability Blog elements Average words Interactivity Author characteristics Content characteristics Number of authors Language and style Female/male authors Media usage Posts written by women/men Topics Company characteristics Reader characteristics Company size Gender Industry Purchasing behavior Communication goal Commenting behavior
For an empirical study of corporate blogs, the given model of a corporate blog as a combination of various characteristics needs to be simplified and the included parameters have to be limited to directly observable blog characteristics. Several previous studies of private weblogs, e.g., Herring, et al. (2004a) and Nardi, et al. (2004), have focused on author characteristics such as gender, age or author goals. However, these studies’ findings were not related to a blog’s performance. Communication theory assumes that in the case when communicators are directly observable by the audience, the audience tends to be similar in terms of demographic characteristics and attitudes and communication effectiveness increases with higher communicator–audience similarity (e.g., Berscheid, 1966; Worchel, et al., 2006), which has also been shown for corporate communication and marketing contexts (e.g., Brock, 1965; Woodside and Davenport, 1974). Consequently, a diverse group of communicators, representing a larger set of backgrounds and attitudes, is also able to attract a diverse and eventually larger audience. In the corporate blog context this would imply that a larger group of authors from different backgrounds, gender and age groups will be able to attract a bigger audience than a very homogeneous group of communicators. Consequently, it can be assumed that a more diverse blog can better address different reader groups. Hence, it is hypothesized that: Hypothesis 1: A higher (lower) diversity will yield higher (lower) blog popularity.
Content characteristics are often cited as key factors to attract readers and built reader loyalty (e.g., Picot and Fischer, 2006; Zerfaß and Boelter, 2005). For corporate blogs, it is often stated that it is crucial to create authenticity and credibility to overcome a general skepticism of consumers towards corporate communication (e.g., Röttger and Zielmann, 2006; Yang and Lim, 2009), a topic which is also prevalent in standard corporate communication theory (e.g., Argenti, 2007; Cornelissen, 2011). In this context, Zerfaß and Sandhu (2005) have established the notion of ‘virtual authenticity’ that can be fostered by engaging in social media, e.g., by establishing a personal CEO–blog or other employee blogs. Consequently it can be hypothesized that: Hypothesis 2: A higher (lower) authenticity will yield higher (lower) blog popularity.
Another stream of research has focused on post–related characteristics. This includes tools that can increase the attractiveness of the posts by adapting posting practices. This comprises for example a high post frequency or the use of media to support written text, as factors that improve reader comfort (e.g., Schmidt, 2007a, 2007c). Hence, it is hypothesized that: Hypothesis 3: A more (less) reader–friendly blog management will yield higher (lower) blog popularity.
Scheidt and Wright (2004) have identified a tendency to adapt sidebar elements to better suit publishers’ needs and meet readers’ expectations. In general, overall blog characteristics, such as blog age, design elements and interactivity may play an important role with respect to blog popularity. Therefore it is hypothesized that: Hypothesis 4: A more (less) developed blog will yield a higher (lower) blog popularity.
Blogs as tools for communication and interaction are part of the new, more people–centered Internet. As it may be characterized as a huge network of linked and interconnected participants and services, a blog publisher’s networking effort may play an important role for establishing a corporate blog (e.g., Cass, et al., 2005; Kaplan and Haenlein, 2010). This effort may include linking to other content, discussing issues of common interest and actively commenting on other blogs, contributing to a more visible and more interesting blog (Ali–Hasan and Adamic, 2007). Therefore, it may be hypothesized that: Hypothesis 5: A higher (lower) networking effort will yield higher (lower) blog popularity.
A second goal of this study is the identification of international differences in corporate blogging. To this end, it is first of all important to understanding how a corporate blog as a corporate activity is subject to its institutional environment According to the new institutional economics (NIE) framework, corporate strategies and decisions are affected by formal institutional factors, including political, social and legal rules, and informal institutional factors relating to culture. These cultural factors are more difficult to identify than formal factors, owing to their intangible nature. The institutional framework also affects the behavior and attitude of individuals, e.g., by influencing a person’s values and preferences. Subsequently, these individuals are also related to a company’s decision making that evolve from their function as employees, customers or other types of stakeholders.
In the context of the blogosphere, the NIE framework can be used to illustrate determinants of corporate blogging practices and the impact of these practices. The formal and informal institutional framework influences or even constraints decisions related to corporate blogging — e.g., Who is allowed to blog? What is the blog’s structure and content in the context of binding legal rules for corporate publications? At the same time it influences readers’ perceptions, values and preferences, which in turn guide the blog–related decisions of the company. Figure 2 presents a corporate blog as unique combination of blogging practices and effects, as introduced at the beginning of this section, which is influenced by formal and informal institutional framework elements as well as individual characteristics, as outlined by Williamson (1996). It implies that on the one hand the institutional setting, e.g., laws and regulations that restrict corporate communication may affect corporate blogging practices, while on the other hand individual characteristics such as personal values of authors and readers also affect the way a corporate blog is performing. The inquiry of the U.S. Security and Exchange Commission with respect to Whole Foods Market, Inc. and the publication and use of data on online platforms such as the blog of its CEO John Mackey, which led to a ten–month suspension of blogging activities, may serve as an extreme example for the impact of the formal institutional framework (Mackey, 2008, 2007). On the other hand, the framework also outlines that corporate blogs may serve companies which are willing to influence their environment. As a tool to improve reputation, it may influence opinions and attitudes of individuals. At the same time, corporate blogs may be used for lobbying and other strategic activities influencing the institutional framework.
Figure 2: A corporate blog within the institutional framework. Source: Author, based on Williamson (1996).
Realizing that companies and employees may use blogs to bridge geographical and institutional boundaries, e.g., by providing a global blogging platform for all employees, by approaching an international audience using a single channel, or by establishing a range of corporate blogs for varying national and international audiences, one can even model corporate blogs as a corporate activity which is subject to different institutional frameworks as proposed by Wolff, et al. (2004). In this context, a company may face for example differing legal frameworks regarding corporate communication and blogging activities, different audience structures as well as differing sets of individual values. In differing contexts, audiences might value different aspects or characteristics of corporate blogs. Consequently, Trammell, et al.  argue that under the influence of varying institutional frameworks ‘different cultural patterns of blogging, and of Internet use in general, will develop over time’. From a global communicator’s perspective, we today deal with Internet audiences and blogospheres that feature significant differences between individual countries or — broader speaking — language areas, as explored by Kelly (2009a, 2009b, 2008; Etling, et al., 2010), who identified significant differences in the structure, content and members of the blogospheres in Germany, Russia and the U.S. Modeling international corporate blogging activities as being subject to different institutional frameworks yields two specific research areas within the framework of corporate blogging practices. Firstly, with respect to the individual elements (rules, code and network), institutional framework effects would result in different blogging practices under different frameworks, i.e., a corporate blog and its characteristics may be the results of formal and informal institutional frameworks, individuals’ characteristics as well as specific corporate blogging governance. Secondly, the interrelationships between the elements might be affected, implying that the size and type of interrelationships such as the effect of corporate blogging rules or software on the size and type of audience may differs between individual frameworks.
Sample and data collection procedure
To evaluate corporate blogging practices, a list of public corporate blogs was compiled. Research focused on three countries: Germany, Russian and the U.S. The three countries were chosen based on their relevance to multinational enterprises, the relevance of social media in these countries, and the size of the population that is using the Internet. The U.S. were selected as the key country for international management research. The country features the highest level of social media penetration in terms of overall status and role for the population (e.g., see the importance of social media for politics and elections), while being home to the majority of social media platforms such as Facebook (Palo Alto, Calif.), YouTube (Mountain View, Calif.) or Twitter (San Francisco, Calif.), and some of the most important blogging platforms including Wordpress (San Francisco, Calif.), Blogger (Mountain View, Calif.) or Typepad (San Francisco, Calif.). Germany was included as the largest member and market within the European Union. Furthermore it has a high level of Internet penetration and substantial overall social media adoption among consumers as well as businesses. Last but not least, Russia has been chosen owing to its importance to international business as an emerging market and the high importance of social media in the Russian society, as Russia has the most engaged social networking audience worldwide (Block, 2010).
All corporate blogs from the selected three countries included in the study had to meet pre–defined criteria to ensure comparability. They had to be set up by medium–size or large companies, i.e., companies with revenues exceeding €100m in 2007. They had to be active, i.e., feature at least one new post during the time period between 7 January and 16 March 2008. They also had to be older than 10 weeks on 16 March 2008. Using an extensive research process, including a large selection of literature on corporate blogs including Weil (2006), several Internet databases on corporate social media activities such as Anderson (2009), as well as an additional Internet search, it was found that 77 English, 20 German and 10 Russian corporate blogs fulfilled these three criteria. The resulting list is provided in Appendix 1.
The complete content as well as the status of all 107 corporate blogs were archived on 17 March 2008, using the software tool ScrapBook (Gomita, 2012), which allows to extract and archive more complex Web site networks. The available data of each individual corporate blog included data at the blog level such as name, design and features, as well as at the individual blog entry level, including the publication date, text, author(s), supporting media content and associated comments. Consequently, the whole dataset comprised 107 corporate blogs with in total almost 700 authors, more than 2,500 blog entries and more than 13,000 comments:
Table 3: : Overview corporate blog sample. G R U.S. Total Blogs 20 10 77 107 Blog posts 467 353 1,797 2,617 Blog authors 103 87 499 689 Comments 1,868 1,911 9,341 13,120
Measures — Dependent variables
While no single measure (or even definition) of corporate blogging success exists owing to the diverse corporate blogging landscape and the variety of communication strategies and objectives, several benefits of corporate blogs have been identified that can be used as objectives of or indicators for corporate blogging strategies. For external corporate blogs, Forrester identifies the benefits to comprise brand visibility and other benefits (Li and Stromberg, 2007; Li and McHarg, 2007). As summarized in Figure 3 below, brand visibility effects include the number of visitors (blog traffic), the number of blog stories that entered traditional media (press mentions), an improved ranking in search engines as well as the number of mentions and discussions in other social media such as social network or private blogs (word–of–mouth). These benefits result in a more positive awareness among target audiences (Chiou and Cheng, 2003) and through a higher visibility online can have an indirect effect on sales (Dellarocas, et al., 2007; Zhu and Zhang, 2010). Other key benefits of corporate blogs include savings on customer insights through customer feedback for example provided in comments on a corporate blog, the reduction of negative user–generated content by positively influencing online discussions, and an increased sales efficiency as employees as well as customers are better informed about the company and its offerings.
Figure 3: Benefits of corporate blogging. Source: Based on Li and Stromberg (2007).
As this study seeks to relate corporate blog characteristics to corporate blog benefits, the selection of dependent variables for this benchmarking study had to focus on indicators that were widely observable for all selected corporate blogs. Consequently indicators such as actual traffic or complex indicators such as reduced impact of negative user–generated content or sales efficiency had to be neglected, as the necessary information would only be available to the individual companies and was in general not available for sharing even in an anonymized academic context. However, two indicators for word–of–mouth and feedback were available as will be outlined shortly.
Word–of–mouth on the Internet initiated through a corporate blog is expressed by the amount of other Web sites picking up discussions and consequently linking to the respective corporate blog, including mentions on social networks or discussions on other weblogs. A common tool to make multiplication throughout the blogosphere visible is the Technorati Authority (http://technorati.com/what-is-technorati-authority/), which measures the number of other blogs linking to one specific blog and its content. It is based on Technorati’s ability — as the largest search engine particularly indexing the blogosphere — to continuously monitor changes in the blogosphere and to provide a real–time picture of associations between blogs. The higher the number of incoming links (restricted to links created within the past six months to better account for frequent changes) the higher the Authority assigned by Technorati. Other estimators try to evaluate the general importance of a Web site within the overall Web content, i.e., Google Page Rank and Alexa ranking. Pages that are frequently updated and receive links from other Web sites are rated higher than static sites that go unnoticed. However, both feature significant drawbacks. Optimized for general Web content and not blogs in particular, both are not frequently updated, feature significant biases towards or against certain language areas, are not available for every blog in the sample and therefore may reflect an inaccurate ranking, while a scaled logarithmic computation adds further distortion (Kirchhoff, et al., 2007). Therefore, this study focuses on Technorati Authority as index–based indicator for word–of–mouth.
A second approach to estimating blog popularity is to evaluate the level of interaction with readers, especially if customer or other stakeholder feedback is an important goal of the blogging company. As outlined by Li and Stromberg (2007), feedback collected through a corporate blog takes the form of comments by readers below the individual blog entries. Consequently, the amount of feedback, in particular the number of comments received on a blog as well as the average length can serve as an approximation for how well a company is able to engage in a discussion with the intended target group. Therefore, the average number of comments per post, which can be calculated from the archived data set, shall serve as the indicator for collected feedback.
Measures — Independent variables
To estimate the individual characteristics and factors, a set of observable blog parameters had to be defined. Herring, et al. (2006, 2004b) selected various author characteristics, text statistics and the use of media and Internet links to describe blogging practices. Scheidt and Wright (2004) focused on visual design elements, such as sidebar elements used in private blogs. Lee, et al. (2008) and Lee, et al. (2006) highlighted the importance of author characteristics for corporate blogs. Fleck, et al. (2007) outlined the role of content, especially the topics covered in blog posts. Turck (2007) evaluated corporate blogs’ authenticity by rating blogs based on their writing style. In order to assess the formality and indicate the degree of authenticity of blog posts, Puschmann (2007a, 2007b) has identified grammar statistics of blog posts as meaningful characteristic. As a conclusion, the following list gives an overview of often used and observable corporate blog parameters:
Table 4: Observable weblog parameters. Blog age/Blog start Number of interactive elements Comment length Number of standard sidebar elements Comments per post Number of total authors Gender of authors Post length Grammar and choice of words Posts by gender Media use per post Posts by week Number of blogroll elements Topics Number of female/male authors Total links
The empirical study will focus on eleven parameters to estimate blog popularity:
Number of authors — This independent variable summarizes the observable number of authors that was engaged with a blog over the pre–defined time horizon. As more authors increase a blog’s diversity and better address varying reader preferences, it can be hypothesized that a higher number of involved authors can yield higher blog popularity, as stated by Hypothesis 1.
Gender of authors — The gender of authors may have an impact on visitor demographics. Therefore, it shall be determined whether a more uniform author mix contributes to a higher popularity. In this context gender disparity represents by how much an author group varies from an assumed parity of female and male authors, i.e., the absolute value of the difference of 0.5 and the observable share of female authors. A more equal representation of gender in an author group may positively affect the diversity and hence contribute to a blog’s popularity as expressed in Hypothesis 1.
Topical dimensions — This independent variable reflects the content of a corporate blog. For this purpose the set of archived blog texts had to be transformed into sets of quantitative data. This was done based on the transformation process proposed by Srnka and Koeszegi (2007). This process enables researchers to convert qualitative elements into statistically analyzable data and follows five steps: (1) material sourcing; (2) transcription; (3) unitization; (4) categorization; and, (5) coding. The sourcing process included the collection of blogs and the archiving of the content. As a result of the archiving as described above, all blogs including the text and design elements were available in html. As second step, the transcription was performed by extracting the individual blog posts. The whole sample for this procedure thus included 1,797 English, 467 German, and 353 Russian posts, i.e., 2,617 posts in total. At the unitization stage, it was decided to analyze the texts on a single post basis, i.e., every post was analyzed individually. In the fourth step, a category scheme was developed, allowing categorizing every blog post with respect to the topics covered by the text. The aim was to explicitly correlate each blog post to a single topic category. The categorization scheme differentiated four major areas, namely company, industry, blog–related and other topics, with several sub–topics identified by screening the German sample of corporate blogs for covered topics. The following list gives an overview of the whole category scheme:
Table 5: Corporate weblog content categories. Company Administration (CA): Company official and executive information, stock information Career/recruiting (CHR): Job experiences, career hints, HR department news Corporate social responsibility (CSR): CSR activities Events (CE): Events arranged by the company History (CH): Corporate history Marketing (CM): Advertisements, marketing department news, studies Operations (CO): Processes, production, production topics Products (CP): Product tests, specifications, introductions Strategy (CS): Corporate strategy, goals, targets, profit/revenue Technology (CT): Technological issues Industry Development/trends (ID): Industry trends Events (IE): Industry events Products (IP): Products of competitors Technology (IT): Technology and new developments by competitors in the industry Other General news (OG) Private/employee life (OP) Society (OS) Blogging Direct blog–related (BD): Direct relation to the corporate blog General blog–related (BG): General blogging–related issues
Finally, the coding process was conducted by two independent coders. Each coder had the task to evaluate and categorize every single blog post with respect to the above detailed classification scheme. As a consequence, every blog post received one classification tag from each of the coders. After conclusion, intercoder reliability was checked using Cohen’s Kappa. Because the high Kappa values indicate a high level of agreement among the coders (0.916 for the U.S. sample, 0.910 for the German sample, 0.892 for the Russian sample), and both coding schemes were combined into a single scheme with two tags for every individual blog post and then used to further analyze each corporate blog individually. The actual value used for the stepwise regression analyses assigned a level of covered topics to each corporate blog individually, being the number of topics that accounted for five percent or more of the total topics covered by the respective blog. Being an indicator for diversity, a positive relationship between this factor and a blog’s popularity can be assumed as expressed by Hypothesis 1.
Formality — Literature on blogs often highlights authenticity as a major determinant of corporate blog popularity. Readers will only become loyal to a blog if the blog character differs from the traditional marketing and communication mix (Zerfaß and Boelter, 2005). Heylighen and Dewaele (2002, 1999) have developed a structured approach to assess the contextuality of language, which can be interpreted as a unifying measure for language and character of text. To this end, they distinguish between formal communication, which ‘conveys information explicitly, through the linguistic expression itself’ , and contextual communication, where the information is only conveyed implicitly through the context of the expression. To evaluate if texts are written rather formal or informal, all words used are grouped into different classes, i.e., nouns, adjectives, prepositions, determiners, pronouns, verbs, adverbs, interjections and conjunctions. Nouns, adjectives, prepositions and determiners have been identified as indicators for formality, while an increased use of pronouns, verbs, adverbs and interjections characterizes contextual expression. A measure of formality is introduced by calculating the difference between the frequencies of formal text elements and informal text elements, i.e., the independent variable F–score (Heylighen and Dewaele, 2002).
Puschmann (2007a, 2007b) first used the F–score to measure the authenticity of selected private and corporate blogs. A similar approach will be used for this study. To derive the frequencies of the different word classes for every single post included in the study, the software TreeTagger, developed by the University of Stuttgart to analyze language (see http://www.ims.uni-stuttgart.de/projekte/corplex/TreeTagger/), was used. This software tool processes an entered text by first partitioning it into single word elements and then analyzing each word element using an underlying language–specific tagset, which allows the software to systematically assign a word category tag. For the application of the TreeTagger to the three selected language groups the Stuttgart–Tübingen–Tagset (STTS) was used for German text (Schiller, et al., 1999), the Penn Treebank Project tagset was used for English text (Santorini, 1990) and the tagset for the Russian National Corpus was used for Russian text (Sharoff, et al., 2008). As basis for the tagging process, it was necessary to create an individual corpus for each of the blogs in all three samples, yielding 107 different blog corpora. Then, each corpus was processed using the above outlined software and tagsets. For each blog in the sample the frequencies of the word classes were calculated individually and an F–score was computed.
In a side step, F–scores were compared to a paired set of press release published by the individual companies, showing that corporate blogs are overall less formal (and thus for the purpose of this analysis more authentic) (König, 2009b). Consequently, F–scores were then used to estimate the effect of formality on corporate blogging popularity. Included in the stepwise regression analyses, it shall be evaluated whether less formal blogs are indeed more popular, as commonly assumed in literature and as stated in Hypothesis 2.
Post frequency — The independent variable post frequency expresses how many blog posts were published on average during the ten–week observation horizon of the study. It is calculated by dividing the total number of blog entries posted on the particular blog by ten to yield an average number of blog entries published per week. Because blogs that rarely publish new content are considered less attractive for readers and in general feature less content, a positive relation between post frequency and blog popularity is assumed as expressed in Hypothesis 3.
Media usage — The use of media elements in blogs has become more important as the technological capabilities of the various blog services evolved and is considered a determinant of traffic to a blog and driver of interactivity (Kaplan and Haenlein, 2010). For this study different media formats in addition to text, namely picture, video and other formats such as polls or slideshows were evaluated. The occurrences of such elements in all blog posts were counted, and consequently the independent variable media usage per post was calculated by dividing the total number of media elements used in a corporate blog over the observed time horizon by the total number of posts published. Contributing to a more comfortable reading experience, a blogs’ popularity is assumed to be positively dependent on the use of media as stated by Hypothesis 3.
Post length — The independent variable post length was calculated as the average number of words of all posts published on an individual blog. Based on the assumption that longer blog posts have a higher potential to foster reader engagement and can transfer more information, a positive relation with respect to a blog’s performance is assumed, as implied by Hypothesis 3.
Blog age — Social networks build over time. Therefore, age of the blog can be estimated as having a positive impact on blog popularity. Conducting a time series analysis for a single corporate blog, König (2009a) showed a positive long–term effect of age on the number of incoming links, reasoning that older blogs can be expected to have more readers due to reader loyalty and have already had more time to make modifications to their corporate blogging practices based on audience feedback (see Figure 4 for an illustration). In addition, blogs that do not meet the expectations are more likely to be abandoned at early stages. The variable blog age is therefore included as an independent variable, being calculated by subtracting the date of the first post on a blog from the last day of the observed time horizon, i.e., 16 March 2008. Age is assumed to be positively related to popularity as expressed by Hypothesis 4.
Figure 4: The corporate weblog life cycle (Daimler–Blog). Source: König (2009a).
Design complexity — Scheidt and Wright (2004) conducted a comprehensive empirical study of design elements used in private blogs, concluding that sidebar elements were the blog design feature that was most often adapted to suit blog needs. Sidebar elements allow readers to quickly navigate through a blog, e.g., through an archive element or lists of latest posts and comments. As a second aspect, blog publishers themselves can use the sidebar to provide additional information for readers, e.g., by providing a list of often read weblogs (blogroll), highlighting important or popular blog posts, offering subscription options (e.g., RSS and e–mail) or simply giving selected information about the author(s). The following list of sidebar elements was developed by screening the previously introduced literature on blogging practices as well as the sample of German corporate blogs:
Table 6: Sidebar elements in corporate weblogs. Post archive Photo archive Search Author list & information Latest comments Subscribe (RSS, e–mail) Blog roll Latest posts Tag cloud Category list Latest trackbacks Top commentators Classics/hot picks Most commented posts Dates Most viewed posts
Consequently the independent variable number of sidebar elements was calculated by counting the number of elements used out of the above list. Based on the fact that a lack of sidebar and navigation elements makes reading the respective blog uncomfortable for readers who wish to gain additional information and further readings, it is assumed that a larger number of sidebar elements results in higher acceptance of corporate blogs as expressed in Hypothesis 4.
Interactivity — Since blogs evolved as social media, the commenting behavior of blog visitors has become significantly more intense (Rainie, 2005). At the same time, it became important for blog publishers to offer commenting functions and other interactive features, such as the opportunity to recommend blog posts via various social networks, e.g., delicio.us (http://delicious.com/) and digg it (http://digg.com/) as English language and Mister Wong (http://www.mister-wong.de/) and Y!GG (http://yigg.de/) as German language Internet services. Consequently, the independent variable interactivity elements was calculated by summing up the number of available activities out of five standard features, namely comment function, e–mail recommendation, rating mechanism, social network recommendation and print function, which usually are provided directly below or next to an individual post. Hence, it can be hypothesized that blog managers can positively affect a blog’s performance by providing more interactivity elements, as stated by Hypothesis 4.
Networking — Technorati Authority measures the number of incoming links, i.e., links from other blogs to the main page of a blog or individual blog posts. With the blogosphere being an interactive and interlinked network, it can be assumed that a high activity to connect with other blogs and Web sites is rewarded with a more intense interaction and a higher number of backlinks. This phenomenon of reciprocity in the blogosphere was observed by a range of previous studies, the most extensive one being the study by Gaudeul, et al. (2009). While most networking activities of bloggers are hard to be observed (such as attending blogger meetings, exchanging e–mail messages or communicating through other social networks such as Twitter or Facebook), a publicly observable indicator for a blogger’s networking activities is the blogroll. The blogroll is a specific sidebar element which represents a list of favorite Web sites or blogs respectively and creates a linking tool within the blogosphere (Picot and Fischer, 2006, Schmidt, 2007c). Hence, the number of links included in a blog’s blogroll is used as an indicator for the networking activities of its publisher, summarized by the independent variable number of blogroll elements. In this context, it is assumed that a more intense networking, i.e., a higher number of blogs placed in the blogroll, has a positive effect on the blog popularity (Marlow, 2004) as stated in Hypothesis 5.
4. Analysis and results
The analysis comprised two steps to identify differences in blogging practices between the samples and effects of blogging practices on blog acceptance. First, analyses of variance (ANOVA) were be conducted to detect significant differences in the means of the independent variables over all three sample sets. Unpaired t–tests revealed significant mean differences that may be caused by differences in the institutional frameworks and/or by varying blogging practices. Facing small sample sizes with respect to the German and Russian sample set, explorative stepwise regression analyses were employed to analyze the data. The used stepwise regression method was forward selection to identify the variables that are included or excluded. It will was conducted for each sample set individually as well as the overall sample set, in order to examine the effect of the outlined independent variables on the popularity indicators. The software used for both the analyses of variance and the stepwise regression analyses was IBM SPSS 19.
Analysis of variance
Appendix 2 presents means and standard deviations for the independent and dependent variables over all three sample sets. The following table shows the ANOVA results, highlighting statistical differences in the means of the independent variables, which may indicate effects of the institutional frameworks on blogging practices:
Table 7: Results of analysis of variance. Variable U.S./G U.S./R G/R Topical dimensions — — — Post frequency — — — Number of authors — — R > G
(p = 0.07)
Blog age — — — Post length U.S. > G
(p < 0.01)
U.S. > R
(p = 0.02)
— Gender disparity — — G > R
(p = 0.05)
Formality (F–score) U.S. > G
(p < 0.01)
R > U.S.
(p < 0.01)
R > G
(p < 0.01)
Number of sidebar elements — U.S. > R
(p < 0.01)
G > R
(p < 0.01)
Number of blogroll elements — — — Media usage per post — — — Interactive elements — — —
Facing small sample sizes with respect to the German (N = 20) and Russian (N = 10) corporate blogs, few statistically significant differences exist. Topical dimensions, post frequency, the number of blogroll and interactivity elements as well as the use of media show no significant mean differences.
Russian corporate blogs have the highest average number of authors, resulting in a statistically significant difference with respect to German corporate blogs. The average number of authors of English corporate blogs is situated in between these values but shows no significant difference in both directions. The fact that Russian corporate blogs feature very large author groups compared to the other samples may be explained by acknowledging the more collectivist character of the Russian culture. Hofstede (1980) for example assigns a higher level of individualism to the U.S. (91) and Germany (67) compared to Russia (39), where individualism measures the extent to which people are willing to act on their own behalf in contrast to a preferred group or community membership. Relating the ANOVA findings to the impact of culture, Russians, who are less driven to act individually but rather engage in groups and communities, might tend to establishing larger author groups to run corporate blogs.
Post length is the largest in English language corporate blogs. German and Russian blog posts are significantly shorter. To some extent this may be due to general language characteristics, as e.g., identical texts might differ in length when translated into various languages. However, as clear inter–language text length patterns have not yet been found by linguistics researchers and discussions on that often produce contradicting results (e.g., Quinn, 2010), the identified differences are assumed to be due to different blogging practices.
The average gender disparity, i.e., the degree of difference between the number of male and female authors, is highest in German corporate blogs and smallest in the Russian sample set. Also statistically, the gender distribution is significantly closer to parity in Russian blogs than in German blogs, while English corporate blogs rank in–between. This significant difference may be due to institutional framework effects. Hofstede (1980) has identified gender roles as a main determinant of culture, by evaluating male and female values and perceptions. As a consequence, masculinity/femininity can affect corporate governance. His study revealed that Germany has a higher masculinity score (66) than the U.S. (62), which in turn features a higher score than Russia (36). Interestingly, the share of men in corporate blog authors mirrors the individual masculinity scores as shown in Figure 5.
Figure 5: Masculinity of culture vs. share of men in corporate weblogs.
The calculated F–scores show the most significant pattern of mean differences among all observed parameters. German blog posts feature significantly smaller F–scores than English and Russian blog posts, while at the same time English posts feature significantly lower F–scores compared to Russian posts. This pattern cannot directly be interpreted as differences in blogging practices with respect to formal or informal language, but rather shows linguistic differences between the three languages. Therefore, these differences are best described as an institutional framework effect, as language is generally considered a part of the institutional framework in a given environment affecting the formation of culture (Huo and Steers, 1993). In terms of the F–score, the Russian language may be characterized as a more formal language, using a higher percentage of formal word types such as nouns and adjectives compared to the German and English language (Paul, et al., 2011). Hence, a direct comparison of the observed F–scores in terms of formality between the three language groups for the purpose of this study is not appropriate. However, the calculated F–scores are suitable for the analysis of each sample individually.
Stepwise regression analyses
The following table presents the results of the stepwise regressions analyses.
Table 8: Results of regression analysis.
Note: a. For the combined regression analyses, F–score as a parameter was excluded because of linguistic considerations.
*** p < .0.01 , ** p < 0.05, * p < 0.10.
Variables Technorati Authority Comments per post G R U.S. Alla G R U.S. Alla Intercept -29.239***
Topical dimensions 2.972*
Out Out Out -4.918***
Post frequency Out Out -63.546*
Out Out Out Out Number of authors 0.952*
Out Out Out Out Blog age 0.018***
Out Out 0.007***
Post length 0.058**
Out Out Out -0.055*
Gender disparity Out Out 716.155*
Out Out Out Out F–score Out Out Out — Out -0.482***
Out — Number of sidebar elements Out Out -80.843*
Out Out 3.231***
Number of blogroll elements 0.231*
Out Out 0.187***
Out Media per post 31.932***
Out Out Out -27.75**
Out Interactive elements Out 7.123*
Out Out Out Out 1.371**
Out Adjusted R2 0.795 0.850 0.581 0.461 0.484 0.603 0.403 0.940 F 19.37*** 26.41*** 18.58*** 19.47*** 6.942*** 7.83** 11.264*** 4.728*** N 20 10 77 107 20 10 77 107
For both dependent variables significant estimator models could be found with adjusted R2 ranging between 0.403 and 0.940. The following section will highlight selected statistical findings as well as the results of the regression analyses with respect to the individual independent parameters.
Number of authors — The stepwise regression analyses yield a significant positive relation between the number of authors and a blog’s Technorati Authority. The higher the number of authors of a corporate blog, the higher is the expected Technorati Authority. The impact is significant for the German, the U.S. and the total sample. With respect to comment frequency, a significant positive relation was not found for any of the samples.
Gender of authors — In all three samples, there is a clear dominance of male post authors. This underrepresentation of women in corporate blogs is existent for both, authors and individual blog posts as depicted in Table 9. Only in the Russian sample, women are more equally represented compared to men, as already outlined by the significantly higher gender parity compared to German corporate blogs. The following table presents the gender distribution with respect to the observed individual authors but also for the number of published posts.
Table 9: Representation of women and men in corporate weblogs. Sample Authors Share of posts written by Male Female Unknown Male Female Unknown G 70.87% 24.27% 4.85% 61.67% 18.42% 19.91% R 50.50% 32.67% 16.83% 36.18% 43.87% 19.94% U.S. 68.20% 29.00% 2.80% 63.15% 23.24% 13.62%
This statistical result is surprising, given a more uniform gender–related distribution within the weblog community. However, the W3B study by Fittkau and Maaß (2008) found that also corporate reader characteristics showed significant discrepancies from general blog reader demographics. While approximately 60 percent of all German weblog readers are male, this number increases to 70 percent if the analysis is restricted to corporate blogs. Therefore, one can interpret the statistical findings as mirror image of the audience.
The stepwise regression analyses revealed a significant positive relation between gender disparity and Technorati Authority for the U.S. as well as the combined sample, implying that blogs may be more popular when they feature a higher disparity. This may be explained by reader demographics. A high share of men in blog readers may be better addressed by male blog authors. On the other hand the more technical character of many blogs, coming from an automotive industry or Internet background may explain this observation as well.
Topical dimensions — The number of topics covered on a corporate blog was found to be positively related to Technorati Authority for the German sample. A different picture emerges with respect to comments per post, where a significant negative impact was identified for all three country samples and the total sample. This implies that the overall popularity (in this case incoming links) increases with an increase in topics covered on a corporate blog, while this would decrease the intensity of discussions (in terms of average number of comments per post).
Formality — Only a significant negative effect was found for the Russian sample, where the F–score is negatively related to the average number of comments per post. This implies that less formal blogs can achieve a higher comment frequency in the Russian blogosphere, which is reasonable given the fact that comments and feedback are indicators for conversations, which may be fostered by a more personal writing style and tone.
Post frequency — The average number of published posts per week was not found important for a blog’s acceptance. Only the U.S. sample shows a negative dependence of Technorati Authority on post frequency with low significance.
Media usage — The stepwise regression analyses present varying impacts of media usage on blog popularity. Most significant is a positive impact with respect to the German sample and Technorati Authority, where the resulting model predicts an increase of popularity by 31.9 for an increase in average media usage per post of one, which is large, given an average Technorati Authority of 20.60. A significant positive effect is also estimated for the English sample and comment frequency, while other stepwise regressions estimate negative impacts of media usage on the number of comments per post. This might signal two–fold effects of media usage. On the one hand, media usage can have a positive effect on overall popularity, attracting more readers and incoming links. However, larger media usage might also be seen to appeal to readers’ entertainment desires, resulting in a lower commenting activity on blogs with higher media usage.
Post length — The stepwise regression analysis for the German sample has identified a small but significant effect of the number of average words in blog posts on the blog’s Technorati Authority, implying German blog visitors prefer longer post texts over shorter ones. For every increase in the number of average words, Technorati Authority is expected to increase by 0.054. Hence, an increase by 55 words would result in a popularity increase of ‘3’. Facing an average Technorati Authority of 20.60, this would represent an increase of 14.6 percent. In contrast, the impact of post length on comment frequency is small but negative for the German, the Russian and the overall sample, implying that shorter blog posts better foster a conversation.
Blog age — The stepwise regression analyses identify a significant positive effect of blog age on Technorati Authority as well as comment frequency, implying that popularity and the intensity of conversations increase as corporate blogs grow older. This finding may have several reasons. A mature corporate blog has had more time to build up a loyal readership than a relatively new blog. Also, corporate blog editors can adapt the blog over time to better meet reader requirements, e.g., through changing blog design and contents.
Design complexity — The number of sidebar elements seems especially important for a blog’s comment frequency, as expressed by a positive significant relation for the total as well as the German and the Russian sample. A clear impact on Technorati Authority could not be observed, with only a slightly significant negative effect for the U.S. sample. This finding implies that a blog can particularly foster conversations by improving the blog’s infrastructure, while the effect on overall popularity is small. This observation may be explained by the fact that readers who comment are more willing to look through an archive or for author information, e.g., because they are interested in texts with similar topics or by the same author.
Interactivity — Interactivity, as expressed by the number of interactivity elements, is crucial for conversations on a blog, as shown by the estimation for comment frequency. It enables readers to actively use a text, e.g., in order to comment or to recommend it. Only slightly positive impacts of this factor could be identified for the Russian sample (for Technorati Authority) and the U.S. sample (for comment frequency). This may be explained to some extent by the small variance of this factor. Only two blogs in the total sample do not provide an interactive feature, while all others (105) at least allow readers to comment. Interactivity can thus be regarded a standard feature, which therefore has little impact on a blog’s acceptance.
Networking — A clear and significant effect of the number of blogroll elements on Technorati Authority was identified, with a positive impact for all four analyzed samples. For commenting such effect could only be identified for the U.S. sample. This supports the finding of several researchers that popularity in terms of linking within the blogosphere is largely based on pro–activity and reciprocity. Blogs that proactively try to link to other blogs are likely to received more incoming links in return (e.g., Gaudeul, et al., 2009). In this very same context, however, the question of causality arises. It remains beyond the scope of this research to identify, what role the observed blogroll elements play within the reciprocity phenomenon. It needs to be stated that a link to another blog within the blogroll can be both the cause of an incoming link or the reaction to an incoming link. Individual case studies evidence shows that companies carefully and strategically select blogroll elements and outgoing links in general, which would tend to support the idea of outgoing links to cause inbound links (e.g., Schabel, 2008; Wilke, 2009). However, there has not yet been a thorough investigation of this phenomenon for corporate blogs.
5. Discussion and conclusion
The observed effects of the independent variables may be used to evaluate the Hypotheses, which were introduced earlier in this paper in Section 2.
Hypothesis 1: Corporate blog diversity
Three aspects of corporate blog diversity were observed, i.e., the number of authors, gender diversity and topic diversity. The study has identified a positive relation between the number of authors and a blog’s popularity, which may indicate the positive effect of diversity. However, gender parity as another diversity factor could not be found to positively affect blog popularity and may even have a negative effect, which may be caused by particular reader demographics as outlined above. With respect to topical dimensions as diversity indicator, it needs to be differentiated whether a blog aims at becoming more popular in general or whether reader feedback should be fostered. To a limited extent a higher number of topical dimensions contributes to a blog’s popularity as it addresses a broader audience. However, the opposite effect can be observed when estimating comment frequency. Here, a limitation to selected topic categories may be beneficial. Though this may limit or decrease the number of readers, it contributes to the establishment of a somewhat specialized community around an issue and therefore helps initiating a conversation within this community. Hence, Hypothesis 1 can be verified only with respect to the number of authors, while it has to be rejected with respect to gender parity. Finally, a differentiated approach needs to be considered with respect to content categories, which should be done depending on the aim of the company.
Hypothesis 2: Corporate blog authenticity
Post authenticity has been assessed by calculating F–scores for all blogs and including these values in the stepwise regressions for the individual country samples. While the analysis verified that blog texts are in general significantly less formal that traditional corporate communication means, a clear relation between formality and popularity could not be identified by the regressions. Only the Russian sample showed some negative impact of formality. This may imply that authenticity is a general feature of blogs and may be a general precondition for blog popularity, independent of the specific value. Therefore, Hypothesis 2 has to be rejected and further tests become necessary in order to analyze the effect of authenticity on corporate blogs.
Hypothesis 3: Reader–friendly blog management
Neither post frequency nor post length seem to have a significant impact on blog popularity. No clear trend could be observed with respect to media usage. Positive as well as negative effects were revealed depending on the specific sample, although the positive estimations in general featured a higher statistical significance. Therefore, Hypothesis 3 needs to be rejected and further tests with respect to reader preferences become necessary to evaluate the impact of post statistics and media usage on reader behavior.
Hypothesis 4: Corporate blog development
Blog age has a small but significant positive impact on both Technorati Authority as well as comment frequency. A higher number of sidebar elements gives proactive readers additional tools to search or use a corporate blog, which particularly translates into a higher comment frequency. The number of interactivity elements had no overall impact on a corporate blog’s acceptance. Therefore Hypothesis 4 is accepted when limited to blog age (for incoming links and comment frequency) and the number of sidebar elements (for comment frequency).
Hypothesis 5: corporate blog networking
The number of blogroll elements has a significant positive impact on a blog’s Technorati Authority. As an indicator for networking effort, it signals that increased networking activities will lead to higher blog popularity in terms of incoming links. Hence, Hypothesis 5 is clearly accepted for Technorati Authority, where networking is considered an important determinant of blog popularity, but rejected for comment frequency.
Implications for theory and practice
A range of implications could be derived for research regarding corporate blogs and for corporate blog management and corporate communication strategies regarding the use of corporate blogs in varying institutional contexts.
For academic research on social media, this study contributes to a theoretical understanding of the effects of corporate blogging practices within an international management context. This study integrated these two research areas, focusing on corporate blogs from Germany, Russia and the U.S. Such integration is of particular relevance given the international proliferation of the Internet, its increasing function as international information and communication channel, and the fading boundaries of business and communication in a globalizing world. At the same time measures were developed or adapted in order to capture different aspects of corporate blogging practices.
This study also generated several insights for corporate communication and marketing. First of all different types of acceptance measures were introduced that allow companies to evaluate popularity and impact of corporate blogging practices. Focusing on linking within the blogosphere (Technorati Authority) and the amount of collected feedback (average number of comments per post), the study then identified factors influencing the level of these two measures.
Technorati Authority, as a measure of overall blog acceptance, was found to be positively influenced by a corporate blog’s diversity. In particular the number of authors and topics on the blog are important determinants. Companies that can increase the number of authors (for example by increasing the blog editing team or by actively motivating employees to voluntarily contribute) or topics (for example by encouraging employees from different departments and functional areas) are thereby able to attract more links from other bloggers and websites. On the other hand, other diversity indicators had an opposite effect. The share of male authors was found to be positively related to Technorati Authority. Networking, i.e., proactively linking to other blogs, was found to have a similar effect, as reciprocity is an important determinant of popularity within the blogosphere. This concerns the blogroll on a corporate blog but can also be extended to other types of links on a corporate blog, for example within individual posts.
Furthermore, several factors with a significant influence on the average number of comments could be identified. The number of navigation functions and sidebar elements such as archive, categorization or search functions had a significant positive effect. The company can therefore increase the amount of feedback collection by making the blog easier to use for interested users for example by making past content better accessible or categorizing content in accordance with reader (or in this context rather commenter) preferences. An inverse effect was identified for post length. Shorter blog posts were associated with a higher number of average comments. Also media usage was overall found to negatively affect commenting, as too frequent use of media elements rather cater to readers that seek entertainment rather than ways of interacting with a company as can be observed by the range of corporate blogs that feature mostly media content rather than text content such as the General Motors FYI Blog, which attracts many visitors but features very little interaction . Consequently, companies need to increase awareness among blog authors and editors regarding the effects of individual post characteristics on the level of interaction. Companies can even conduct trainings of corporate bloggers regarding writing style or good text and media usage. In contrast to Technorati Authority, comment frequency was adversely affected by the number of topics. The average number of comments was higher for corporate blogs covering fewer topics. This implies that companies can better engage readers and attract feedback and contributions by limiting the topical scope of the activity. This adverse relationship also requires companies to decide what benefits they actually seek from corporate blogging activities as it might be difficult to achieve traffic– and feedback–oriented objectives in parallel.
Blog age was found to positively contribute to both link popularity as well as comment frequency. Older corporate blogs have a higher acceptance in terms of both popularity measures. For companies this implies that popularity builds over time and companies can benefit from showing some patience with their corporate blogging activities. Many corporate blogs such as the German Fischer–Blog (by Fischerwerke GmbH & Co. KG) are very quickly abandoned when companies become disappointed with the size of audience or amount of collected feedback. However, corporate blogging as a social media activity has a strong social relationship aspect and the desired social networks around a corporate social media activity grow slowly. Therefore companies should not only look at the current absolute benefits from a corporate blog but should rather monitor the development over time to arrive at conclusions about their blog’s performance.
Last, but not least, the study was able to reject some of the general (mostly theoretical) assumptions about corporate blogging practices. First of all, it was shown that post frequency is not an important factor for corporate blog popularity, indicating that quality rather than quantity is key to success in the blogosphere (at least in terms of incoming links and number of comments). Second of all, the common belief about authenticity as a means to rising blog popularity can be de–mystified. While seems to be a common feature of corporate blogs to be less formal than other corporate communication activities, there is no specific impact on a blog’s popularity.
Figure 6 summarizes the key findings of this study for corporate blogging practice:
Figure 6: Key findings for corporate blogging practice.
Limitations and future research directions
Several limitations derive from the collection and the composition of the dataset underlying the conducted analyses. In particular the limitation to directly observable blog characteristics constrained the evaluation of corporate blogging practices, limiting the choice of dependent variables (for example excluding simple traffic or reader numbers that are only observable by a blog’s administrator), but also limiting the choice and construction of independent variables. Consequently, several potentially important factors were excluded as they were not observable from outside such as the integration with other corporate communication activities or additional diversity aspects (for example author age or the authors’ roles and functions within the company). Other important determinants were simplified owing to the restriction to observable characteristics. In particular the authenticity dimension (which against popular theory was not found to influence corporate blog acceptance) was modeled as the formality of writing style only, ignoring other aspects of authenticity such as the credibility of the individual author or the credibility of individual messages communicated in the blog posts. The diversity dimension was limited to simple observable characteristics such as author numbers or gender, neglecting other important aspects of diversity such as diversity of opinion or ethnic diversity. For example, it needs to be acknowledged that an author group of 10 authors can be homogeneous or heterogeneous depending on aspects such as ethic background, education, income, etc. It was unfortunately beyond the scope of this research to analyze diversity at such level of detail. Also, networking efforts were explained through the existence of a blogroll and the number of blogs in the blogroll only, ignoring more complex networking activities (including offline activities such as participation in industry and social media events) that cannot precisely be observed from a corporate blog directly. In addition, a limitation to the networking dimension is the unknown causality. It requires additional research to identify whether outbound links are a reaction to or a cause of inbound links. Also company characteristics were largely ignored (for example size of the company, type of products or industry), although these might cause differences in acceptance as for example corporate blogs by Internet and software companies (such as Google or Yahoo) might by more popular just because they are more relevant to Internet users. Despite these limitations, the current study represents an initial step in exploring the possible determinants of corporate blog acceptance including links by other blogs or interaction with the audience.
A key limitation of this study is the use of popularity indicators, i.e., comment statistics and inbound links from other blogs, rather than actual success measures. While this type of benchmarking allows comparing practices between corporate blogs, it is difficult to derive specific recommendations for individual communication strategies. As outlined in Section 3, the meaning of success in social media for companies is extremely diverse. To correctly apply the findings from this study, companies have to first evaluate the role of popularity for their corporate blog strategy, before in detail considering the implication of specific blog characteristics.
Extensions of this research seem promising in at least three directions. First, it is worthwhile to apply the methodology to other institutional frameworks, i.e., to corporate blog samples from other countries. Second, additional research seems necessary to identify effects on traffic and reader numbers, which are often the most widely used type of popularity indicator by companies as they allow a direct comparison to other types of communication and marketing. Last but not least, an in–depth analysis of corporate blogging strategies and means and indicators of success will be helpful. Such research would enable a better application of the findings of this study to individual blogging and communication activities. Eventually, these proposed extensions will be necessary to tailor corporate blogging practices not only to general institutional frameworks but to more detailed target group characteristics as well as company objectives.
About the author
Nils König is an Assistant Professor for Marketing and Corporate Communication at the American University of Central Asia in Bishkek, Kyrgyz Republic. His research focusses on the role of social media in modern corporate communication.
E–mail: koenig_n [at] mail [dot] auca [dot] kg
1. Schmidt, 2007b, p. 1,409.
2. Schmidt, 2007b, p. 5.
3. Schmidt, 2007b, 1,418.
4. Trammell, et al., 2006, p. 716.
5. Heylighen and Dewaele, 1999, p. 1.
6. This might be one of the main reasons why the FYI Blog was in 2011 turned into a pure media portal, featuring pictures and videos of General Motors products, abandoning the blog format and thus the commenting function.
N. Ali–Hasan and L.A. Adamic, 2007. “Expressing social relationships on the blog through links and comments,” Proceedings of the 1st Annual Meeting of the North American Chapter of the Association for Computational Linguistics, at http://www-personal.umich.edu/~ladamic/papers/oc/onlinecommunities.pdf, accessed 4 February 2013.
C. Anderson, 2009. “Fortune 500 business blogging wiki,” at http://www.socialtext.net/bizblogs/index.cgi, accessed 23 April 2009.
P.A. Argenti, 2007. Corporate communication. Fourth edition. Boston: McGraw–Hill/Irwin.
P.A. Argenti, R.A. Howell and K.A. Beck, 2005. “The strategic communication imperative,” Sloan Management Review, volume 46, number 3, pp. 83–89.
E. Berscheid, 1966. “Opinion change and communicator–communicatee similarity and dissimilarity,” Journal of Personality and Social Psychology, volume 4, number 6, pp. 670–680.http://dx.doi.org/10.1037/h0021193
B. Block, 2010. “Russia has most engaged social networking audience worldwide” (20 October), at http://www.comscore.com/Press_Events/Press_Releases/2010/10/Russia_Has_Most_Engaged_Social_Networking_Audience_Worldwide, accessed 4 February 2013.
T.C. Brock, 1965. “Communicator–recipient similarity and decision change,” Journal of Personality and Social Psychology, volume 1, number 6, pp. 650–654.http://dx.doi.org/10.1037/h0022081
J. Cass, K. Munroe and S. Turcotte, 2005. Corporate blogging: Is it worth the hype? Waltham, Mass.: Backbone Media, Inc., at http://www.backbonemedia.com/blogsurvey/, accessed 4 February 2013.
J.–S. Chiou and C. Cheng, 2003. “Should a company have message boards on its Web sites?” Journal of Interactive Marketing, volume 17, number 3, pp. 50–61.http://dx.doi.org/10.1002/dir.10059
J. Cornelissen, 2011. Corporate communication: A guide to theory and practice. Third edition. London: SAGE.
C. Dellarocas, X.M. Zhang and N.F. Awad, 2007. “Exploring the value of online product reviews in forecasting sales: The case of motion pictures,” Journal of Interactive Marketing, volume 21, number 4, pp. 23–45.http://dx.doi.org/10.1002/dir.20087
B. Etling, K. Alexanyan, J. Kelly, R. Faris, J. Palfrey and U. Gasser, 2010. “Public discourse in the Russian blogosphere: Mapping RuNet politics and mobilization” (18 October), Cambridge, Mass.: Berkman Center for Internet & Society, Harvard University, at http://cyber.law.harvard.edu/publications/2010/Public_Discourse_Russian_Blogosphere, accessed 4 February 2013.
S. Fittkau and H. Maaß, 2008. W3B–Ergebnisband. Hamburg: Fittkau & Maaß.
M. Fleck, L. Kirchhoff, M. Meckel and K. Stanoevska–Slabeva, 2007. “Applications of blogs in corporate communication,” Studies in Communication Sciences, volume 7, number 2, pp. 227–245.
A. Gaudeul, L. Mathieu and C. Peroni, 2009. “Blogs and the economics of reciprocal attention,” at http://ideas.repec.org/p/pra/mprapa/11298.html, accessed 4 February 2013.
Gomita, 2012. “ScrapBook,” at http://amb.vis.ne.jp/mozilla/scrapbook/, accessed 11 April 2011.
S.C. Herring and J.C. Paolillo, 2006. “Gender and genre variation in weblogs,” Journal of Sociolinguistics, volume 10, number 4, pp. 439–459.http://dx.doi.org/10.1111/j.1467-9841.2006.00287.x
S.C. Herring, L.A. Scheidt, I. Kouper and E. Wright, 2007. “A longitudinal content analysis of weblogs: 2003–2004,” In: M. Tremayne (editor). Blogging, citizenship, and the future of media. New York: Routledge, pp. 3–20.
S.C. Herring, I. Kouper, L.A. Scheidt and E. Wright, 2004a. “Women and children last: The discursive construction of weblogs,” Into the blogosphere: Rhetoric, community, and culture of weblogs, at http://blog.lib.umn.edu/blogosphere/women_and_children.html, accessed on 11 June 2012.
S.C. Herring, L.A. Scheidt, S. Bonus and E. Wright, 2004b. “Bridging the gap: A genre analysis of weblogs.” HICSS ’04: Proceedings of the Proceedings of the 37th Annual Hawaii International Conference on System Science, pp. 101–111.
F. Heylighen and J.–M. Dewaele, 2002. “Variation in the contextuality of language: An empirical measure,” Foundations of Science, volume 7, number 3, pp. 293–340.http://dx.doi.org/10.1023/A:1019661126744
F. Heylighen and J.–M. Dewaele, 1999. “Formality of language: Definition, measurement and behavioral determinants,” Brussels: Free University of Brussels, at http://pespmc1.vub.ac.be/Papers/Formality.pdf, accessed 4 February 2013.
G. Hofstede, 1980. Culture’s consequences: International differences in work–related values. Beverly Hills, Calif.: Sage.
Y.P. Huo and R.M. Steers, 1993. “Cultural influences on the design of incentive systems: The case of East Asia,” Asia Pacific Journal of Management, volume 10, number 1, pp. 71–85.http://dx.doi.org/10.1007/BF01732225
A.M. Kaplan and M. Haenlein, 2010. “Users of the world, unite! The challenges and opportunities of social media,” Business Horizons, volume 53, number 1, pp. 59–68.http://dx.doi.org/10.1016/j.bushor.2009.09.003
J. Kelly, 2009a. Mapping the global blogosphere. Berlin: Newthinking Communications.
J. Kelly, 2009b. “Passing the baton 2009,” at http://www.usip.org/files/file/5.pdf, accessed 4 February 2013.
J. Kelly, 2008. “Pride of place: Mainstream media and the networked public sphere,” Cambridge, Mass.: Berkman Center for Internet & Society, Harvard University, at http://cyber.law.harvard.edu/, accessed 4 February 2013.
L. Kirchhoff, A. Bruns and T. Nicolai, 2007. “Investigating the impact of the blogosphere: Using PageRank to determine the distribution of attention,” Association of Internet Researchers (Vancouver), at http://eprints.qut.edu.au/10517/, accessed 4 February 2013.
N. König, 2011. “Blogs in der Unternehmenskommunikation,” In: J. Rump, F. Schabel and S. Grabmeier (editors). Auf dem Weg in die Organisation 2.0: Mut zur Unsicherheit. Sternenfels: Wissenschaft & Praxis.
N. König, 2009a. “‘Einblicke in einen Konzern’: Unternehmensblogs als Anlaufstelle für Bewerber am Beispiel des Daimler–Blogs,” In: A. Hohenstein and K. Wilbers (editors). Handbuch E–Learning: Expertenwissen aus Wissenschaft und Praxis. Köln: Deutscher Wirtschaftsdienst.
N. König, 2009b. “What makes corporate blogs successful? A cross–cultural empirical study of corporate blog characteristics,” Proceedings of the COST Action 298 Conference 2009.
P. Kotler, H. Kartajaya and I. Setiawan, 2010. Marketing 3.0: From products to customers to the human spirit. Hoboken, N.J.: Wiley.
H.–H. Lee, S.R. Park and T. Hwang, 2008. “Corporate–level blogs of the Fortune 500 companies: an empirical investigation of content and design,” International Journal of Information Technology and Management, volume 7, number 2, pp. 134–148.http://dx.doi.org/10.1504/IJITM.2008.016601
S. Lee, T. Hwang and H.–H. Lee, 2006. “Corporate blogging strategies of the Fortune 500 companies,” Management Decision, volume 44, number 3, pp. 316–334.http://dx.doi.org/10.1108/00251740610656232
C. Li and T. McHarg, 2007. “Calculating the ROI of blogging: A case study” (24 January), Cambridge, Mass.: Forrester Research, at http://www.forrester.com/, accessed 4 February 2013.
C. Li and C. Stromberg, 2007. “The ROI of blogging” (24 January), Cambridge, Mass.: Forrester Research, at http://www.forrester.com/, accessed 4 February 2013.
J. Mackey, 2008. “Back to blogging,” at http://www2.wholefoodsmarket.com/blogs/jmackey/2008/05/21/back-to-blogging/, accessed on 23 April 2009.
J. Mackey, 2007. “Temporary hold on my blog,” at http://www2.wholefoodsmarket.com/blogs/jmackey/2007/07/17/temporary-hold-on-my-blog/, accessed on 23 April 2009.
C.A. Marlow, 2004. “Audience, structure and authority in the weblog community,” at http://alumni.media.mit.edu/~cameron/cv/pubs/04-01.html, accessed 4 February 2013.
B. Nardi, D.J. Schiano, M. Gumbrecht and L. Swartz, 2004. “Why we blog,” Communications of the ACM, volume 47, number 12, pp. 41–46.http://dx.doi.org/10.1145/1035134.1035163
Nielsen, 2010. “Led by Facebook, Twitter, global time spent on social media sites up 82% year over year” (22 January), at http://blog.nielsen.com/nielsenwire/global/led-by-facebook-twitter-global-time-spent-on-social-media-sites-up-82-year-over-year/, accessed 4 February 2013.
M. Paul, A. Finch and E. Sumita, 2011. “Word segmentation for dialect translation,” CICLing’11: Proceedings of the 12th International Conference on Computational Linguistics and Intelligent Text Processing, Part II, pp. 55–67.
A. Picot and T. Fischer (editors), 2006. Weblogs professionell: Grundlagen, Konzepte und Praxis im unternehmerischen Umfeld. Heidelberg: dpunkt Verlag.
M.E. Porter, 2001. “Strategy and the Internet,” Harvard Business Review, volume 79, number 3, pp. 62–78.
C. Puschmann, 2007a. “Corpora, blogs and linguistic variation: Arguments for using structured web data in corpus development“ (8 November), at http://ynada.com/presentations.html, accessed 4 February 2013.
C. Puschmann, 2007b. “Variation and ‘genrefication’ in blogs“ (28 February), at http://ynada.com/presentations.html, accessed 4 February 2013.
A. Quinn, 2010. “Do most languages need more space than English?” (14 September), at http://english.stackexchange.com/questions/2998/do-most-languages-need-more-space-than-english, accessed 19 March 2012.
L. Rainie, 2005. “The state of blogging” (2 January), Pew Internet & American Life Project, at http://pewinternet.org/Reports/2005/The-State-of-Blogging.aspx, accessed 4 February 2013.
U. Röttger and S. Zielmann, 2006. “Weblogs — unentbehrlich oder überschätzt für das Kommunikationsmanagement von Organisationen?” In: A. Picot and T. Fischer (editors). Weblogs professionell: Grundlagen, Konzepte und Praxis im unternehmerischen Umfeld. Heidelberg: dpunkt Verlag, pp. 31–50.
B. Santorini, 1990. “Part–of–speech tagging guidelines for the Penn Treebank Project,” University of Pennsylvania, School of Engineering and Applied Science, Department of Computer and Information Science; Technical report, MS–CIS–90–47.
R. Schabel, 2008. Einführung eines Weblogs bei der Daimler AG. Stuttgart: Hochschule der Medien Stuttgart.
L.A. Scheidt and E. Wright, 2004. “Common visual design elements of weblogs,” In: L.J. Gurak, S. Antonijevic, L. Johnson, C. Ratliff and J. Reyman (editors). Into the blogosphere: Rhetoric, community, and culture of weblogs. St. Paul: University of Minnesota, and at http://blog.lib.umn.edu/blogosphere/common_visual.html, accessed 4 February 2013.
A. Schiller, S. Teufel, C. Stöckert and C. Thielen, 1999. “Guidelines für das Tagging deutscher Textcorpora Guidelines für das Tagging deutscher Textcorpora mit STTS (Kleines und großes Tagset),” Stuttgart: Institut für maschinelle Sprachverarbeitung Universität Stuttgart, at http://www.sfs.uni-tuebingen.de/resources/stts-1999.pdf, accessed 4 February 2013.
J. Schmidt, 2007a. “Blogging practices in the German–speaking blogosphere,” Bamberg: Universität Bamberg, at http://nbn-resolving.de/urn:nbn:de:0168-ssoar-9953, accessed 4 February 2013.
J. Schmidt, 2007b. “Blogging practices: An analytical framework,” Journal of Computer–Mediated Communication, volume 12, number 4, pp. 1,409–1,427, and at http://jcmc.indiana.edu/vol12/issue4/schmidt.html, accessed 4 February 2013.
J. Schmidt, 2007c. “Stabilität und Wandel von Weblog–Praktiken: Erste empirische Befunde,” In: S. Kimpeler, M. Mangold and W. Schweiger (editors). Die digitale Herausforderung: Zehn Jahre Forschung zur computervermittelten Kommunikation. Wiesbaden: Verlag für Sozialwissenschaften, pp. 51–60.
S. Sharoff, M. Kopotev, T. Erjavec, A. Feldman and D. Divjak, 2008. “Designing and evaluating a Russian tagset,” Proceedings of the 6th International Conference on Language Resources and Evaluation (LREC), pp. 279–285, and at http://www.lrec-conf.org/proceedings/lrec2008/pdf/78_paper.pdf, accessed 4 February 2013.
R.D. Smith, 2005. Strategic planning for public relations. Second edition. Mahwah, N.J.: Lawrence Erlbaum.
K.J. Srnka and S.T. Koeszegi, 2007. “From words to numbers: How to transform qualitative data into meaningful quantitative results,” Schmalenbach Business Review, volume 59, number 1, pp. 29–57.
K.D. Trammell, A. Tarkowski, J. Hofmokl and A.M. Sapp, 2006. “Rzeczpospolita blogów [Republic of Blog]: Examining Polish bloggers through content analysis,” Journal of Computer–Mediated Communication, volume 11, number 3, pp. 702–722, and at http://jcmc.indiana.edu/vol11/issue3/trammell.html, accessed 4 February 2013.
M. Turck, 2007. “Traits of the Fortune 500 blogs,“ at http://www.socialtext.net/bizblogs/index.cgi?traits_of_the_fortune_500_blogs, accessed 23 April 2009.
S.H. Wilke, 2009. “The Daimler–Blog — A case study. An analytical approach to the benefits of corporate weblogs with respect to company intentions & expectations,” In: I. Peters, C. Puschmann, V. Trkulja and K. Weller (editors). Proceedings of the SsSoft09 (Düsseldorf), pp. 1–36.
O.E. Williamson, 1996. The mechanisms of governance. New York: Oxford University Press.
A.G. Woodside and J.W. Davenport, 1974. “The effect of salesman similarity and expertise on consumer purchasing behavior,” Journal of Marketing Research, volume 11, number 2, pp. 198–202.http://dx.doi.org/10.2307/3150562
S. Worchel, V. Andreoli and J. Eason, 2006. “Is the medium the message? A study of the effects of media, communicator, and message characteristics on attitude change,” Journal of Applied Social Psychology, volume 5, number 2, pp. 157–172.http://dx.doi.org/10.1111/j.1559-1816.1975.tb01305.x
S.–U. Yang and J.S. Lim, 2009. “The effects of blog-mediated public relations (BMPR) on relational trust,” Journal of Public Relations Research, volume 21, number 3, pp. 341–359.http://dx.doi.org/10.1080/10627260802640773
A. Zerfaß and D. Boelter, 2005. Die neuen Meinungsmacher: Weblogs als Herausforderung für Kampagnen, Marketing, PR und Medien. Graz: Nausner & Nausner.
A. Zerfaß and S. Sandhu, 2005. Virtuelle Authentizität: Die Nutzung von Weblogs als Kommunikationsinstrument für das Top–Management. Bonn: Deutsche Gesellschaft für Publizistik und Kommunikationswissenschaft (DGPuK).
F. Zhu and X.M. Zhang, 2010. “Impact of online consumer reviews on sales: The moderating role of product and consumer characteristics,” Journal of Marketing, volume 74, number 2, pp. 133–148.http://dx.doi.org/10.1509/jmkg.74.2.133
Appendix 1: Corporate blog sample
Sub–sample of German corporate blogs ahlers. (AO Deutschland Medien GmbH) Frosch Blog (Werner & Mertz GmbH) Alles über EVE (Delphi Corporation) Frosta Blog (FRoSTA AG) AMD Notebook Test Blog (Advanced Micro Devices, Inc.) Jokers–Blog (Verlagsgruppe Weltbild GmbH) Ausbildungsblog (Festo AG & Co. KG) Langenscheidt Weblog (Langenscheidt Verlagsgruppe) Bosch Trainee–Blog (Robert Bosch GmbH) Lotus Germany (IBM Corporation) Bredl bloggt (Telekom TA AG) Magix Blog (MAGIX AG) Daimler Blog (Daimler AG) Payback Blog (Loyalty Partners GmbH) Das Recruiting von Accenture (Accenture Ltd.) RheinNeckarWeb (BASF SE) DocMorris Blog (DocMorris) T–Systems Automotive Blog (T–Systems Enterprise Services) Frischegarantie (Wüstenrot AG) Umweltmanagement Swisscom (Swisscom AG) Sub–sample of Russian corporate blogs allsoft.ru blog (SAO Softline) Блог Яндекса (Yandex LLC) Press Club (VimpelCom) Веблог “Лаборатории Касперского” (Kaspersky Lab) блог ГМК “Норильский никель” (MMC Norilski Nikel) Как живет IT в Intel (Intel Corporation) Блог издателей деловой литературы (Mann, Ivanov & Ferber) Официальный блог — Google Россия (Google Inc.) Блог МегаФон Москва (MegaFon) Просто о сложном (Alfa Group) Sub–sample of U.S. corporate blogs A thousand words (Eastman Kodak Company) Islamic Finance blog (PriceWaterhouseCoopers LLP) Accelerating High Performance Business (Accenture Ltd.) IT@Intel (Intel Corporation) Alliconnect (GlaxoSmithKline plc) JNJ BTW (Johnson & Johnson) Amazon Webservices Blog (Amazon.com Inc.) Jonathan’s Blog (Sun Microsystems, Inc.) Baby Babble! (Groupe Danone) Kia BUZZ (Hyundai Kia Automotive Group) Backstage at Sundance (Hewlett–Packard Company) Marriott in the Kitchen (Marriott International, Inc.) BenettonTalk (Benetton Group) Marriott on the move (Marriott International, Inc.) BrewBlog (SAB Miller plc) n–gage blog (Nokia Corporation) Cadillac Drivers’ Log (General Motors Corporation) Nickel from Norilsk (MMC Norilski Nikel) CheckOut (WalMart–Stores Inc.) Official Google Blog (Google Inc.) Chrysler Blog (Chrysler LLC) Open for Discussion (McDonald’s Corporation) Cisco High Tech Policy Blog (Cisco Systems, Inc.) open mike (Pitney Bowes Inc.) Coca–Cola Conversations (The Coca–Cola Company) Play On (Xerox Corporation) Corporate Reporting blog (PriceWaterhouseCoopers LLP) PluggedIn (Eastman Kodak Company) Creating History Blog (Arcelor Mittal) PwC People blog (PriceWaterhouseCoopers LLP) CSR@Intel (Intel Corporation) Randy’s Journal (The Boeing Company) Dell Shares (Dell Inc.) Research@Intel (Intel Corporation) Digital Straight Talk (Cox Communications, Inc.) Small & Medium Business Community Blog (Hewlett–Packard Co.) Direct2Dell (Dell Inc.) Small Business Community Blog (Microsoft Corporation) Dr. Laundry (Clorox Company) Sony Electronics Blog (Sony Corporation) Ed Gottsman: Weblog (Accenture Ltd.) Southwest Airlines Blog (Southwest Airlines Corporation) EDS’ Next Big Thing (Electronic Data Systems) Technology@Intel (Intel Corporation) Emerson Process Experts (Emerson Electric Company) The Blog (Avaya Inc.) Exchange (NYSE Euronext Inc.) The Bovine Bugle (Groupe Danone) Experienced Consultants Blog (Accenture Ltd.) The Calculating World with You and Wing, (Hewlett–Packard Co.) Fastlane Blog (General Motors Corporation) The Future of Documents (Xerox Corporation) Finance & Treasury blog (PriceWaterhouseCoopers LLP) The Gender Agenda (PriceWaterhouseCoopers LLP) From Edison’s Desk (General Electric Company) The Official Palm Blog (Palm, Inc.) FYI Blog (General Motors Corporation) The Penguin Blog (Pearson PLC) GM TunerSource (General Motors Corporation) The Student LoanDown (Wells Fargo & Co.)
Appendix 2: Corporate blog sample
Pooled sample and German sample
Note: Pooled sample: N=107. Correlations above |0.16| are significant at p < 0.05. For Germany: N20. Correlations above |0.38| are significant at p < 0.05.
All Germany 1 2 3 4 5 6 7 8 9 10 11 12 13 1. Technorati Authority — -0.07 0.18 0.25 0.33 -0.11 -0.08 0.55 0.36 0.80 0.42 -0.23 0.51 2. Comments per post 0.13 — -0.28 -0.16 -0.16 0.14 0.11 0.11 0.59 -0.21 -0.03 -0.05 -0.08 3. Post frequency 0.18 -0.02 — -0.07 0.30 -0.03 0.17 -0.19 -0.22 0.11 0.07 -0.19 0.15 4. Post length 0.02 -0.08 -0.24 — 0.37 0.05 -0.11 0.08 0.02 -0.13 0.18 -0.40 -0.26 5. Number of authors 0.57 0.08 0.55 -0.02 — -0.50 -0.05 0.02 -0.08 0.03 -0.22 -0.39 -0.04 6. Gender disparity -0.02 0.04 -0.29 0.05 -0.40 — 0.02 0.10 0.24 0.00 0.37 0.13 -0.01 7. Blog age 0.30 0.12 0.05 -0.01 0.12 0.09 — -0.23 -0.07 -0.24 -0.20 -0.45 0.06 8. Number of blogroll elements 0.42 0.12 0.02 -0.03 0.31 -0.11 0.05 — 0.63 0.45 0.35 -0.04 0.37 9. Number of sidebar elements 0.01 0.31 -0.09 0.13 2.22 0.00 -0.07 0.19 — 0.30 0.35 0.06 0.36 10. Media per post 0.03 0.03 0.20 0.29 0.17 0.01 -0.03 0.01 -0.01 — 0.53 0.02 0.54 11. Interactivity -0.01 0.03 -0.09 0.19 0.06 -0.11 -0.16 0.08 0.16 -0.02 — -0.23 -0.02 12. F–score -0.02 -0.12 -0.02 -0.03 -0.06 -0.03 -0.05 0.06 -0.15 0.03 -0.17 — 0.18 13. Topical dimensions 0.14 0.04 0.13 -0.07 0.28 -0.10 -0.05 0.22 0.21 0.20 -0.03 0.01 — All mean 157.8 5.40 2.45 331.3 6.58 0.36 568.6 9.94 5.29 0.88 2.28 68.37 4.05 sd 873.7 17.63 2.60 183.1 9.22 0.17 354.9 17.29 1.71 1.09 0.96 5.87 1.98 G mean 20.60 11.48 2.345 216.8 5.15 0.39 504.4 5.55 5.30 0.66 2.10 62.65 3.85 sd 28.16 38.69 1.71 159.8 5.72 0.15 271.0 10.67 2.18 0.64 0.85 5.47 2.01
Russian and U.S. sample
Note: For Russia: N=10. Correlations above |0.55| are significant at p < 0.05. For the U.S.: N=77. Correlations above |0.19| are significant at p < 0.05.
Russia U.S. 1 2 3 4 5 6 7 8 9 10 11 12 13 1. Technorati Authority — 0.42 0.24 -0.02 0.62 -0.03 0.33 0.63 -0.02 0.03 -0.03 -0.03 0.15 2. Comments per post 0.48 — 0.17 0.12 0.39 -0.04 0.34 0.46 0.10 0.32 0.17 -0.09 0.25 3. Post frequency -0.19 0.17 — -0.26 0.61 -0.25 0.09 0.20 0.10 0.21 -0.17 -0.11 0.15 4. Post length -0.16 -0.26 -0.29 — -0.03 0.02 -0.02 -0.06 0.05 0.30 0.20 0.01 -0.09 5. Number of authors -0.18 0.61 0.47 -0.45 — -0.36 0.12 0.53 0.13 0.23 0.10 -0.14 0.35 6. Gender disparity -0.23 -0.61 -0.65 0.47 -0.57 — 0.10 -0.08 -0.19 -0.02 -0.20 0.08 -0.08 7. Blog age -0.09 -0.09 -0.23 -0.09 0.30 0.20 — 0.11 -0.09 -0.02 -0.11 -0.01 -0.09 8. Number of blogroll elements 0.92 0.38 -0.23 -0.12 -0.19 -0.24 0.00 — 0.21 0.06 0.07 -0.08 0.28 9. Number of sidebar elements 0.26 -0.04 -0.60 0.39 -0.53 0.34 -0.18 0.24 — -0.02 0.11 -0.00 0.18 10. Media per post -0.34 -0.48 0.15 0.80 -0.31 0.35 0.03 -0.28 -0.11 — -0.05 -0.09 0.18 11. Interactivity 0.19 0.27 0.21 -0.33 -0.00 -0.09 -0.72 -0.02 -0.14 -0.34 — -0.27 -0.01 12. F–score -0.02 -0.63 -0.04 -0.31 -0.37 0.34 -0.09 0.05 0.04 -0.10 0.20 — -0.04 13. Topical dimensions 0.00 0.28 0.04 0.22 0.12 -0.53 -0.04 0.02 0.24 0.01 -0.45 -0.64 — R mean 14.90 4.84 3.51 234.9 10.10 0.27 540.1 15.10 3.30 1.19 2.10 77.51 4.10 sd 29.81 3.05 4.61 150.9 8.31 0.17 371.9 41.32 1.25 1.43 0.88 4.30 1.73 U.S. mean 214.8 3.89 2.33 373.5 6.49 0.36 589.0 10.42 5.55 0.89 2.35 68.67 4.09 sd 1038 7.10 2.47 177.9 10.06 0.18 375.5 13.47 1.47 1.14 0.99 4.37 2.03
Received 13 June 2012; revised 25 July 2012; accepted 20 January 2013.
This paper is licensed under a Creative Commons Attribution–NonCommercial–ShareAlike 3.0 United States License.
International corporate blogging practices and effects
by Nils König
First Monday, Volume 18, Number 2 - 4 February 2013