Cultural Differences in Relationship Cultivation Strategies and
Consumer Reactions on Facebook
A cross-cultural comparison of online interactions between multinationals and
consumers in Colombia, Germany and the United Kingdom
Jacqueline Heinz (11354100) Master’s Thesis
University of Amsterdam
Graduate School of Communication
Master’s Programme Communication Science
Supervisor: dhr. dr. Theo B. Araujo Date of submission: 30.06.17
1 Abstract
Multinationals address a global audience on social media, thus the importance of local culture for online communications and interactions has steadily gained importance. In order to build relationships with their audience organizations actively engage on Facebook, a platform which is inherently very interactive and social and therefore facilitates dialogues between the corporate site and the stakeholders. The present study analyses how multinationals use
Facebook to interact with their publics in three countries: Colombia, Germany and the United Kingdom. The focus was initially on the corporate site which concentrated on how they attempt to engage in conversations with their followers. The second focus was on how the consumers truly reacted to these approaches. A content analysis on six corporate Facebook pages with 900 corporate posts was conducted. Overall the multinationals communicated very similarly across the different cultures, thus indicating that the
individualistic-collectivistic cultural values may not be the only influence on online communication. The findings strengthen the assumption that SNS, like Facebook facilitate the development of standardised virtual cultures and values, particularly where individual cultural differences diminish. The consumer engagement surprisingly differed strongly between the two individualistic countries (namely The UK and Germany) showing that the amount of consumer engagement is not only dependent of the message content but also on the culture itself.
Keywords: Online relationship cultivation, Hofstede’s dimensions of national culture,
2 Introduction
Social media is actively included in organisational marketing and brand building strategies (Kaplan & Haenlein, 2010) because it has become part of everyone’s daily lives around the globe (Men & Tsai, 2012). Thus, the importance of local culture for online communication has emerged as an increasingly relevant subject for cross-cultural research (Hermeking, 2005) because multinationals address a global audience on social media. As social media constitutes a powerful tool to build relationships with consumers (Tsai & Men, 2017), it is even more important for multinationals to embrace cultural differences (Waters & Lo, 2012) as users around the globe have different communication preferences (Singh & Matsuo, 2004).
Research has found that organisations that do not employ one global communication approach for all cultures but rather adapt their strategies to local tendencies are the most successful on social media (Stelzner, 2011). However, this assumption is contradicted by scholars who argue that global social media adaptions lead to a new virtual culture (Pfister & Soliz, 2011) because globally-standardised social networking sites (SNSs) like Facebook and Twitter are changing and unifying the way in which the internet is used. Nonetheless, the impact of culture cannot be ignored because culture shapes the ways in which individuals interact and socialise with other members of society and it influences the motivations, lifestyles and product choices of consumers (Singh & Matsuo, 2004).
SNSs have been found to be more effective than traditional marketing in terms of influencing the attitudes and behaviours of users (Thackeray, Neiger, Hanson, & McKenzie, 2008). Early studies investigated how companies engage and maintain relationships with their audiences on SNSs (Correia Loureiro & Concalves Gomes, 2016). Among others, scholars have started examining which strategies companies follow to maintain a dialogue and build relationships with their followers through Facebook (Bortree & Seltzer, 2009; Waters,
3 Facebook messages and pages created by both organisations and stakeholders, comparing the United States and China. More recently, Correia Loureiro and Concalves Gomes (2016) have analysed Facebook pages and messages created by both users and organisations in Brazil and Portugal. The current study follows this approach, but adds to the existing research by only analysing corporate messages and not taking page descriptions into account.
Rather than evaluating what users post on the brand pages, the cultural different consumer engagement holds particular interest in the current study. Consumer engagement has become a critical element for any brand’s marketing strategy (Malhotra, Malhotra & See, 2013) and it has been widely researched. Many studies have investigated which types of posts receive the highest number of likes, shares or comments (Cvijkj & Michahelles, 2013; De Vries, Gensler, & Leeflang, 2012). Nonetheless, no study to date has measured the consumer engagement of the comments in a cross-cultural setting for online relationship cultivation strategies.
Many cross-cultural studies focus on China or Korea in comparison to the United States (US) due to their strong cultural differences (Men & Tsai, 2012; Luarn, Lin, & Chiu, 2015). However, there remains a gap in terms of cross-cultural comparisons of relationship cultivation strategies in other cultures. Colombia, Germany and the United Kingdom (UK) strongly vary among the six dimensions of national culture formulated by Hofstede
(Hofstede, Hofstede & Minkov, 2011) and have different communication styles (De Mooij & Hofstede, 2011). Moreover, these three countries have never been the subject of analysis in this context. The current study not only aims to analyse how multinationals try to engage with their audiences, but whether the consumers react in a culturally different manner. Therefore, the research question of this thesis is as follows:
4
How do multinationals’ use of online relationship cultivation strategies and the
corresponding consumer reactions on Facebook differ across Colombia,
Germany and the UK?
The current paper is based on a diverse strategic and theoretical framework to answer the aforementioned research question. The theoretical framework comprises: a) cultural differences described by Hofstede’s (1980) dimensions of national culture and Hall’s (1976) categorisation of context dependent communication styles; b) the online relationship
cultivation strategies first introduced by Kent and Taylor (1998) and later adapted for SNSs by Waters et al. (2009); and finally, c) cross-cultural differences in consumer engagement (Tsai & Men, 2017).
The following section will elaborate the three theoretical frameworks used for this study, before subsequently presenting the hypotheses. The procedure of the study will be explained in the methodology section, which is followed by results and ending with the discussion and limitations.
5 Theoretical Framework
Dimensions of National Culture
Hofstede (1980) first formulated five dimensions that characterise a national culture, which were later extended to six (Hofstede et al., 2010). The six dimensions of national culture are power distance, individualism versus collectivism, masculinity, as opposed to femininity, uncertainty avoidance, long-term versus short-term orientation and indulgence, as opposed to restraint (De Mooij & Hofstede, 2010; Gudykunst, 1997). Colombia, Germany and the UK are among the many countries to have been studied by Hofstede (2001). According to Hofstede’s (2017) country comparison model (see illustration 1), the three countries strongly differ on two dimensions: long-term orientation and individualism. These two dimensions are assumed to strongly influence the communication strategies of
multinationals on Facebook. On the other hand, the dimension of indulgence is expected to influence consumer engagement across the three countries.
Illustration 1.
6
Dimension of Indulgence. The dimension of indulgence describes the extent to which
people of a culture try to control or follow their desires and impulses (Hofstede, 2017). Germany has a culture of restraint, scoring lower than the other two countries (Hofstede, 2017b). As Hofstede (2017b) describes, countries that score low on this dimension do not give priority to leisure time and control the gratification of their desires, meaning that they do not easily get swept away to fulfil their desires. The UK scores higher than Germany in terms of indulgence, but is still lower than Colombia (see illustration 1). Indulgent societies such as Colombia have a tendency towards optimism. They see leisure time as more important, act more easily as they please and are more willing in realising their impulses and desires to enjoy life and have fun.
Dimension of long-term orientation. Cultures that score high on the long-term
orientation dimension are more pragmatic and they encourage thrift and change to prepare for challenges of the future, like Germany and the UK. Countries that score low on this
dimension – for example, Colombia – are more normative, prefer to maintain traditions and norms and do not prepare for future changes as quickly. By contrast, the German culture is very pragmatic, concentrated on achieving results and investing in the future, reflecting features that signalise their general long-term orientation (Hofstede, 2017b). British people score in between the two divergent countries and are moderately long-term-oriented (Hofstede, 2017c).
Dimension of individualism. The other dimension on which the three cultures
strongly vary is individualism versus collectivism (I-C). The importance of in-group belonging is one of the major characteristics that differentiates I-C cultures (Gudykunst, 1997). Colombia is one of the most collectivistic countries in the world, whereby the needs, goals and values of the in-group (e.g. family values and well-being) are seen as more
important than those of the individual (Gudykunst, 1997). This stands in complete opposite to individualistic societies, in which loose relations are the norm (Nishimura, Nevgi, & Tella,
7 2008) and the needs and goals of the individual are valued more highly than those of the in-group (Gudykunst, 1997). The British are considered one of the most individualistic and private sets of people in in the world (Hofstede, 2017c), whereas Germany is rather individualistic and somewhere in between the other two (Hofstede, 2017a/ 2017b).
Although I-C exists in all cultures, one tends to be more dominant than the other. Cultural I-C also has a direct influence on communication because the I-C tendencies
influence and guide the behaviour, rules and norms of the cultures (Gudykunst et al., 1996).
Communication Context Differences across Cultures
One way to differentiate communication styles across and within cultures is by applying Hall’s (1976) typology of high- (HC) and low-context (LC) cultures according to the degree of context dependence. Context in this sense applies to the situational information that someone needs to acquire to understand the meaning of the communication (Men & Tsai, 2012). Cultures do not communicate exclusively in a LC or HC manner; rather, both styles are used in every culture, although one communication style tends to be more prevalent (Richardson & Smith, 2007).
High-context communication. HC communication is more indirect, implicit and
ambiguous and a lot of knowledge about the information tends to already be internalised by the people communicating (Men & Tsai, 2012; Waters & Lo, 2012). In this indirect form of communication, the content of the verbal message is usually minimised, giving strong room for interpretation of implicit messages (Gudykunst et al., 1996). Research has agreed that in collectivistic cultures like Colombia the HC communication style is more prevalent
(Gudykunst et al., 1996; De Mooij, Hofstede 2011).
Low-context communication. LC communication is the exact opposite, being
explicit, direct and transmitted in a clear and straightforward manner (Waters & Lo, 2012). Speaking one’s mind, telling the truth and communicating consistently with one’s feelings
8 are characteristics valued and expected in LC communication (Gudykunst et al. 1996). This explicit communication style has been found to be predominant in individualistic cultures such as Germany and the UK (Gudykunst et al., 1996; De Mooij & Hofstede, 2011).
Online Relationship Cultivation Strategies on Facebook
Waters et al. (2009) examined how Facebook is used by non-profit organisations in the US to engage in a dialogue with their publics. For this study, they derived a framework from Kent and Taylor’s (1998) original “dialogic theory” for websites to cultivate
relationships with followers on SNSs. This framework aims to evaluate how companies try to stimulate a dialogic reciprocal discourse on SNSs. Their framework comprises three-essential online relationship cultivation strategies, which are: a) organisational disclosure, b)
information dissemination and c) interactivity.
The strategy of organisational disclosure refers to the willingness of the organisation to provide information about itself on its Facebook page and engage in open discourse with its public by providing – for example – its contact information. Companies should provide group descriptions, historical information, a link to the website, contact information of the company or simply share their mission statement and goals (Men & Tsai, 2012; Waters et al., 2009).
The second strategy – organisational information dissemination – addresses the interests, concerns and needs of the public while the organisation disseminates its
information. This type of information can appear in the form of posts with photos, videos or links to external sites with news about the organisation or its website, informing about organisational accomplishments or answering questions from consumers (Men & Tsai, 2012; Waters et al., 2009).
Finally, interactivity is an important strategy for dialogic relationship cultivation (Men & Tsai, 2012; Waters et al., 2009). The great advantage of Facebook is that it allows
9 organisations to engage interactively and directly with their consumers, through posting about upcoming events or linking to the website to purchase one of the organisation’s products (Waters & Lo, 2012).
Cross-cultural use of Online Relationship Cultivation Strategies on Facebook.
Men and Tsai (2012) expanded Waters et al.’s (2009) theoretical model
cross-culturally and compared organisational SNS messages in the US and China. They found that the corporate messages on American Facebook pages significantly differed from Chinese messages. American companies provided information directly related to the company or the products, following a hard-sell approach that fits their individualistic LC culture (Men & Tsai, 2012). By contrast, Chinese companies posted a lot of additional information to their followers, posted a lot of jokes, human interest stories, photos or videos with celebrity endorsers or simply posted greetings or jokes to engage with the SNS followers, like a friend (Men & Tsai, 2012). Such messages indicate an indirect and implicit communication style, which is typical for a collectivistic HC culture like China (Men & Tsai, 2012). Due to these findings, cultural difference in the use of the three online relationship cultivation strategies in Colombia, Germany and the UK are also expected and explained in the following paragraphs.
Organisational Disclosure. Men and Tsai (2012) found evidence that overall Chinese
companies were less open in terms of company disclosure on their Facebook page compared to the individualistic US companies. Other studies that have analysed online relationship cultivation strategies only look at organisational disclosure at the page level (Correia Loureiro & Concalves Gomes, 2016; Waters et al., 2009; Waters & Lo, 2012). The layout of a
Facebook page is very standardised (Cvijkj & Michahelles, 2013), which facilitates organisational disclosure at the page level because it is easy to fill out the requested fields. This is the reason why this paper measures organisational disclosure at a corporate message
10 level, where the actual interaction with followers and fans takes place. For example, Singh and Matsuo (2014) found evidence of cultural differences on websites between Japan and the US. As Japan is a highly collectivistic society, they like looking out for others in the group. This was also depicted in their different websites; for example, by offering and promoting fan club memberships, strongly emphasising their culture, tradition, and history. By contrast, US websites were found to be more direct, success-oriented and informative (Singh & Matsuo, 2014). This finding confirms that individualistic LC cultures tend to communicate more openly, share more information about themselves and do not like ambiguous and unclear messages (Gudykunst et al., 1996; De Mooij & Hofstede, 2011). This leads to the assumption that:
H1: Compared with their Colombian counterparts, multinationals provide more
explicit and direct organisational disclosure in their British and German
Facebook messages.
Information Dissemination. Another study investigating how culture influences the
organisational use of relationship cultivation strategies to interact with stakeholders on Facebook was conducted by Waters and Lo (2012). Their hypothesis expected US non-profit organisations – among others – to apply more information dissemination strategies on
Facebook than their collectivistic HC counterparts, Turkey and China. However, the authors found contradicting results. Their findings only supported the traditional cultural expectations for the individualistic US and the collectivistic Turkey. Conforming with their individualistic tendencies, US organisations had higher levels of information dissemination than the Turkish ones. Surprisingly, although China is a normative and collectivistic HC culture, they openly disseminated a lot of information about their successes and organisational achievements (Waters & Lo, 2012). In addition, Men and Tsai (2012) also found vague results for the information dissemination strategy on SNSs. In terms of providing information, US
11 companies outperformed the Chinese ones on certain levels (for example, featuring videos and news links), but showed the same levels for the usage of photos and making or
announcements. This paper aims to find clarification for these unclear results and assumes that individualistic LC cultures like the UK and Germany will disseminate more information overall on Facebook than Colombians due to their preference for direct and explicit
information. This assumption will be tested with the following hypothesis:
H2: Compared to Colombian Facebook pages, multinationals disseminate more:
a) pictures in their information-disseminating posts on German and British
Facebook pages;
b) videos/graphics in their information-disseminating posts on German and
British Facebook pages; and
c) news links and announcements in their information-disseminating posts on
German and British Facebook pages.
Interactivity. In terms of the interactivity, unclear results have been found across
several studies. Men and Tsai (2012) found that certain interactivity features (for example, action features) were used more by American companies on Facebook than by Chinese ones on Renren (the Chinese equivalent to Facebook). Waters and Lo (2012) could not fully support the notion that collectivistic HC cultures (China and Turkey) would be less engaging on Facebook than the individualistic LC American organisations, because Chinese
organisations were unexpectedly more engaging (Waters & Lo, 2012). However, Correia Loureiro and Concalves Gomes (2016) found that normative Portuguese companies use more interactivity features on Facebook than the pragmatic-oriented Brazilian companies. Knowing that Colombia scores even lower than Portugal in terms of Hofstede’s (2017a) long-term orientation dimension (see illustration 1), it is assumed to find similar findings. Therefore, the final relationship cultivation hypothesis reads as follows:
12
H3: Compared with German and British organisations, multinationals post more
interactive messages in their Colombian Facebook messages.
Online Consumer Engagement
With the new interaction dynamics enabled through SNSs, the focus of interest has been placed on customer behaviour beyond the actual purchase. The term online consumer engagement has become a central construct to describe the interactions between consumers and brands on SNSs (Cvijkj & Michahelles, 2013). Measuring the number of likes, shares and comments has proven meaningful and practical to evaluate consumer engagement on Facebook (Hyesun, Ahn, & Yu, 2016). De Vries et al. (2012) interpreted the like and
comment counts of organisational messages and define them as an indication for the general popularity of a post. The more popular a post, the wider the audience that it reaches on Facebook and thus the better it is for the organisation (Hyesun et al., 2016).
Much research has analysed what kind of organisational messages receive the most likes or comments (De Vries et al., 2012). However, there are also considerable cultural differences in terms of how consumers prefer to interact and behave on SNSs (Tsai & Men, 2017). Tsai and Men (2017) examined consumer engagement motivation in a cross-cultural context, finding evidence supporting the assumption that culture indeed influences consumer engagement activities on SNS brand pages. Their results showed that HC-oriented Chinese users engaged (liked, shared and commented) significantly more with the brand on the Chinese SNS compared with the LC-oriented American users (Tsai & Men, 2012).
Consumer engagement on interactivity strategy posts. De Vries et al. (2012) showed
that interactive messages containing questions or contests influence the number of received likes. Call for actions – such as explicitly asking the users to do something in the name of the brand (e.g. sharing or commenting on the post) – are actively incorporated in organisational
13 messages as they also positively influence consumer engagement (Hyesun et al., 2016). Hyesun et al.’s (2016) findings also indicate that content-related rewards (winning
something) also positively influence the number of likes. The interactivity strategy proposed in this paper measures questions directed towards users, contests, call for actions and
giveaways (see variable explanation in appendix A). Furthermore, Hofstede’s (1980) dimensions of national culture also apply to individuals and not only to organisational behaviour. As this study expects multinational companies to communicate according to Hofstede’s (1980) cultural dimensions of individualism, it is also assumed for users to behave according to the indulgence dimension. As Colombians tend to follow their impulses and are more willing to follow their desires (Hofstede, 2017a), it would be reasonable for engaging posts following the interactivity strategy to receive more likes, shares, a higher number of comments and overall more positive comments from Colombian users:
H4: Compared with British and German users, Colombian users:
a) give more likes on interactive organisational posts;
b) share interactive organisational posts more;
c) leave more comments overall on interactive organisational posts; and
d) write more positive comments on interactive organisational posts.
Consumer engagement on information dissemination strategy posts. Hyesun
et al. (2016) attested that the more information that a post contains about the organisation, the more likes and comments are given. The strategy of information dissemination measures how information is transmitted to the users through photos, news links, videos or even announcements (see variable explanation in appendix A). For these individual features, the amount of consumer engagement has been measured in several papers. Findings from Rauschnabel, Praxmarer and Ivens (2012) demonstrate that in general the more media elements (videos, pictures, graphs) that are included in
14 Facebook posts, the more likes that are given. De Vries et al. (2012) also found in their
study that posts with videos increased the like count, although news links or photos neither influenced the number of likes nor comments. Even though differing findings have been found, informational messages are commonly used by practitioners (Correia Loureiro & Concalves Gomes, 2016) and thus are also analysed for this study.
The hypothesis has previously been formulated that multinationals post more messages that follow the information dissemination strategy on German and British Facebook pages due to their individualistic LC cultural orientation. Although Germans are very restrained and composed compared to Colombians (Hofstede, 2017b), it could be possible that that regarding topics that are more informative their engagement will be higher. Explicit posts disseminating information about the company and products could be more popular in individualistic societies, where ambiguous messages are less commonly used (Men & Tsai, 2012). Although the British are not as restrained as Germans, it could be the case that due to their assumed same preferences for clear information, their consumer engagement will be higher regarding the same messages.
H5: Compared with Colombian users, German and British users:
e) give more likes on information dissemination posts;
f) share information dissemination posts more;
g) leave more comments overall on information dissemination posts; and
h) write more positive comments on information dissemination posts.
Methods
Research Design
In order to answer the research question and the hypotheses, a quantitative content analysis was conducted, in a similar vein as previous studies focusing on the same subject
15 (Men & Tsai, 2012; Waters et al., 2009). Organisational Facebook posts from six global brands were analysed based on formal features, using a codebook (see appendix B).
Sample and Procedure
The overall sample comprises 900 coded organisational Facebook posts and an additional range of metrics of each post. The global brands analysed are from the food and beverage sector, the car industry and the technology branch. The companies were chosen based on Interbrand’s listing of the top 100 most successful global brands in 2016
(Interbrand, 2016). Therefore, the first six companies that had a Facebook page in all three relevant countries – Colombia, Germany and the UK – were chosen. These companies are Coca-Cola, Microsoft, Samsung, Mercedes Benz, BMW and McDonald’s.
For each company in each country (n = 150), the 50 most recent organisational posts and the full related statistics were downloaded. Every company has an individual Facebook page for each of the three countries for which the posts were downloaded. The data was retrieved with Netvizz version 1.44, an app that allows extracting data from Facebook groups and pages for research purposes (Netvizz, 2017). For this paper, only messages that were posted by the page – namely coming from the company – with full statistics about the number of likes, comments and reactions were downloaded via Netvizz. This resulted in one dataset per country, which added up to three distinct datasets per organisation (n = 150). The posts (N = 900) received on average 1,504 likes and 396 comments. Every post featured a unique identifier and a link to track it back. The messages of the corporate posts were coded with a codebook (see appendix B). All of the variables in the codebook are based on previous studies, although not all of them were used for the analysis due to restructuration of the paper and analysis. Additionally, a sentiment analysis with the SentiStrength software was run for the comments, which allowed assessing how positive or negative the comments were.
16
Intercoder Reliability Test
The intercoder reliability was measured using two statistical reliability tests to compare the two coders, who coded 10% of the total amount. The Kappa values of the
intercoder reliability varied between κ = .634 and 1.00. The Percent Agreement scores ranged from 78% agreement to 100%. Both indices take values from .00 (no agreement) up to 1.00 (perfect agreement), respectively from 0% up to 100%. In general, for intercoder reliability indices values above .80 are considered acceptable or good (Lombard et al., 2002). Some variables did not receive a Kappa value in SPSS because there was no variance between the two coders, meaning that both coded the exact same value (see table 1). The consumer engagement variables (likes, shares, comments, sentiment) did not need a intercoder reliability because they were retrieved through specific software. After the intercoder
reliability test, the codebook was enhanced with examples and more information based on the actual data to increase the clarification.
Table 1.
Intercoder Reliability Test
Variables Cohen's Kappa p % agreement
Organization 100%
Country 100%
Message Disclosure - Description 100%
Message Disclosure - History 100%
Message Disclosure - Miss-Stat 100%
Information Dissemination - Photographs 1 .000 100%
Information Dissemination - NewsLinks 100%
Information Dissemination - Videos .956 .000 96%
Information Dissemination - Announcements/PR 100%
Interactivity - Sharing to one's page 1 .000 100%
Interactivity - Navigation .908 .000 96%
Interactivity - Partnership/Giveaways .847 .000 98%
Interactivity - Call for Action .634 .000 94%
17
Independent Variable
Country. The independent variable is country (measurement level
nominal-categorical). The chosen countries are Colombia, Germany and the UK. Additionally, the country was recoded into three dichotomous variables to allow different forms of analyses.
Dependent Variables
The detailed description for the variables of organisational disclosure, information dissemination and interactivity are provided in appendix A.
Organisational Disclosure. Previous studies have analysed this at the page level,
comparing individual features on the companies’ Facebook “about” sections (e.g. logo, URL, company history, description, contact information, mission statement) (Men & Tsai, 2012; Waters & Lo, 2012). This study reformulated three out of these five original variables to measure organisational disclosure at the message level, where the actual interaction with the consumers takes place: message description (n = 3), message history (n = 8), message
mission statement (n = 16). In order to measure message description as a whole construct, an
additive ordinal-scaled variable was recomputed: message disclosure (M = .03, SD = .17). Nonetheless, even though message disclosure theoretically ranged from 0 to 3 (0 = no organisational disclosure in the message, 3 = all message disclosure levels used), only the values 0 (n = 873) and 1 (n = 27) had cases. This is the reason why this variable is treated with a nominal measurement level.
Information Dissemination. The dichotomous variables to measure this construct
were taken from previous studies (Men & Tsai, 2012; Waters & Lo, 2012). These are:
photographs (n = 546), news links (n = 6), videos/audios (n = 338) and announcements/PR (n
= 11). On Facebook organisations have the possibility to post photos and videos to enhance the impact of their posted information (Men & Tsai, 2012). Previous studies have treated these two variables as information dissemination and not as interactivity features because
18 they can also be viewed as information, just the same as announcements and news links (Correia Loureiro & Concalves Gomes, 2016; Men & Tsai, 2012). As both news links and announcements and PR had the lowest number of cases, they were recoded into a new dummy variables news links and announcements (n = 17). The three variables for this
strategy were measured individually because pictures and videos are formal features, whereas announcements and press releases cover the actual content of the message.
Interactivity. The dichotomous variables that measured interactivity were mainly based
on Men and Tsai’s (2012) study: sharing to one’s own page (n = 397), navigation (n = 108),
solicitation of response (n = 176). Two additional variables were modified and added to gain
a broader range of message interactivity. Men and Tsai’s (2012) variable “action features for online participation” was modified into call for action (n = 139), based on Lovejoy and Saxton (2012). The variable partnership/giveaway (n = 93) was included based on Correia Loureiro and Concalves Gomes’ (2016) analysis. Previous studies have only explored which individual features were used more, but never at an accumulated level. For this paper, all of the aforementioned interactivity variables were recoded and added into a new six-point Likert scale variable interactivity (0 = no interactivity, 5 = all levels of interactivity). As the last two levels had very few cases (n < 5), the last three levels were recoded into one, resulting in a four-point Likert scale (M = .40, SD = .49). Due to the recoding, the variable interactivity has an ordinal measurement level, which will be also treated as close to metrical due to the equidistance of the items (Fields, 2009).
Consumer Engagement. The consumer engagement is measured with the following
four variables:
Likes. The number of likes (M = 1503.71, SD = 4415.09) that a post received overall
was automatically generated through the Netvizz download.
Shares. The number of times that an organisational post was shared (M = 204.54, SD
19
Comments. The total number of comments (M = 396.20, SD = 1654.95) that a post
received was retrieved with the Netvizz download.
Comment Sentiment. For the overall sentiment of the comments, a sentiment analysis
was run with SentiStrength, which allows estimating the strength of positive (1 = not
positive, 5 = extremely positive) and negative (-1 = not negative, -5 extremely negative) short texts, as well as differentiating them from neutral texts (SentiStrength, 2017). The sum of the positive and negative sentiments was added up and the mean for every message with
comments (n = 781) was calculated (M = 13.28, SD= 33.48).
Results
Comparing Relationship Cultivation Strategies
Hypothesis H1. A chi-square test of independence was performed to examine the
assumption that multinationals provide more organisational disclosure at a message level on their British and German Facebook pages compared with Colombian ones. The chi-square test of independence was statistically not significant (X2(2, N= 900) = 0.92, p = .632). This result indicates that multinationals do not provide different amounts of organisational disclosure at a message level, which leads to the rejection of H1.
Hypotheses H2. This hypothesis was tested with three sub-hypotheses, assuming that
multinationals disseminate more information on German and British Facebook pages than Colombian ones, with pictures (H2a), videos (H2b) and news links/announcements (H2c). The three sub-hypotheses were all tested with a binary logistic regression, taking the variable Colombia as the constant:
For H2a, the logistic regression model was significant as a whole (X2(2) = 14.528, p = .001, N = 900), with a weak Nagelkerke’s R2 of .02. Overall, the model predicted 60.7% of
20 the pictures correctly. Table 2 shows that the use of pictures in information dissemination posts significantly differs in Germany compared with Colombia. The odds ratio of an information dissemination post with a picture increases 0.71 times if the multinational posts on a German Facebook page. Based on this result, H2a can only be partially accepted for Germany and Colombia. For the UK and Colombia, the hypothesis has to be rejected. Table 21.
Summary of Binary Logistic Regression Analysis Predicting Information Dissemination with Pictures.
The regression model for H2b was statistically significant (X2(2) = 26.438, p < .001, N = 900). The model predicted 62.4 % of the cases correctly and had a weak R2 of .04
(Nagelkerke). Posts on German Facebook pages significantly varied compared with Colombia in terms of disseminating information with videos (see table 3). The odds ratio decreases 0.53 times, showing that the likelihood of German posts containing a video is lower than in Colombia. This stands in exact opposition to what H2b assumed. Moreover, no
statistically significant difference between the UK and Colombia was found. Due to these results, H2b has to be rejected.
1 All the binary logistic regression tables follow the APA format suggested by: Adelheid, AM Nicol, and M.
Pexman Penny. (2012) Presenting your findings: A practical guide for creating tables.
Germany 0.53 0.17 1.71 [1.22, 2.39] 9.74 .002
UK -0.05 0.16 0.95 [0.69, 1.31] 0.11 .742
Constant 0.28 0.12 1.33 5.84 .016
Predictor B SE OR 95% CI Wald statistic p
21 Table 3.
Summary of Binary Logistic Regression Analysis Predicting Information Dissemination through Videos.
H2c had to be rejected because the regression model was statistically non-significant (X2(2) = 2.14, p = .344, N = 900). No differences in the number of information-disseminating posts with news links and announcements could be found across the three countries, leading to a rejection of H2c.
Hypothesis H3. A one-way ANOVA was performed to test whether multinationals
post more interactive messages on Colombian Facebook pages compared with Germany and the UK. Levene’s F test showed that the homogeneity of variances assumption was not met with Levene’s F(2, 897) = 12.03, p < .001 and thus the Welch’s F test was performed.
Welch’s correction resulted in an insignificant Welch F(2, 587.90) = 1.90, p = .150, ɳ2 = .00. No significant difference between the countries and the number of interactive messages could be verified, which leads to a rejection of H3.
Comparing Consumer Engagement across Countries
Hypotheses H4. Cultural consumer engagement differences in interactive posts was
measured with four sub-hypotheses. For each of the four sub-hypotheses, a one-way ANOVA was conducted. Levene’s F test of homogeneity of variance was significant for all four tests,
Germany -0.76 0.18 0.47 [0.33, 0.66] 18.88 .000
UK 0.03 0.16 1.03 [0.74, 1.42] 0.03 .869
Constant -0.28 0.12 0.75 5.84 .016
Predictor B SE OR 95% CI Wald statistic p
22 indicating that equal variances across the three country groups could not be assumed. This is why the Welch correction was analysed throughout.
The Welch correction for H4a (Welch’s F(2, 383.59) = 8.34, p < .001) – which assumed that Colombian users like more interactive posts than the users in the other two countries – was significant. Welch’s F test indicates that the means of the number of likes significantly differed across the three countries. The effect size was small (ɳ2 = .04), meaning that the difference was weak. Table 4 shows the descriptive statistics for the groups. A Games-Howell post-hoc test was run to detect the specific differences between the groups. The results showed that Colombian users gave significantly more likes than Germans
(MDifference = 1838.43, p = .004). Colombians also gave significantly more likes on interactive
posts compared with British users (MDifference = 2214.82, p < .001). In coherence with the
findings, H4a can be fully accepted.
Again, the Welch correction for H4b – predicting that Colombian users share
interactive posts more than the other two nationalities – was statistically significant but small (Welch’s F(2,297.06) = 3.80, p = .023, ɳ2 = .02). Table 4 shows the descriptive statistics for the groups. To evaluate exactly between which countries the mean of the number of shares differed, a Games-Howell post-hoc test was run, indicating that Colombians shared
interactive posts significantly more than German users (MDifference = 496.18, p = .026). No
significant differences in the mean number of shares was found between Colombia and the UK (p = 0.75). Therefore, H4b can only be partially accepted for Colombia and Germany but not for Colombia and the UK.
The overall number of comments was assumed to be higher on Colombian interactive posts compared with German or British ones (H4c). Welch’s F test was not significant (Welch’s F(2,404.95) = 1.44, p = .239, ɳ2 = .01) for this hypothesis. Accordingly, no difference in the number of comments could be tested, leading to a rejection of H4c.
23 Welch’s F test was significant (Welch’s F(2,287.68) = 13.78, p < .001, ɳ2 = .04) for H4d, which argued that Colombian users comment more positively on interactive posts compared with the other two nationalities. This indicates that the mean sentiment of the comments significantly differed weakly across the countries (see descriptive statistics in table 4). A following Games-Howell post-hoc test showed that Colombians only posted
significantly more positive comments on interactive posts than Germans (MDifference = 0.19, p
< .001), whereas they did not comment more positively than British ones (p = .966). Surprisingly, British users were found to comment significantly more positively than
Germans (MDifference = 0.18, p = .001). These results support H4d only partially for Colombia
and Germany. Table 4.
Means and Standard Deviations for the likes, shares, comments, and positive comments for interactive posts
Hypotheses H5. Consumer engagement differences on information dissemination
posts was also measured with four sub-hypotheses. For each of the four sub-hypotheses, a one-way ANOVA was conducted. For all of the sub-hypotheses, Levene’s F test of homogeneity of variance was significant, proving that equal variances across the three country groups could not be assumed. Instead, the Welch correction was analysed throughout. M SD M SD M SD Likes 2879.08 7751.55 1040.65 2647.24 664.26 2025.60 Shares 563.87 2705.17 67.69 309.67 129.67 837.96 Number of Comments 631.71 2533.87 355.94 1579.29 292.01 1314.92 Positive comments 0.30 0.51 0.10 0.28 0.28 0.58
Colombia Germany United Kingdom
24 The Welch F test for H5a – predicting that assumed German and British users like more information dissemination posts than Colombians – was statistically significant (Welch’s F(2,549.33) = 5.19, p = .006, ɳ2 = .02). This indicates weak differences in the average number of likes for information dissemination posts across the three countries (see descriptive statistics in table 5). A Games-Howell post-hoc test showed that Colombian users liked these posts significantly more than German users (MDifference = 1281.30, p = .006).
Colombians also gave more likes on information dissemination posts compared with British users (MDifference = 1288.67, p = .006). These findings are the exact opposite of the prediction,
leading to a rejection of H5a.
H5b – which claimed that German and British users share information dissemination posts more than Colombian users – had very weak significant differences between the groups (Welch’s F(2,458.01) = 4.48, p = .012, ɳ2 = .01). Table 5 shows the descriptive statistics for the groups. The Games-Howell post-hoc test showed only one significant difference, whereby Colombian users shared significantly more information dissemination posts than Germans (MDifference = 324.26, p = .037). By contrast, no significant difference was found
between Colombia and the UK (p = .184). These results do not support H5b.
The overall number of comments was alleged to be higher on German and British information dissemination posts than in Colombia in H5c. Welch’s test was insignificant (Welch’s F(2,562.71) = 0.19, p = .827, ɳ2 = .00), proving that there is no significant difference between the countries in terms of the mean count of comments on information dissemination posts. Based on these findings, H5c was rejected.
Finally, H5d predicted that German and British users leave more positive comments on information dissemination posts compared with Colombians. Welch’s F test was
significant (Welch’s F(2,438.48) = 19.749, p < .001, ɳ2 = .04), meaning that small differences in the mean sentiment were found between the countries. Table 5 shows the descriptive
25 statistics for the groups. The subsequent Games-Howell post-hoc test demonstrated two significant differences, whereby Colombian users commented more positively on information dissemination posts than Germans (MDifference = 0.17, p < .001) and British users left more
positive comments on information-disseminating posts compared with Germans (MDifference =
0.17, p < .001). British and Colombian users did not differ in terms of the sentiment of their comments (p = 1.00). These results do not support H5d.
Table 5.
Means and Standard Deviations for the likes, shares, comments, and positive comments for information dissemination posts
Discussion
The goal of the present study was to understand how the use of online relationship cultivation strategies and the corresponding consumer reactions on multinational corporate Facebook pages differ between cultures.
The first set of hypotheses addressed the three online relationship cultivation strategies. Overall, multinationals in Colombia, Germany and the UK did not differ in the amount of organizational disclosure in posts, disseminating information with news links and announcements and in terms of interactivity. The only pair of countries that differed in terms of disseminating information with pictures and videos were Colombia and Germany. The
M SD M SD M SD
Likes 2378.44 6629.22 1097.14 2511.57 1089.78 2757.29
Shares 394.37 2246.82 70.12 291.67 153.45 739.30
Number of Comments 438.54 2110.63 408.90 1542.17 359.90 1232.09
Positive comments 0.28 0.47 0.11 0.26 0.28 0.52
26 findings concerning the relationship strategies on Facebook contribute to the theory of cross-cultural online relationship cultivation in several ways, as outlined below.
First, this paper is a pioneer in terms of measuring organizational disclosure at the message level. Actively informing about the history and mission of the organization could be more fruitful to build consumer relationships than simply filling out this information in the “about” section of the brand page, a part that users might not actively seek because the six organizations of this study are globally known. This global cognition of the brands might also be a reason why the multinationals are less actively disclosing information about themselves. Another reason could be that organizations already assume general knowledge about the organization among their brand page followers. The little amount of organizational disclosure implemented in corporate messages did not vary across the three countries. This finding indicates that organizations are not keen in providing information about themselves through messages, but rather prefer to post about other content. Future research should try to verify these findings across other cultures and aim to compare it with smaller local organizations to ascertain whether there are differences across multinationals and small and medium-sized enterprises.
Second, the unclear results of this analysis for the information dissemination strategy with pictures, videos, and news links and announcements is in line with the previous unclear results from Men & Tsai (2012) and Waters & Lo (2012). Both previous studies provide contrasting findings for the individual dimensions of information dissemination. The aim of this paper was to clarify the previous findings, although no coherent results were found because the individualistic British posts contained just as many pictures, videos and news links and announcements as collectivistic Colombian posts. This paper enhances the cross-cultural research in this field, providing evidence that other cross-cultural factors rather than simply Hofstede’s (1990) dimension of individualism may play a role in corporate information
27 dissemination strategies. Accordingly, other dimensions of culture should be taken into consideration for future research.
Third, of the three papers that previously analysed the interactivity strategy in a cross-cultural context, only Men and Tsai’s (2012) study supports the notion that individualistic cultures make more use of the interactivity strategy. Waters & Lo (2012) found contradicting results: one of their findings verified differences in the usage of the interactivity strategy between Turkey and the US, although their findings failed to verify the differences in interactivity strategies between the US and China. Colombia is a collectivistic culture, just like Turkey and China and was found to be just as interactive as the individualistic countries. As Water and Lo (2012) accentuate, the internet and especially SNSs like Facebook allow cultures to communicate and behave according to their cultural values, although it also gives room for the emergence of online cultures, where cultural norms shift. This assumption is enforced by Pfister and Soliz (2011), who also argue that intercultural communication is altered through SNSs like Twitter, allowing new public spaces to evolve, where virtual communication challenges traditional forms of communication. This argumentation could be plausible for the findings of this study because no significant differences could be established in the use of the interactivity strategy between Colombia, the UK and Germany. Therefore, future research should take this into account. Including more countries in the analysis could provide more insight into whether this difference holds true across further individualistic and collectivistic countries or if it simply applies to these three cultures.
The next set of hypotheses measured the consumer engagement for posts following the interactivity strategy. Colombians liked, shared and commented more positively on interactivity posts compared with Germans, while they also gave more likes than British user. The users in the three countries did not strongly differ in the number of comments.
28 indicating that the assumed differences in terms of engagement between these two cultures might be less strong than argued.
The last set of hypotheses tested the consumer engagement for posts following the information dissemination strategy. Germans and British users did not engage more on these posts than Colombians. Indeed, the opposite was the case, whereby Colombians liked, shared and commented more positively than Germans on information-disseminating posts. No differences in the number of comments were found between the countries and the British users did not differ from Colombians in terms of consumer engagement. In addition, the comments were more positive on British information dissemination posts compared with German ones, indicating surprising differences in the consumer engagement behaviour between the two individualistic countries.
One possible explanation for the similar consumer engagement behaviour for British and Colombian users on interactive and information disseminating posts is that although the UK stands in complete contrast to Colombia in terms of the individualism and
communication context style (Hofstede, 2017; Gudykunst et al. 1996), they are on equal levels on the indulgence dimensions (Hofstede, 2017). Just like Colombians, the British enjoy realising their impulses and desires in terms of enjoying life and having fun (Hofstede,
2017c). Moreover, Facebook is a standardized SNS, meaning that the page structure and possibilities of consumer reactions to corporate posts are the same worldwide, namely liking, commenting and sharing (Cvijkj & Michahelles, 2013). Given the happy and indulgent nature of Colombians and British users, it makes sense that they would also give in more easily to liking or sharing a post, whereas the restrained Germans would think twice about their interaction with the organization. Although evidence has been found that coming from a certain culture influences one’s online behaviour (Tsai & Men, 2017), the reasons for engaging with and reacting to corporate messages requires more clarification. Consumer
29 reactions cannot simply be explained by cultural differences, but rather should be analysed in terms of the message content in future research.
Practical Implications
Practitioners should implement more organizational disclosure at the message level to share their values, history and mission to create enhance the relationship-building with their followers. Although the results indicate that Facebook is a virtual culture where
communication becomes similar, organizations should still care about cultural differences. As Ess and Sudweeks (2006) argue, practitioners should strive for a global strategy while
emphasizing cultural differences, because previous studies have indicated that organisations are more successful when emphasizing cultural differences on social media (Ess &
Sudweeks, 2006; Stelzner, 2011)
Limitations and Future Research
The findings of this study should be interpreted with caution, given the results only apply to the six multinationals that were analysed. The subject of analysis only comprised three countries, which have never previously been compared in terms of online relationship cultivation strategies and consumer engagement on Facebook. In order to make general statements about how individualistic- and collectivistic-oriented cultures really implement their online relationship strategies, future research should analyse more countries to better evaluate how they influence corporate SNS use and consumer reactions.
Limitations in terms of the online relationship cultivation strategies also have to be pointed out. The second strategy – information dissemination – should address the needs, concerns and interest of the public (Men & Tsai, 2012), meaning that organizations should think about what information is important, relevant and demanded among their brand page followers. However, this construct has been measured with some contradicting variables such
30 as posted pictures, videos, news links and announcement and press releases. Although the variables were completely based on previous literature, they do not seem to measure the same content. Pictures and videos are only features of a message, which could be entertaining, educational, product-related or even unrelated to the company. On the other hand,
announcements and press releases are features that are content-related and indicate that that the topic is organizational, whereas pictures and videos can be in any given organizational message for any given topic. Future research should try to reformulate and specify the variables to measure what the message content is and whether it addresses organizational accomplishments and information or if it only promotes entertaining and unrelated content.
Furthermore, consumer engagement in the form of the number of likes, shares, comments and sentiment are not indicators measuring the effort of the organization to actually engage in a conversation with the audience. To evaluate whether the online relationship cultivations strategies really improve dialogue with the public, it should be analysed whether organizations actually reply to user comments and not only one-sidedly send out their messages.
Finally, the content analysis used for this study is a descriptive research technique that does not allow further interpretation (Weber, 1990).
Conclusion
In conclusion, the findings of this study indicate that the information dissemination and interactivity strategies proposed by Waters et al. (2009) are frequently used by
multinationals across all countries, albeit mostly not according to the expected cultural values. In all of the cultures, the multinationals tried to disseminate information about the company or products using strong visuals and trying to interactively engage with their audience. Research has shown that consumer engagement depends on the organizational messages but could also be greatly influenced on the cultural background. The current paper
31 contributes to the research on online relationship cultivation strategies by proving that
Facebook allows for new overarching virtual cultures to evolve, where the same communication patterns are used (Waters & Lo, 2012).
32 References
Adelheid, AM Nicol, and M. Pexman Penny. (2012) Presenting your findings: A practical guide for creating tables. Washington DC: American Psychological Association. Bortree, D.S. & Seltzer, T. (2009). Dialogic strategies and outcomes: An analysis of
environmental advocacy groups’ Facebook profiles. Public Relations Review, 35, 317-319. Doi: 10.1016/j.pubrev.2009.05.002
Correia Loureiro, S.M. & Gonçalves Gomes, D. (2016). Relationship Between Companies and the Public on Facebook: The Portuguese and the Brazilian Context. Journal of
Promotion Management, 22(5), 705-718. Doi: 10.1080/10496491.2016.1185493
Cvijikj, I. P., & Michahelles, F. (2013). Online engagement factors on Facebook brand pages. Social Network Analysis and Mining, 3(4), 843-861.
Doi: 10.1007/s13278-013-0098-8
De Mooij, M. & Hofstede, G. (2010). The hofstede model. International
Journal of Advertising, 29(1), 85-110, Doi: 10.2501/S026504870920104X
De Mooij, M. & Hofstede, G. (2011). Cross-cultural consumer behavior: A review of research findings. Journal of International Consumer Marketing, 23(3-4), 181-192. Doi: http://dx.doi.org/10.1080/08961530.2011.578057
De Vries, L., Gensler, S., & Leeflang, P. S. (2012). Popularity of brand posts on brand fan pages: An investigation of the effects of social media marketing. Journal of
interactive marketing, 26(2), 83-91. Doi: https://doi.org/10.1016/j.intmar.2012.01.003
Ess, C., & Sudweeks, F. (2006). Culture and computer-mediated communication: Toward new understandings. Journal of Computer-Mediated Communication, 11(1), 179–191. Doi: 10.1111/j.1083-6101.2006.tb00309.x
Field, A. (2009). Discovering statistics using SPSS. Third edition. London: SAGE Publications Ltd.
Gudykunst, W.B. (1997). Cultural Variability in Communication. An Introduction.
Communication Research, 24(4), 327-348.
Doi: https://doi.org/10.1177/009365097024004001
Gudykunst, W.B., Matsumoto, Y., Ting-Toomey, S., Nishida, T., Kim, K., & Heyman, S. (1996). The Influence of Cultural Individualism-Collectivism, Self Construals, and Individual Values on Communication Styles Across Cultures. Human Communication
Research, 22(4), 510-543. Doi: 10.1111/j.1468-2958.1996.tb00377.x
33 Hermeking, M. (2005). Culture and internet consumption: Contributions from cross-cultural
marketing and advertising research. Journal of Computer-Mediated Communication,
11(1), 192–206. Doi: 10.1111/j.1083-6101.2006.tb00310.x
Hofstede, G. (1980). Culture’s consequences: International differences in work-related values. Beverly Hills: Sage.
Hofstede, G. (2001) Culture’s consequences: Comparing values, behaviors, institutions, and organizations across nations. Thousand Oaks: Sage.
Hofstede, G., Hofstede, G. J., & Minkov, M. (2010). Cultures and organizations: Software of the mind (Rev. 3rd ed.). New York, NY: McGraw-Hill
Hofstede, G. (2017). The Hofstede Centre – Country Comparison. Retrieved from: https://geert-hofstede.com/countries.html
Hofstede, G. (16.05.2017a). The Hofstede Centre – Country Comparison. Retrieved from: https://geert-hofstede.com/colombia.html
Hofstede, G. (16.05.2017b). The Hofstede Centre – Country Comparison. Retrieved from: https://geert-hofstede.com/germany.html
Hofstede, G. (16.05.2017c). The Hofstede Centre – Country Comparison. Retrieved from: https://geert-hofstede.com/united-kingdom.html
Hyesun, J., Ahn, H.J., & Yu, G.J. (2016). What makes people react to the posts on the brand pages of mobile social network games? Online Information Review, 40(3) pp. 435-448. Doi: http://dx.doi.org/10.1108/OIR-07-2015-0236
Interbrand. (13.06.2017). Best global brands 2016 rankings. Retrieved from: http://interbrand.com/best-brands/best-global-brands/2016/ranking/ Kaplan, A.M, Haenlein, M. (2010). Users of the world, unite! The challenges and
opportunities of social media. Business Horizons, 53, 59-68. Doi:10.1016/j.bushor.2009.09.003
Kent, M. L., & Taylor, M. (1998). Building dialogic relationships through the World Wide Web. Public Relations Review, 24, 321–334. Doi: https://doi.org/10.1016/S0363-8111(99)80143-X
Luarn, P., Lin, Y.P., Chiu, Y.P. (2015). Influence of Facebook brand-page posts on online Engagement. Online Information Review, 39(4), 505 – 519.
Doi: http://dx.doi.org/10.1108/OIR-01-2015-0029
Lombard, M., Snyder-Duch, J., Campanella Bracken, C. (2002). Content analysis in mass communication: Assessment and reporting of intercoder reliability. Human
34 Lovejoy, K. & Saxton, G.D. (2012). Information, community, and action: How
nonprofit organizations use social media. Journal of Computer-Mediated
Communication, 17, 337–353. Doi: 10.1111/j.1083-6101.2012.01576.x
Malhotra, A., Malhotra, C. K., & See, A. (2013). How to create brand engagement on Facebook. MIT Sloan Management Review, 54(2), 18.
Men, L.R., & Tsai, W.H.S. (2012). How companies cultivate relationships with publics on social network sites: Evidence from China and the United States. Public Relations Review,
38, 723-730. Doi:10.1016/j.pubrev.2011.10.006
Netvizz (13.06.2017). Netvizz. Retrieved from:
https://tools.digitalmethods.net/netvizz/facebook/netvizz/faq.php
Nishimura, S., Nevgi, A., & Tella, S. (2008). Communication style and cultural features in high/low context communication cultures: A case study of finland, japan and india.
Teoksessa A. Kallioniemi (toim.), Uudistuva ja kehittyvä ainedidaktiikka. Ainedidaktinen symposiumi,8, 783-796.
Pfister, D. S., & Soliz, J. (2011). (Re)conceptualizing intercultural communication in a networked society. Journal of International and Intercultural Communication, 4(4), 246-251. Doi: 10.1080/17513057.2011.598043
Rauschnabel, P. A., Praxmarer, S., & Ivens, B. S. (2012). Social media marketing: How design features influence interactions with brand postings on Facebook. In Eisend, M., Langner, T., & Okazaki, S. (Eds.), Advances in Advertising Research (Vol. III) (pp. 153-161). Gabler Verlag.
Richardson M. R. & Smith, S. W. (2007). The influence of high/low-context culture and power distance on choice of communication media: Students’ media choice to communicate with professors in japan and america. International Journal of
Intercultural Relations, 31, 479–501. doi: 10.1016/j.ijintrel.2007.01.002
Sentistrength (14.06.2017). Retrieved from: http://sentistrength.wlv.ac.uk/
Stelzner, M. (2011). 2011 social media marketing industry report. Retrieved September 30, 2011, Retrieved from:
http://www.socialmediaexaminer.com/social-media-marketing-industry-report-2011/
Singh, N. & Matsuo, H. (2004). Measuring cultural adaptation on the Web: a content analytic study of U.S. and Japanese Web sites. Journal of Business Research, 57, 864-872. Doi: 10.1016/S0148-2963(02)00482-4
35 Thackeray, R., Neiger, B., Hanson, C., & McKenzie, J. (2008). Enhancing promotional
strategies within social marketing programs: Use of Web 2.0 social media. Health
Promotion Practice, 9(4), 338–343. Doi: https://doi.org/10.1177/1524839908325335
Tsai, W.H.S. & Men, L.R. (2017) Consumer engagement with brands on social network sites: A cross-cultural comparison of China and the USA, Journal of Marketing
Communications, 23(1), 2-21, Doi: 10.1080/13527266.2014.942678
Waters, R.D., Burnett, E., Lamm, A., & Lucas, J. (2009). Engaging stakeholders through social networking: How non-profit organizations are using Facebook. Public
Relations Review, 35, 102-106. doi:10.1016/j.pubrev.2009.01.006
Waters, R.D. & Lo, K.D. (2012). Exploring the impact of culture in the social media sphere: A content analysis of nonprofit organizations’ use of facebook. Journal of
Intercultural Communication Research, 41(3), 297-319. Doi:
http://dx.doi.org/10.1080/17475759.2012.728772
36 Appendix A
Table A1.
Description of the Online Relationship Cultivation Variables
Strategy Variable Description History Mission Statatement Photographs News Links Videos/Audios Announcements/PR
Sharing to one's page
Navigation
Partnership/Giveaways
Call for Action
Solicitaiton of Response Organizational Disclosure - Message Level Information Dissemination Interactivity
Any descritpion about the organization's programs, services or activities which adresses the company as a whole.
Descriptions of the establishment, founding year, their first shop, product, or mentioning of milestones.
Information which covers the organizations core purpose. Addressing the organization's good deeds like CSR actions and how they try to help the world.
Any form of question which seems like an interst, attempt, or effort from the side of the organization to get responses. Also polls, games, quizzes and contests aim for a solicitation of response.
Images that imply a connection to the company.
Hyperlinks which reffer to external media coverage, not the own website or blog.
Videos, Animations, Gifs or Audiofiles which imply a connection to the company.
A link to the company's website or blog. Description
Announcements of company activities that were conducted or are about to take place and which have a long-term effect or duration. Also direct download to press releases in the post.
A link that navigates to any external content which is not the own website/blog and not external media coverage.
Information about the brand partnering up with a social mdia influencer, blogger, another company and/or providing free giveaways (not inform of a contest).
Any request in the message which aims to animate the followers to do something in the name of the organization.
37 Appendix B
Codebook Aim of the study
The aim of this study is to analyze how cultural dimensions (HC/LC and I-C) influence online relationship cultivation strategies of the top 10 global brands who have Facebook brand pages in Colombia, Germany and the United Kingdom. A cross cultural comparison is conducted in the previously mentioned countries and the resulting consumer engagement of these relationship cultivation strategies is measured.
Selection criteria and coding instructions
• The companies were already selected beforehand and are listed in the codebook. • The links provided below, allow a fast access to the specific organizational Facebook
page for each of the three countries.
• There are two levels of analysis. The first one is on the Facebook page level and the secons one is on the message/post level.
• In a first step, keep the Facebook page in your browser open to access information about followers, likes, organizational information about the disclosure of the organization.
• As a second step, code the information about the messages/posts, which were all downloaded into an Excel sheet. On that sheet information about the post type, post id, content/message, likes and reactions is provided and does not have to be recoded. • In case of doubt, the link to the specific post is also available in the Excel sheet and
should be consulted to be sure if it is coded right.
• If for certain reasons the variable cannot be coded and neither a yes or no seems to be fit the code 999 can be entered manually.
• The sentiment analysis for the comments will be computed automatically and does not have to be coded manually.
38 Links of organizations official Facebook page according to the country
01: Coca Cola https://www.facebook.com/cocacolacolombia/ https://www.facebook.com/CocaColaDE/ https://www.facebook.com/cocacolaGB/ 02: Microsoft https://www.facebook.com/MicrosoftUK/ https://www.facebook.com/MicrosoftDE/ https://www.facebook.com/MSFTColombia/ 03: Samsung https://www.facebook.com/SamsungColombia/ https://www.facebook.com/SamsungDeutschland/ https://www.facebook.com/SamsungUK/ 04: Mercedes Benz https://www.facebook.com/mercedesbenzdeutschland/ https://www.facebook.com/MercedesBenzUK/ https://www.facebook.com/MercedesBenzColombia/ 05: BMW https://www.facebook.com/BMWDeutschland/ https://www.facebook.com/bmwuk/ https://www.facebook.com/BMW.Colombia/ 06: McDonalds https://www.facebook.com/mcd/ https://www.facebook.com/McDonaldsColombia/ https://www.facebook.com/McDonaldsUK/
