Industrial Marketing Management: Qi Deng, Yun Wang, Michel Rod, Shaobo Ji

Industrial Marketing Management 99 (2021) 1–15
Contents lists available at ScienceDirect
Industrial Marketing Management

journal homepage: www.elsevier.com/locate/indmarman
Speak to head and heart: The effects of linguistic features on B2B brand
engagement on social media
Qi Deng a, b, Yun Wang c, *, Michel Rod d, Shaobo Ji c
a
Zhejiang Gongshang University, 18 Xuezheng Street, Xiasha University Town, Hangzhou 310018, China
b
Rowe School of Business, Dalhousie University, 6100 University Ave, PO Box 15000, Halifax, Nova Scotia B3H 4R2, Canada
c
Sprott School of Business, Carleton University, 7045 Nicol, 1125 Colonel By Drive, Ottawa, Ontario, K1S 5B6, Canada
d
Faculty of Business, University of New Brunswick, University of New Brunswick Saint John, P.O. Box 5050, Saint John, New Brunswick E2L 4L5, Canada
A R T I C L E I N F O A B S T R A C T
Keywords: Brand engagement on social media increasingly draws B2B brands’ attention as it may produce positive WOM
Brand engagement and bring branding and financial benefits. However, B2B marketers face challenges in creating compelling brand
B2B marketing posts on social media. Beyond ‘knowing what to post’, what is even more challenging for B2B marketers is a lack
Social media
of knowledge of ‘knowing how to communicate’, i.e., knowing how to design the non-informational cues in
Brand linguistics
brand posts to stimulate brand engagement and generate social media WOM. This research makes initial attempts
to address this gap by investigating the impacts of post language on B2B brand engagement on social media.
Building on the model of B2B effective communication and theories in linguistics, we identify six linguistic
features (i.e., post length, language complexity, visual complexity, emotional cues, interpersonal cues, and
multimodal cues in rich media) that influence brand engagement, captured using Twitter likes and retweets.
Through analyzing 229,272 tweets collected from 156 B2B brands in 10 industries, we found that, in general,
linguistic features that facilitate the central or peripheral route processing will have positive effects, while those
that hinder the processing will have negative impacts on brand engagement. This research contributes to our
knowledge of B2B social media communication by revealing the power of brand language in driving brand
engagement and introducing linguistics as a valuable conceptual lens for maximizing the benefits of B2B mar
keting content on social media. This research also highlights the interpretative nature of social media commu
nication – B2B brands must go beyond the content purpose and strategy decisions to consider the specific
language use and communication style of the message.
1. Introduction goods and services or their sellers” (Berger, 2014, p. 261), WOM can evoke
rational and emotional responses from the brand audience (Juntunen,
Business-to-business (B2B) firms are increasingly embracing social Ismagilova, & Oikarinen, 2020; Leek, Canning, & Houghton, 2016;
media for their marketing activities. According to a recent report, social Swani, Milne, Brown, Assaf, & Donthu, 2017), and consequently in
media has surpassed email, brand websites and blogs and become the crease brand equity, differentiate a brand from its competitors, create
top channel for marketing content distribution (Content Marketing brand outcomes and monetary value, and impact B2B firms’ stock per
Institute, 2021). As for paid content distribution, 83% of B2B marketers formance (Bruhn, Schoenmueller, & Schäfer, 2012; Krings, Palmer, &
used social media advertising or promoted posts in 2020, rising from Inversini, 2021; Liu, 2020; Lynch & De Chernatony, 2007; Pitt, Plangger,
60% in 2019 (Content Marketing Institute, 2021). It has been reported Botha, Kietzmann, & Pitt, 2019; Trusov, Bucklin, & Pauwels, 2009;
that 75% of B2B buyers and 84% of C-level or vice-president level buyers Zhang & Du, 2020). However, while social media is increasingly adop
are influenced by social media when making purchasing decisions ted in industrial communications and branding activities, B2B firms
(Articulate, 2019). Social media word-of-mouth (WOM) is deemed continue to encounter problems in integrating social media into their
important for B2B firms. Defined as “informal communications directed at marketing efforts (Järvinen & Taiminen, 2016), despite the fact that B2B
other consumers about the ownership, usage, or characteristics of particular brands’ social media presence has a positive impact on all four brand
* Corresponding author.
E-mail addresses: [email protected] (Q. Deng), [email protected] (Y. Wang), [email protected] (M. Rod), [email protected] (S. Ji).
https://doi.org/10.1016/j.indmarman.2021.09.005
Received 27 February 2021; Received in revised form 31 July 2021; Accepted 20 September 2021
Available online 29 September 2021
0019-8501/© 2021 Elsevier Inc. All rights reserved.
Q. Deng et al. Industrial Marketing Management 99 (2021) 1–15
relationship strength indicators - interactivity enhances perceived 2020) takes both rational and emotional appeals into account, thus
partner quality, while responsiveness positively influences commitment providing a solid foundation for a systematic investigation of the effects
(Karampela, Ewelina, & McLean, 2020). Among these problems, a major of linguistic features. By investigating 229,272 tweets collected from
difficulty lies in creating brand posts that stimulate brand engagement 156 B2B brands in 10 industries, we empirically identify six linguistic
and generate social media WOM. In fact, brand post creation was re dimensions, including post length, language complexity, visual
ported to be the activity with which B2B marketers have most difficulty; complexity, emotional cues, interpersonal cues, and multimodal cues in
regardless of company size and content marketing budget (Content rich media, that affect brand engagement on social media via either the
Marketing Institute, 2021). central route or the peripheral route of information processing. In this
Underlying the challenges in B2B social media post creation is, way, we contribute to B2B brand post and engagement research by of
firstly, the fact that extant literature and theories are mostly grounded in fering guidance on how to design engaging posts and communicate
the B2C area (Kaplan & Haenlein, 2010), and secondly, the lack of clear effectively on social media. Our findings also contribute to B2B mar
and comprehensive knowledge of ‘knowing how’ rather than ‘knowing keting practice by providing many easy-to-implement tactics that can
what’ (Leek, Houghton, & Canning, 2019; Nguyen, Yu, Melewar, & help B2B brands stimulate engagement and facilitate WOM on social
Chen, 2015). The limited research addressing B2B social media posts media.
mainly focuses on the performance of the posts, such as WOM referrals, The remainder of the paper is structured as follows. Section 2 begins
their subsequent brand outcomes (Bruhn et al., 2012; Pitt et al., 2019) with a discussion of social media in B2B marketing and the importance
and return-on-investment (Kumar & Mirchandani, 2012), and general of brand engagement, followed by a review of current research on B2B
management issues, such as multi-account/cross-platform management brand posts. We then identify research gaps, state our research position,
(Mehmet & Clarke, 2016; Zhang, Gosselt, & de Jong, 2020). With the and introduce our theoretical foundations. Sections 3 and 4 detail our
exception of some recent research that considers B2B and B2C post hypothesis development and methodology. Our empirical results are
comparison (Swani et al., 2017), the extant research focus has been on then presented in Section 5, followed by an in-depth discussion of both
‘knowing what’, i.e., understanding what informational contents (e.g., theoretical and managerial implications of the research and the limita
marketing themes/topics) to use in brand posts for particular B2B tions of the present study and opportunities for future research in Sec
marketing purposes. There is a paucity of research on ‘knowing how’, i. tion 6.
e., knowing how to design the non-informational cues (e.g., language/
tone/media) in brand posts to stimulate brand engagement and generate 2. Theoretical background
social media WOM (Juntunen et al., 2020; Leek et al., 2019; Nguyen
et al., 2015). However, ‘knowing how’ is undoubtedly important 2.1. B2B marketing and brand engagement on social media
because non-informational cues can influence brand engagement by
affecting both how the brand post is cognitively processed and the extent Among various social media applications in B2B marketing, the most
to which the audience is emotionally bonded to the brand (Juntunen promising one appears to be branding through B2B content strategies
et al., 2020; McShane, Pancer, & Poole, 2019). (Brennan & Croft, 2012; Juntunen et al., 2020; Leek et al., 2016; Leek &
In B2B marketing literature, there seems to be a debate on whether Christodoulides, 2011; Swani, Brown, & Milne, 2014). Research has
rational or emotional appeal works best in brand communication on suggested that, in B2B social media content marketing, the most
social media (Cortez, Gilliland, & Johnston, 2020). One perspective important activity is to increase brand engagement (Leek et al., 2019;
argues that the online context is not effective for creating an emotional McShane et al., 2019; Swani et al., 2017). In the digital age, multiple
bond (Barari, Ross, Thaichon, & Surachartkumtonkun, 2020; Steinhoff, stakeholders (external and internal) may all participate in social media
Arli, Weaven, & Kozlenkova, 2019), while another suggests that discussions, read brand posts, and become key influencers (Brennan &
including emotional appeals in B2B social media communication is Croft, 2012; Huotari, Ulkuniemi, Saraniemi, & Mäläskä, 2015). Brands’
effective for creating positive brand attitudes and behavioural engage success in getting various stakeholders involved and engaged on social
ment (Sundström, Alm, Larsson, & Dahlin, 2021; Swani et al., 2017). media (e.g., liking and sharing brand posts) can drive positive WOM for
Recent research suggests that both rational and emotional appeals are the brand and, consequently, raise awareness, generate leads, and build
accessible and processable to social media users who view B2B brands’ trust and credibility for the brand (Alboqami et al., 2015; Leek et al.,
posts (Cortez et al., 2020). Depending on personal characteristics (e.g., 2019; Lynch & De Chernatony, 2004; Swani et al., 2017; Trusov et al.,
expertise) and situational factors (e.g., business task involvement), 2009). This positive WOM has been proven to positively influence brand
stakeholders such as buyers, potential customers, professionals, partners outcomes and financial outcomes in the B2B context (Liu, 2020; Pitt
on social media may respond to B2B brand posts differently through two et al., 2019; Trusov et al., 2009). Furthermore, along with the stake
information processing routes, i.e., central route or peripheral route. holders’ engagement behaviours, much behavioural data is stored on
Whereas some stakeholders may focus on rational information con social media, such as the number of views, liking, and shares, click-
tained in brand posts and engage with brands driven by utilitarian through rates, and comments. The exploitation of this behavioural
motives, others may engage in seeking emotional benefits that are data can generate business insights concerning many B2B marketing
usually related to their underlying needs for personal expression, social tasks such as product/service improvements and competitor bench
approval, and self-esteem (Cutler & Javalgi, 1993). Therefore, we argue marking (Nam, Joshi, & Kannan, 2017; Quinton, 2013; Upreti et al.,
that it is necessary to develop a strategy for simultaneously delivering 2021). Hence, it is surely important for B2B brands to create compelling
both rational and emotional values through the brand post on social brand posts to better engage with a range of stakeholder groups on social
media. media (Brennan & Croft, 2012; Swani et al., 2017).
Used appropriately, language can improve information quality for
better cognitive processing and create socio-psychological meanings for 2.2. B2B brand posts on social media
interpersonal and emotional associations-building (Cortez et al., 2020;
McShane et al., 2019; Zhang & Du, 2020). As such, in this paper, we aim Thus far, research on brand posts on social media has mostly been in
to answer the following research question: how do linguistic features of the B2C context (Huotari et al., 2015; Kaplan & Haenlein, 2010; Pitt
B2B posts influence brand engagement on social media? We apply et al., 2019), while some studies compare B2C and B2B brand posts as a
Cortez et al.’s (2020) B2B communication model and investigate the focal point (Swani et al., 2014, 2017; Swani, Milne, & Brown, 2013).
effects of linguistic features of B2B posts on brand engagement by Only recently have scholars begun to specifically address the effective
considering their influence on central and peripheral information pro ness of brand posts on social media within the B2B arena. This work
cessing. The comprehensive B2B communication model (Cortez et al., concentrates on ‘knowing what’, namely, summarising major B2B
2
marketing themes/topics and corresponding informational contents of in information processing, such as cognition, affect, attention, and
the brand post for better engaging with social media audiences (Jun behavioural intention, but avoids a sequential ordering from attention to
tunen et al., 2020; Leek et al., 2019; Sundström et al., 2021; Zhang et al., behaviour. It is, therefore, appropriate to apply this model to investigate
2020). While there is no consistent and comprehensive summary B2B brand posts on social media because social media communication is
regarding the effective factual information in the B2B brand post, a regarded as a non-linear, co-creative process in which participants are
central underlying conclusion, however, appears to be that B2B brand deemed not always to be attentive (Mehmet & Clarke, 2016). The ELM is
ing lacks sufficient knowledge of how to create compelling social media a “dual-process theory of attitude formation and change arguing that
posts to communicate with the audience (Swani et al., 2017). persuasion can act via a central or peripheral route and that personal
Research suggests that, to maximise the WOM benefits of brand posts attributes determine the relative effectiveness of these processes” (Angst
on social media, ‘knowing how to design the messages’ is equally & Agarwal, 2009, p. 341). Thus, depending on the personal attributes
important as ‘knowing what messages to post’ (Nguyen et al., 2015). For and situational factors such as individual expertise, involvement, and
example, suitable post design and language use in a brand post can in personality, a range of stakeholder groups and the individuals in each
fluence the message quality (e.g., informativeness, variety, interactivity) group may respond to the same social media message differently
to affect the viewer’s cognitive processing of the informational content through the central route or the peripheral route (Cortez et al., 2020).
(Zhang & Du, 2020); the specific features such as embedded links, Moreover, the social media environment makes the situational factors
media, and hashtag are suggested to influence message fluency and then more complicated and leads to more dynamic individual motives for
brand engagement (McShane et al., 2019). Moreover, recent research brand engagement, including information seeking, socialising with
recognises the co-existence of cognitive and emotional dimensions un others, and satisfying socio-psychological needs such as self-assurance/
derlying the brand engagement concept (Money, 2004; Zhang & Du, − achievement (Kaplan & Haenlein, 2010).
2020) and suggests the positive effect of emotional appeals on brand Therefore, both rational appeals (i.e., focusing on practical, func
engagement (Leek et al., 2019; Pitt et al., 2019; Swani et al., 2017). Our tional, or utilitarian needs and emphasising informational features and
review finds that research on ‘what messages to post’ has examined post tangible benefits) (Grigaliunaite & Pileliene, 2016) and emotional ap
content and developed appropriate post themes, topics, and marketing- peals (i.e., relating to emotions and feelings and attempting to stir up
related discourses to establish an emotional appeal (e.g., building trust, emotions to motivate purchase) (Swani et al., 2014) have possibilities to
invoking empathy) (Juntunen et al., 2020; Sundström et al., 2021). be accessed and processed by the social media users in B2B settings.
Unfortunately, empirical evidence is currently scant regarding the effect Then, the effective communications of rational appeals through the
of non-informational cues on emotional appeals (Swani et al., 2017). central route and emotional appeals through the peripheral route could
This is considered a major oversight, given that non-informational ele enhance brand values, build brand associations, and activate engage
ments such as post-composition, language use, and communication style ment behaviours (Cortez et al., 2020; Lynch & De Chernatony, 2007). By
are useful in offering peripheral cues and creating new socio- recognising the effectiveness of both central route and peripheral route
psychological meanings for higher emotional involvement and engage processing, we argue that the real question for B2B marketers becomes
ment (Cortez et al., 2020). how to design/communicate the brand messages in ways that both
Therefore, in this paper, we investigate the non-informational ele central and peripheral routes could be facilitated to increase brand
ments of B2B brand post composition and design from a linguistic engagement. In particular, we propose that the key to the answer lies in
perspective. We observe that some research has touched upon the in the linguistic features of the brand post.
fluences of particular post elements on brand engagement based on a Regarding the linguistics aspect, we refer to Halliday’s (1976, 1978)
linguistic view, but this work referred to broad or vague umbrella terms theories on systemic functional linguistics and apply metafunction and
such as message styles, message quality, post tactics, and post features multimodality concepts to build our analytical foundation. According to
(Juntunen et al., 2020; Leek et al., 2016, 2019; McShane et al., 2019; Pitt metafunction theory, text can be understood through three perspectives,
et al., 2019; Zhang & Du, 2020) and this work usually addresses the ideational (including experiential and logical), interpersonal, and tex
effect of linguistic features on either rational appeals or emotional ap tual (Halliday, 1978). The ideational function of the text enables viewers
peals. As we argued, linguistic features have the potential for simulta to make sense of the social action, event, and activity from reading the
neously enhancing both cognitive information processing and emotional text, and the interpersonal function allows viewers to enact the social
response generation (Cortez et al., 2020; Mehmet & Clarke, 2016). relationships and feel the moods, emotions, and attitudes of others via
Therefore, we contend that a systematic investigation of the impact of the text. More importantly, the viewer’s sensemaking of the ideational
linguistic features on brand engagement in the B2B context by consid and interpersonal meanings is socially constructed and influenced by the
ering its influences on the communications of rational appeals and textual function of the language (e.g., text composition, grammar use)
emotional appeals is warranted. This is especially important as we (Mehmet & Clarke, 2016). Hence, linguistic features play an active role
acknowledge increased attention to the discursive elements of B2B in building and construing the social reality experienced by the viewers
communications in enhancing band legitimacy (Gustafson & Pomir (Halliday, 1978). By integrating metafunction theory with the effective
leanu, 2021). Moreover, we suggest that it would be most appropriate to B2B communication model, we suggest that proper use of textual met
apply linguistics theories to the dataset as the research aims to exploit afunction in a B2B brand post might enable the brand’s ideas in the post
language resources (Mehmet & Clarke, 2016). Next, we introduce our to be accurately processed and the brand-viewer social ties to be built
theoretical foundation, the comprehensive model of effective B2B and enhanced. Specifically, the textual metafunction influences the
communication (Cortez et al., 2020) and linguistics theories, including viewers’ cognitive and peripheral processing routes. Textual meta
metafunction and multimodality (Halliday, 1976, 1978). function denotes the mode/composition and linguistic components of
the message, which exhibits multimodality. Multimodality is defined as
2.3. Two routes of information processing & linguistic metafunction and communication that is not limited to written language but includes
multimodality various modes and literacies such as images, videos, and kinetic move
ment; namely, everything from the inclusion of emoticons to the
Gilliland and Johnston (1997) provided the first comprehensive placement of images to the organisation of the message to the method of
model of the effective B2B communication process; Cortez et al. (2020) delivery thus createing meaning for ideational and interpersonal inter
revisited and refined the model by incorporating the influences caused pretation (Mehmet & Clarke, 2016). Multimodality provides flexibility
by social media on industrial communication and the role of emotional for researchers with the ability to change and adapt depending on their
appeals in branding. The effective B2B communication model is based specific purposes (Jewitt, 2009). In our research, the following multi
on the Elaboration Likelihood Model (ELM), which involves all elements modal elements are incorporated as they are deemed salient in social
3
media communications: post length, post structure, word use, social logic is that greater post length leads to a higher level of post informa
media special symbols (e.g., hyperlink, URL, emoji, at-mention), and tiveness (amount of information contained in a message), which in turn
rich media (e.g., gif, image and video). We aim to investigate the impact influences brand followers’ engagement behaviours (McShane et al.,
of linguistic features (the various modal elements) on brand engagement 2019; Schultz, 2017). Applying this logic to the present study, we argue
by treating the various modal elements as influencers that facilitate/ that post length can influence brand engagement. In B2B social media
hinder the central route or the peripheral route information processing. marketing, information sharing has been found to play an important role
in shaping brand engagement (Leek et al., 2019; Swani et al., 2017).
3. Conceptual framework and hypotheses development Given the nature of business marketing, B2B stakeholders are usually
actively seeking information about brands/products on social media
The conceptual framework that guides our hypotheses development (Cortez et al., 2020; Sundström et al., 2021). As such, brand posts that
is presented in Fig. 1. It depicts the effects of linguistic features, grouped contain more information are more likely to fulfil business users’
into two categories based on whether they influence central or periph informational needs and if viewed as relevant, lead to enhanced
eral route processing on brand engagement. We argue that the linguistic engagement (Leek et al., 2019). Therefore, we hypothesise:
features that facilitate the central or peripheral route processing will
H1. Post length has a positive effect on brand engagement.
have positive effects, while those that hinder the processing will have
negative impacts on brand engagement.
3.1.2. Language and visual complexity
Message complexity refers to the efforts needed to process and un
3.1. Central route processing influencer derstand a message. At the linguistic level, complexity focuses more
specifically on the “ease with which people are able to convert elements
3.1.1. Post length of the message into meaning” (McShane et al., 2019, p.7). Marketing
Post length has been found to influence brand engagement on social research reveals that message complexity can significantly influence the
media (Antoniadis, Paltsoglou, & Patoulidis, 2019; de Vries, Gensler, & persuasiveness of advertising (Cox & Cox, 1988; Davis, Horváth, Gretry,
Leeflang, 2012). In B2C literature, post length has either a positive effect & Belei, 2019). Furthermore, this research indicates that marketing
(Antoniadis et al., 2019; Sabate, Berbegal-Mirabent, Cañabate, & Leb message persuasiveness can be maximized when message receivers can
herz, 2014), a negative effect (Lee, Hosanagar, & Nair, 2018; Schultz, process and understand the message with ease, revealing that, when
2017), or no significant effect (de Vries et al., 2012) on brand engage designing marketing messages, brands need to consider the match
ment. Despite the inconsistent findings, the well-developed underlying
Fig. 1. Brand Engagement in B2B Social Media: A Conceptual Framework.
4
between the cognitive resource that is available to message receivers and makes it more difficult for readers to interpret the message (Alter &
and the resource required to process the message (Anand & Sternthal, Oppenheimer, 2009; Davis et al., 2019; McShane et al., 2019). For
1990; Burgers, Konijn, Steen, & Iepsma, 2015). On social media, users instance, a tweet with these features (e.g., For #ValentinesDay, tell
constantly face ubiquitous information overload (Rodriguez, Gummadi, #IBM your #secretlovelanguage, but make it tech. https://ibm.
& Schoelkopf, 2014). Using Twitter as an example, 500 million tweets co/1KiLeh7) will be more difficult for readers to process than a clean
are posted per day on average (Brandwatch, 2020), while users spend 6 one (e.g., For Valentine’s Day, tell IBM your secret love language, but
min per day on Twitter (eMarketer, 2021) with an average session make it tech.). Additionally, given the interactive nature of these fea
duration of 3.53 min (Statista, 2019). On Facebook, approximately 1500 tures, when users click on these features, they will navigate away from
eligible posts appear in a user’s feed each day (Backstrom, 2013), while, the current message, which further decreases readers’ likelihood of
on average, users only spend 34 min per day using the platform engaging with the message (de Vries et al., 2012; Schultz, 2017).
(Hootsuite, 2021). This tendency of exposure to vast amounts of infor Together, we argue that these social media-specific features will nega
mation within such short durations demands social media users to tively influence brand engagement. Thus, we hypothesise:
optimise their cognitive resources (McShane et al., 2019). Consequently,
H3a. Hashtag has a negative effect on brand engagement.
social media users are likely to engage with brand posts that are easy-to-
process (Davis et al., 2019). Research has also shown that cognitively H3b. Cashtag has a negative effect on brand engagement.
busy individuals tend to perceive easy-to-process messages as more ac
H3c. URL has a negative effect on brand engagement.
curate and trustworthy (Alter & Oppenheimer, 2009). Both of these
message features were preferred by B2B marketing stakeholders and
positively related to the share of information (Chiu, Hsu, & Wang, 2006; 3.2. Peripheral route processing influencer
Lau & Chin, 2003; Tsai & Men, 2013). Therefore, we argue that brand
posts that are easy to process will generate more brand engagement. 3.2.1. Emotional cues
At the linguistic level, many linguistic features have been found to be Emotional cues in brand communication affect how the communi
associated with the processing ease of a message. Sentence length (e.g., cation is processed and influence the communication outcomes (Percy,
words per sentence) and word length (e.g., characters per word) have 2012). On social media, emotionally charged brand messages were
been identified and widely used as common indicators of message found to stimulate higher involvement and psychological arousal and
complexity (Davis et al., 2019). Research has shown that individuals enhance message receivers’ participating and sharing behaviours
need to consume more working memory load to process long sentences (Berger & Milkman, 2012; Kim & Johnson, 2016). In B2B marketing,
and long words (Mikk, 2008). As such, incorporating long sentences and emotional communications provide peripheral cues that satisfy various
words in brand posts will decrease the processing ease and require more stakeholders’ divergent socio-psychological motives to build emotional
cognitive resources to be employed (Khawaja, Chen, & Marcus, 2014). associations and achieve emotional brand values, thus, potentially
Except for long sentence and long words, style words such as preposi activating engagement intent and WOM behaviours (Cortez et al.,
tions and conjunctions are also indicative of more complex language 2020). Previous research found that emotional appeals increase B2B
(Tausczik & Pennebaker, 2010). Prepositions signal that more complex posts’ popularity on social media (Leek et al., 2019; Swani et al., 2017).
and concrete information about a topic is provided, and conjunctions Furthermore, research has shown that the more emotional a tweet is, the
integrate multiple complex ideas together and often introduce more more frequently and faster the tweet will be shared (Stieglitz & Dang-
complex sentence structures (Tausczik & Pennebaker, 2010). Research Xuan, 2013; Thelwall, Buckley, & Paltoglou, 2011). Thus, we argue
has also shown that prepositions and conjunctions reflect the cognitive that post emotionality (the extent to which a brand post is emotional)
complexity of a message (Buck, Minor, & Lysaker, 2015; Tausczik & will positively affect brand engagement.
Pennebaker, 2010). Thus, using prepositions and conjunctions will in
H4a. Emotionality has a positive effect on brand engagement.
crease language complexity and necessitate that readers expend more
effort to understand and interpret brand posts. Taken together, we B2B marketing literature has acknowledged the importance of
hypothesise: adopting certain linguistic cues in communications to stress the psy
chological and emotional resonance between the buyer and the seller
H2a. Average sentence length has a negative effect on brand
and to create strong emotional connections with the customer (Lynch &
engagement.
De Chernatony, 2007). As such, when communicating with stakeholders
H2b. Long words have a negative effect on brand engagement. on social media, B2B brands should tweak their language to stress
appropriate emotional brand values and socio-psychological values that
H2c. Prepositions have a negative effect on brand engagement.
match stakeholders’ needs (Rich & Smith, 2000). One of the most
H2d. Conjunctions have a negative effect on brand engagement. important brand values in B2B relationships is trust (Morgan & Hunt,
1994), and recent research confirms that social media communication
Except for traditional linguistic features, social media-specific fea
practices directly impact B2B customers’ perceptions of trust (Rose,
tures, such as hashtag (a metadata tag that is prefaced by the hash
Fandel, Saraeva, & Dibley, 2021). Within the B2B context, while func
symbol, e.g., #TimeToAct), cashtag (a company ticker symbol that is
tional/rational brand values transferred and accumulated in actual
prefaced by the U.S. dollar sign, e.g., $TSCO), and URL (a reference to a
purchase situations may still dominate the development of trust and
web resource, e.g., ibm.co/3jYd9IH), can also influence message pro
commitment, research suggests that stakeholders on social media can be
cessing ease. Social media platforms have created these specific features
influenced by emotional brand values such as reassurance and security
to help users deal with the vast amount of information. For example,
(Jussila, Kärkkäinen, & Aramo-Immonen, 2014) as the brand serves as a
hashtag and cashtag can help users search for posts under similar topics
risk-reduction heuristic for B2B customers (Brown, Zablah, Bellenger, &
or related to the same company, track discussions, and distribute their
Johnston, 2011). Research on general semantics and linguistics suggests
messages in a more targeted way. URLs can provide the link to external
that word use indicating certainty tends to endorse the brand and
resources that may provide users with more information. In being
demonstrates resoluteness and confidence (Pitt et al., 2019), which leads
created to facilitate users searching, sharing, and engaging with infor
to the message receiver’s perceived expertise and credibility of the
mation, these features have become central to social media communi
brand (Money, 2004) and a feeling of self-assurance (Zhang et al., 2020).
cations (Davis et al., 2019; McShane et al., 2019). However, despite their
Similarly, achievement words reinforce brands’ capabilities and
convenience, incorporating these features in social media messages can
competence within specific areas and signals to message receivers the
increase the visual complexity, which disrupts the processing fluency
benefits of engaging with them (Leek et al., 2019). Moreover,
5
achievement words help fulfil brand followers’ needs for self- H5b. Use of at mentions has a positive effect on brand engagement.
actualisation and significantly relates to emotive resonance and
H5c. First plural pronouns have a positive effect on brand
deeper and more enduring customer relations (Ringsberg & Forquer
engagement.
Gupta, 2003). Hence, we hypothesise:
H5d. Second pronouns have a positive effect on brand engagement.
H4b. Certainty words have a positive effect on brand engagement.
H4c. Achievement words have a positive effect on brand engagement. 3.2.3. Multimodal cues (Rich Media)
Advertising has a long history of using vivid messages to influence
3.2.2. Interpersonal cues stakeholders’ attitudes toward brands. On social media, brands often
In B2B marketing, building interpersonal relationships has been create vivid posts by incorporating modal cues in rich media, such as
viewed as a central facet of brand communications (Leek et al., 2019) photo, gif, and video, with the aim of enhancing the persuasiveness of
and managing these relationships using social media is increasingly seen their posts. Vividness refers to the extent to which a brand post stimu
as valuable (Cartwright, Davies, & Archer-Brown, 2021). The traditional lates different senses and can be achieved by including dynamic ani
approach to B2B marketing has highlighted the roles of personal re mations, colours, or pictures (de Vries et al., 2012; Steuer, 1992).
lationships and interactions between sales representatives and cus According to Steuer (1992), vividness is based on its sensory breadth, i.
tomers (Huotari et al., 2015). This need to develop interpersonal e., the number of senses engaged. As such, text would be low in vivid
relationships is enhanced by the very nature of social media, as one of ness, whereas photo and video would be high in vividness. Unlike simple
the ideological and technological foundations of social media is text-only posts, brand posts presented in a multimodal format capture
networking (Kaplan & Haenlein, 2010). Because of its interactive na and stimulate multi-sensory interaction of sight, sound, and motion,
ture, social media has been viewed as a communication channel where thus, enable readers to have a more immersive experience with the
B2B brands specifically focus on developing, enhancing, and sustaining content. Previous research suggests that a vividly presented message is
their interpersonal relationships with stakeholders (Cortez et al., 2020). inherently interesting, attention-getting, thought-provoking, image
Linguistic research has a long history of studying language’s role in producing, emotionally arousing and easy upon which to elaborate
shaping relationships, and the findings from this research suggest that (Kim, Kardes, & Herr, 1991; Sreejesh, Paul, Strong, & Pius, 2020).
language can be used to improve relationships (Otterbacher et al., Moreover, vividly presented messages are suggested to reinforce the
2017). Some language features, when adopted properly, can soften hi intimacy and authenticity of the communication (Mehmet & Clarke,
erarchical power relationships, reduce social distance, and convey 2016). Therefore, we argue that brand posts using modal cues in rich
closeness (Gretry, Horváth, Belei, & van Riel, 2017), thus facilitating an media can increase the post vividness and are, thus, preferred by brand
interpersonal relationship. We argue that adopting these linguistic fea followers on social media. Therefore, we hypothesise:
tures can help B2B brands develop interpersonal relationships with their
H6a. Use of gifs has a positive effect on brand engagement.
stakeholders on social media.
Linguistic features, such as emojis, at mention, and personal pro H6b. Use of photos has a positive effect on brand engagement.
nouns, can be used to facilitate interpersonal communications. Emojis,
H6c. Use of video has a positive effect on brand engagement.
when used in brand communications, can create a light mood by making
the communications less serious and more friendly (Kaye, Wall, &
4. Methodology
Malone, 2016), conveying an attitude of welcoming conversational
communication and informal interaction (Crystal, 2006). Furthermore,
4.1. Data collection
through conveying non-verbal cues that are normally absent in
computer-mediated communications, emojis can help decrease the
Our data is drawn from Twitter. With 192 million monetisable daily
psychological distance and consequently increase the perceived in
active users (Twitter, 2021a), Twitter has become one of the most
timacy (Derks, Bos, & Von Grumbkow, 2007). At mention makes a brand
important social media platforms for B2B marketing (Leek et al., 2019;
message more targeted to specific individuals or groups and transforms
McShane et al., 2019; Pitt et al., 2019). According to the 2020 B2B
the message for the masses into direct communication. This direct
Content Marketing Report, 82% of B2B content marketers used Twitter
communication assembles the traditional interpersonal communication
for content marketing in the last 12 months (Content Marketing Insti
(face-to-face or remotely via phone, email, or videoconference) that was
tute, 2021), making it the second most popular social media platform for
the mainstay of interactive information exchange in business markets
B2B marketing. With its public nature, Twitter also provides a natural
(Leek et al., 2019), thus is likely to be preferred by B2B stakeholders.
setting where brand engagement can be observed unobtrusively.
Second person pronouns direct attention inward toward the person
To create a dataset, we used the top 200 ranked B2B brands in the
reading the brand messages. As such, using second-person pronouns
B2B Social Media Report as our sample (Brandwatch, 2015). These
increases self-referencing and perceived relevancy of the messages and
brands were identified and ranked based on the B2B social ranking index
conveys a feeling of personalisation. Research has shown that person
that evaluates their social media performance and have been used in
alised brand message can increase customer involvement (Cruz, Leon
other research as a pool of B2B brands (McShane et al., 2019; Pitt et al.,
hardt, & Pezzuti, 2017). Thus, using second-person pronouns is more
2019). Facepager, a free, open-source software developed for fetching
likely to fulfil the interpersonal need of B2B stakeholders. First-person
publicly available data on various social media platforms such as
plural pronouns (e.g., we, us) can signal a sense of group identity
Facebook, Twitter, Reddit, and YouTube, was used to capture the most
(Tausczik & Pennebaker, 2010). When used in advertisements, first-
recent tweets from these brands (Jünger & Keyling, 2019). Employing
person plural pronouns were found to enhance brand attitudes by
Twitter application program interface (API), Facepager allows us to
creating a sense of closeness with the brand (Cruz et al., 2017). Despite
collect the full tweets, the number of likes and retweets, date and time of
the different mechanisms, a common argument that can be drawn from
the tweets, media types and other specific elements incorporated in the
these findings is that these linguistic features all facilitate building an
tweets, such as hashtag, cashtag, user mentions, and URLs (Jünger &
interpersonal relationship. Taken together, we argue that using these
Keyling, 2019). Data collection was conducted in late January 2021. For
linguistic features in brand posts can facilitate relationship building with
each brand, we scraped the most recent 3200 tweets as this represents
brand followers, thus enhancing their engagement. Thus, we
the maximum number of historical tweets that can be accessed via the
hypothesise:
public Twitter API. When scraping the data, we excluded brand replies
H5a. Use of emojis has a positive effect on brand engagement. (i.e., brand responses to other Twitter users) because these tweets are
6
invisible to brand followers unless they also follow the other Twitter and have been widely used as the measurements of brand engagement in
users mentioned. For brands with multiple official Twitter accounts, we previous research (Davis et al., 2019; Leek et al., 2019; McShane et al.,
scraped tweets from the account with the highest number of followers. 2019). On Twitter, “like” represents the lowest engagement behaviour
We excluded tweets posted after December 31, 2020, to filter out the as it requires only a single click on the liking button, which is virtually
potential changes in brand engagement after data being recorded. We effortless, instantaneous, and reflexive (Labrecque, Swani, & Stephen,
then removed retweets as they are not originally posted by brands and 2020; Leek et al., 2019). The like action mainly symbolises support for
tweets that are not in English. Since some brands do not have Twitter the brand or indicates that the brand tweet has been seen and positively
accounts or have closed their Twitter accounts because of being merged received (Leek et al., 2019; Swani & Labrecque, 2020). Retweet allows
or acquired, our final sample included 229,727 tweets from 156 brands users to spread brand tweets publicly to their own networks. Before
that represent 10 industries, namely, aerospace, agriculture & food retweeting, users have the option to add their own comments and/or
production, business software, chemical, construction, energy, heavy media; and to do so, users usually need to read and process the brand
industry, industrial technology, medical & pharmaceutical, and military tweets well. As such, retweet is reflective and requires more cognitive
defense (see Appendix A for sample description). resources (Labrecque et al., 2020; Leek et al., 2019; Swani & Labrecque,
2020). Although retweeting can be performed through clicking a few
4.2. Measurement share buttons, the retweeted brand post will appear in a user’s followers’
timelines and can be seen by all of the user’s personal relationships,
4.2.1. Dependent variables allowing a higher visibility of one’s activities to others (Labrecque et al.,
In Table 1, we report the measurements and descriptive statistics of 2020; Leek et al., 2019; Swani & Labrecque, 2020). As such, retweeting a
our variables. To measure brand engagement, we used two common brand post shows a form of brand identification and strong brand
social media metrics, namely, likes and retweets. On Twitter, liking and endorsement and represents the highest engagement behaviour (Lab
retweeting brand posts are the most common engagement behaviours recque et al., 2020; Leek et al., 2019). In this paper, we followed pre
that brand followers can perform. As such, likes and retweets have vious research (Davis et al., 2019; Leek et al., 2019; McShane et al.,
become the prevalent metrics and objectives in social media marketing 2019) and measured brand engagement using like and retweet counts
the tweets received to address both lowest and highest brand engage
Table 1 ment behaviours.
Variables, measurements, and descriptive statistics.
4.2.2. Independent variables
Variable Description Mean SD Min Max
To measure the linguistic variables for our large dataset, we used the
Dependent 2015 Linguistic Inquiry and Word Count (LIWC). LIWC is computational
Variables
linguistic software that has been widely adopted in social research
Like Number of likes 22.76 372.70 0 134,717
Retweet Number of retweets 6.34 91.16 0 36,525 (Pennebaker, Boyd, Jordan, & Blackburn, 2015). It can be used to
Independent measure the linguistic style of text by assessing approximately 90 pre
Variables defined linguistic categories, including general descriptors (total word
Post length Total number of 23.10 9.87 1 58
count, words per sentence, etc.), style words (e.g., pronouns, articles,
words
Average Average number of 12.37 6.01 1 53
auxiliary verbs, preposition, conjunction, etc.), words tapping psycho
sentence length words per sentence logical constructs (e.g., affect, cognition, biological processes, drives,
Long word Percent of words 36.07 12.90 0 100 etc.), personal concern categories (e.g., work, home, leisure activities,
with 6 or more letters etc.), informal language markers (e.g., fillers, assent, swear words, etc.),
Preposition Percent of 13.52 6.22 0 60
and punctuation categories (e.g., periods, exclamation, ellipsis, etc.)
prepositions
Conjunction Percent of 3.16 3.72 0 50 (Pennebaker et al., 2015).
conjunctions In the present study, we used LIWC to measure the linguistic vari
Hashtag Number of hashtags 1.40 1.36 0 13 ables, namely, post length (i.e., total word count), average sentence length
Cashtag Number of cashtags 0.02 0.19 0 4
(i.e., word count per sentence), long word (i.e., word with six or more
URL Number of URLs 0.68 0.49 0 4
Emotionality Sentiment strength of 0.88 0.98 0 7
letters), preposition, conjunction, certainty, achievement, first-person plural
the tweet pronoun, and second-person pronoun. These variables were expressed as
Certainty Percent of certainty 0.81 2.12 0 100 the percentage of total words except for post length and average sentence
words length.
Achievement Percent of 3.04 4.20 0 100
For emotionality, we measured the level of sentiment using SentiS
achievement words
Emoji Tweet contains emoji 0.07 0.25 0 1 trength (Thelwall, Buckley, & Paltoglou, 2012). SentiStrength is a
At mention Number of at 0.47 0.82 0 17 sentiment analysis tool that can be used to classify emotions in short
mentions informal messages like tweets. It has been proven to provide a higher
First-person Percent of first plural 2.81 3.84 0 50 accuracy rate than standard machine learning approaches (Thelwall
plural pronoun pronouns
Second-person Percent of second 1.16 2.79 0 50
et al., 2012). For each text analysed, SentiStrength can report a positive
pronoun pronouns sentiment score on a scale of 1 (neutral) to 5 (strongly positive) and a
Gif Tweet includes a gif 0.02 0.13 0 1 negative sentiment score on a scale of − 1 (neutral) to − 5 (strongly
Photo Tweet includes photo 0.55 0.50 0 1 negative). To measure the emotionality (the total amount of sentiment)
Video Tweet includes video 0.08 0.28 0 1
of tweets, we followed Stieglitz and Dang-Xuan (2013) and used the
Control Variables
Weekend Tweet was posted on 0.09 0.29 0 1 following calculated score:
weekend
Posting Year Dummy variables for – – – –
emotionality = (positive sentiment score–negative sentiment score) − 2
(11) Year the tweet was
This score, which ranges from 0 to 8, captures the degree of
posted
Posting Month Dummy variables for – – – –
emotionality because both positive and negative sentiments are
(11) Month the tweet was included. In the formula, 2 was subtracted to avoid confusion where a
posted positive number, i.e., 2, indicates no sentiment in the case of the score
Brand (155) Dummy variables for – – – – ranging from 2 to 10.
brands
Other internet linguistic symbols (hashtag, cashtag, url, and at
7
mention) were measured using the count of each symbol incorporated in Table 2
tweets. Regarding the visual elements (emoji, gif, photo, video), we Standardised estimation results for brand engagement.
created separate dummy variables to indicate whether each of these Ln Like Ln Retweet
elements was contained in one tweet.
Post Length Total word count 0.028*** 0.037***
Language Complexity Average sentence − 0.004** − 0.004*
4.2.3. Control variables length
We included a series of control variables to account for their po Long word − 0.006*** − 0.006***
tential influence. Research has shown that social media users’ activity Preposition 0.001 − 0.006***
Conjunction − 0.006*** − 0.005**
level is different between weekdays and weekend (Wagner, Baccarella, Visual Complexity Hashtag − 0.015*** 0.000
& Voigt, 2017), which implies that people may engage with brands at Cashtag − 0.022*** − 0.021***
different levels on the weekend versus weekdays. Recent research URL − 0.068*** − 0.018***
revealed that brand tweets posted during weekends generate higher Emotional Cues Emotionality 0.019*** 0.010***
Certainty 0.016*** 0.016***
engagement than those posted during weekdays (McShane et al., 2019).
Achievement 0.016*** 0.008***
Thus, we included a dummy variable that indicates whether a tweet was Interpersonal Cues Emoji 0.019*** 0.015***
posted on the weekend to control for this effect. In the present study, we At mention 0.038*** 0.035***
collected tweets from 156 different brands. Previous research has First plural pronoun 0.031*** 0.014***
identified brand as an important factor that influences brand engage Second pronoun − 0.025*** − 0.016***
Multimodal Cues (Rich Gif 0.022*** 0.027***
ment. Brand characteristics such as brand reputation and brand equity Media) Photo 0.077*** 0.057***
can strongly influence brand engagement on social media (Van Doorn Video 0.097*** 0.099***
et al., 2010). To rule out the potential influences, we created and Control Variables Weekday vs. Weekend 0.019*** 0.012***
included dummy controls for brands by which the tweets were posted. Posting Year (11) X X
Posting Month (11) X X
This also helps rule out the potential influences of the unique brand
Brand (155) X X
audience profiles and inherently controls for the effects of industry and Unstandardized Constant 3.29 2.10
brand follower size. Lastly, we included dummy controls for the years N 229,727 229,727
and months when the tweets were posted to minimise other potential F-Value 2493.01*** 1228.82***
influences raised by posting time and historical Twitter updates. Since R2 0.68 0.51
Adjusted R2 0.68 0.51
Twitter does not allow access to brands’ historical follower size data,
*
these dummy variables of posting time (year and month), together with p < 0.05.
**
brand dummy variables, help minimise the potential influences of his p < 0.01.
***
torical brand follower size changes. p < 0.001.
4.3. Analysis method length) and less long words (i.e., words with six or more characters) are
liked and retweeted more frequently, supporting H2a and H2b.
To test our hypotheses, we ran multiple linear regressions to estimate Regarding the effects of prepositions and conjunctions, we found them
the effects of various post characteristics on brand engagement. The to both have negative effects on brand engagement, except for prepo
dependent variables are the numbers of likes and retweets. Given that sition on the number of likes. Thus, H2c is only partially supported and
the dependent variables are positively skewed, we followed previous H2d is supported. Despite of being complex, prepositions can increase
research (McShane et al., 2019; Schultz, 2017) and used the natural the concreteness of the posts by providing more contextualised and
logarithmic transformation of like and retweet counts, i.e., Ln(Like+1), detailed information (Larrimore, Jiang, Larrimore, Markowitz, & Gor
Ln(Retweet+1), as our dependent variables. Here, 1 was added to avoid ski, 2011). Compared to retweeting, liking behaviour is less cognitively
taking logs of zero. The statistical analysis was performed using IBM loaded and mainly reflects the brand endorsement expressed by social
SPSS Statistics Version 26. SPSS is a statistical software that has been media followers, thus is less utilised in the central route processing
widely used in social science research, thus provides a solid tool for (Sabate et al., 2014). Consequently, the negative effects of a specific
current data analysis. In the next section, we report the statistical linguistic feature on liking behaviour via central route processing are
analysis results. more easily negated by its positive effects. As such, while the use of
prepositions negatively influences brand engagement via increased
5. Results language complexity, this effect is likely countered by the increased post
concreteness at the same time. Similar to linguistic features that increase
The standardised estimation results are presented in Table 2. As a language complexity, the social media-specific features that increase
whole, the model for like is significant (F = 2493.01, p < 0.001) and visual complexity also have negative effects on brand engagement. More
explains the variance of the dependent variable well (R2 = 68%, Adj. R2 specifically, the more cashtags and URLs a tweet contained, the less it
= 68%). Similarly, the model for retweet is significant (F = 1228.82, p < was liked and retweeted. Thus, H3b and H3c are supported. With
0.001) and explains the variance of the dependent variable well (R2 = regards to hashtags, we found it negatively influences the number of
51%, Adj. R2 = 51%). Compared to previous research (i.e., McShane likes but not retweets. Thus, H3a is partially supported. One possible
et al., 2019), the current model provides more explanatory power for explanation is that hashtag, as a linguistic feature in social media
both like and retweet. One potential explanation is that, while McShane communications, may affect engagement behaviours differently and
et al. (2019) focused on the central route processing and mainly through different mechanisms because of the new psychological mean
examined brand engagement from the perspective of information pro ings socially constructed through its use on social media (Halliday,
cessing fluency, the current model considered the central route and 1978; Crystal, 2011). For example, by sharing a hashtag (e.g., “#cli
peripheral route processing simultaneously. matecrisis” and “#InventorsDay”), the viewer acknowledges the ideo
Regarding the effects of central route influencers, we found that logical values expressed by the hashtag and is willing to be self-
linguistic features that facilitate the central route processing positively identified and socially presented as a member of the community.
influence brand engagement, while those that hinder the central route Hence, while hashtag may increase the visual complexity to discourage
processing negatively affect brand engagement. As we expected, tweets engagement behaviour (liking), it may make the shared values salient
with more words tend to receive more likes and retweets, supporting H1. and initiate emotional contacts between the brand and the viewer to
Furthermore, tweets with shorter sentences (i.e., lower average sentence encourage sharing (Sundström et al., 2021).
8
In terms of the effects of peripheral route influencers, we found that representation needs and further facilitate a mutual-trust relationship.
brand tweet features that facilitate peripheral route processing generally Therefore, enhancing brand engagement behaviours such as liking and
enhance brand engagement. As we expected, tweets with more sharing on social media enables B2B brands to develop long-term re
emotional communications received more likes and retweets. Specif lationships and build stronger loyalty with the stakeholders, which in
ically, we found that emotionality, certainty words and achievement crease customer considerations when future business opportunities arise
words all have positive effects on brand engagement, supporting H4a, (Cartwright, Liu, & Raddats, 2021; Tiwary, Kumar, Sarraf, Kumar, &
H4b, and H4c. Regarding the interpersonal linguistic cues, we found that Rana, 2021).
incorporating emojis, at mentions, and first-person plural pronouns in Additionally, enhancing brand engagement will improve B2B
tweets enhance brand engagement, supporting H5a, H5b, H5c. Addi brands’ social media presence and help them expand the business
tionally, we found that the more second-person pronouns a tweet con community. On Twitter, liking and retweeting both help the spread of
tained, the less it was liked and shared. Thus, H5d is not supported. brand posts. When a user retweets a brand tweet, the brand tweet will
Despite second-person pronouns promoting self-referencing and appear in all the user’s followers’ timelines with a retweet icon and the
creating a conversational atmosphere, research has shown that the use user’s handle (Twitter, 2021b). When a user likes a brand tweet, the
of second-person pronouns predicts low-quality relationships (Tausczik brand tweet liked also has a possibility of appearing in the user’s fol
& Pennebaker, 2010). Second-person pronouns can reflect the position lowers’ timeline (Twitter, 2021b). As such, increasing liking and sharing
in a social hierarchy; those higher in the social hierarchy use second- behaviours enables B2B brands to disseminate information to stake
person pronouns at a greater frequency (Kacewicz, Pennebaker, Davis, holders who are otherwise outside of the brands’ own network, largely
Jeon, & Graesser, 2014). Thus, using second-person pronouns in brand expanding the brand audience network (Leek et al., 2019; Swani et al.,
tweets may communicate a sense of unequal social position, which 2013). According to a recent report, Twitter users have 707 followers on
might not be preferred by B2B brand audiences. As a result, the use of average (Brandwatch, 2020). This means that a single retweet could
second-person pronouns can discourage brand engagement. Further spread a brand post to over 700 additional users, not accounting for the
more, we found that tweets that contained rich-media modalities (gif, potential multiplier effect that is likely to occur if any of these 700 users
photo and video) were liked and retweeted more frequently, supporting retweet again and shared the brand post with their networks. Today, the
H6a, H6b and H6c. customer experience in B2B firms is fundamentally different. Customers
seek information about vendors when they perceive they need it, and
6. Discussion very often, the customers are embedded in various social networks and
both gain information from these networks and are influenced by them
In this paper, we found that linguistic features of brand posts impact (Articulate, 2019; Cortez et al., 2020). Therefore, through enhancing
brand engagement on social media via either the central route or the brand engagement on social media, B2B brands can reach more audi
peripheral route of information processing. Thus far, research on brand ences, thus create brand awareness, and may generate referrals later on
social media posts has mostly focused on B2C contexts (Kaplan & (Luo, Toth, Liu, & Yuan, 2021).
Haenlein, 2010), with limited research comparing B2C and B2B brand
posts as a focus (Swani et al., 2017). Very little research has addressed 6.1. Theoretical implications
B2B brand posts specifically, focusing on the performance and strategic
management issues (Liu, 2020; Zhang et al., 2020) or concentrating on The findings of this research have a few theoretical implications.
‘what messages to post’ – the informational content of the brand post Firstly, the present study integrates linguistic theories with a B2B
(Juntunen et al., 2020). Notwithstanding these few studies, the question communication model (dual routes for information processing) and
of ‘how to design the post’ has not yet been separately examined in empirically tests the influence of linguistic features on brand engage
detail. Therefore, our research contributes to B2B social media content ment. Our research underscores the importance of treating brand social
marketing and brand engagement literature by unveiling how B2B media posts as exhibiting linguistic multimodality (Halliday, 1976;
brands can enhance brand engagement on social media through Jewitt, 2009; Mehmet & Clarke, 2016) and suggests that the various
adopting the proper linguistic features. modal elements in brand posts impact the audiences’ information pro
B2B brands can benefit from the enhanced brand engagement in cessing routes and subsequently, their liking and sharing behaviours.
several ways. Through enhancing brand engagement, B2B brands can Furthermore, our analysis bridges linguistic theories with B2B commu
reinforce their relationships with customers and other stakeholders, nication research and suggests their mutual affinity. Being multimodal,
which has been identified as B2B companies’ primary goal of adopting linguistic features of the brand social media post help audiences inter
social media (Agnihotri, 2020; Cartwright, Davies, & Archer-Brown, pret the informational/functional message through the central pro
2021). On social media, liking a brand post represents a positive atti cessing route and fulfil the interpersonal and emotional meanings
tude and acknowledgement toward the content and conveys one’s through the peripheral route (Cortez et al., 2020; Halliday, 1976, 1978).
relationship with the brand itself (Labrecque et al., 2020; Swani & Hence, it is undoubtedly essential to examine the various modalities in
Labrecque, 2020). It has been found to be primarily driven by cus the brand social media post and build knowledge of how these linguistic
tomers’ motive to maintain the brand-relationship connection (Dhaoui components influence the central/peripheral information processing
& Webster, 2021; Labrecque et al., 2020; Swani & Labrecque, 2020). As routes in predicting brand engagement.
such, through posting contents that stimulate liking behaviour, B2B Our results support that appropriate design and use of linguistic
brands can better address stakeholders’ needs to connect with brand and features facilitate the central route processing to increase brand
facilitate the process of establishing sustained long-term relationships engagement. More specifically, brand posts presented in greater length
with customers and stakeholders (Agnihotri, 2020). Different from (indicating higher informativeness) and with lower language/visual
liking, sharing a brand post allows users to take partial or full ownership complexity have a higher level of engagement. There is no doubt that
of the contents and spread the contents to their own networks. Research stakeholders are likely to hold a utilitarian motive for professional/
has found that sharing behaviour is predominantly driven by self- functional information and rational appeals. By filling a brand post with
representation (Labrecque et al., 2020; Swani & Labrecque, 2020). high informativeness and, at the same time, presenting it in a simple and
B2B research has indicated that emotional bonds between B2B brands easy-to-understand mode, brands could enhance the central route pro
and customers are likely to be developed if brand communications cessing to increase stakeholders’ perceived usefulness of the post and
convey the shared values and reflect customers’ extended self perceived ease of use, which in turn, leads to a positive attitude toward
(Sundström et al., 2021). Therefore, through posting contents that the brand, and then engagement and WOM behaviours.
stimulate sharing behaviour, B2B brands can address stakeholders’ self- Our research also supports that linguistic features can facilitate
9
peripheral processing to build an emotional and personal connection market and people who may be interested in brand/product/service
between the viewer and the brand to increase brand engagement (Cortez information. Hence, staying active in creating and updating social media
et al., 2020; Swani et al., 2014). Specifically, adding emotional cues, posts is critical.
interpersonal cues, and multimodal cues in rich media increases brand Our research offers guidance for B2B content marketers regarding
liking and sharing. The implication for B2B marketing is that it em their social media content creation. Specific suggestions are provided
phasises that B2B brand posts are likely to be read by various stake regarding the language use of brand posts to increase brand engagement
holders situated in various settings and holding dynamic motives such as and WOM behaviours. First, we suggest B2B marketers write longer but
socialising, experiencing, and self-expression (Belk, 2013; Kaplan & less-complex posts to maximise the informativeness of the post while
Haenlein, 2010). Such interpersonal situations (e.g., friends) are more simultaneously decreasing the required cognitive load for processing (e.
likely to activate peripheral route information processing, and viewers g., write shorter sentences, use simple words, use fewer preposition and
tend to privilege subjective cues to generate emotional responses and conjunction words). In addition, the social media-specific features
association with the brand (Cortez et al., 2020). Hence, our results (hashtag, cashtag, URL) increase the visual complexity of the text and
support literature and theory recognising the power of emotional ap the viewer’s cognitive processing load and lead to decreased engage
peals in increasing brand equity (Lynch & De Chernatony, 2004) and ment and WOM (liking and sharing). However, the use of social media
contributes to the literature by revealing the role of linguistic features in features (e.g., hashtag, at-mention, emoji) may also relate to emotional
delivering emotional appeals. Previous research focuses on informa appeals and responses due to the socio-psychological meanings
tional content (themes/topics), which polarises the advertising appeal to embedded in specific social media behaviours (e.g., sharing). For
be either rational or emotional. Our findings further illustrate that B2B example, the share of a hashtag (e.g., “#climatecrisis” and “#Inven
firms could actually deliver emotional appeals through linguistic fea torsDay”) may indicate that the viewer acknowledges the ideological
tures regardless of the factual informational content being either values embedded in the hashtag and is willing to be self-identified and
rational or emotional because the message receiver is influenced by the socially presented as a member of the community. Hence, B2B content
emotion, mood, and attitudes expressed by the message no matter its marketers may need to develop a comprehensive knowledge of the social
subject matter (Mehmet & Clarke, 2016). It further suggests that B2B media language and features and then make trade-offs regarding their
marketing tasks that fit social media content are expandable. Tasks like impact on the audience’s cognitive processing and emotional bonding.
direct selling and promotion that are suggested to be avoided in social We provide particular guidance regarding the emotional bonding
media might become promising by utilising a viewer preferred post between B2B brands and the audience. We suggest that B2B marketers
design. For example, images and emojis might be effective in conveying could include more emotional words, add rich media (e.g., gif, image,
commercial messages because they create authenticity and are not and video), and use emojis, at-mention, and first plural pronouns (e.g.,
treated as advertisements (Mehmet & Clarke, 2016). ‘we’ and ‘us’) in social media posts, to provide more emotional and
Research suggests that linguistic differences can be used to assess interpersonal cues for the audience situated in low business task
individual differences in personality, translating their internal thoughts involvement and high personal/informal settings (Cortez et al., 2020).
and psychology (Pennebaker & King, 1999; Tausczik & Pennebaker, Swani et al. (2017) have made suggestions for B2B content creation,
2010). Even in the B2B marketing context, brands have the need and such as incorporating corporate brand names. Our research provides
potential for differentiation by establishing a unique and consistent further insights by suggesting that, while B2B brands include corporate
identity (Michaelidou, Siamagka, & Christodoulides, 2011). The first brand names in their posts, they may not rule out the use of “we” and
task, then, is to determine which language style and verbal tone are “us” because these pronouns indicate an authentic friend-like conver
preferred by social media users. In this aspect, our findings have im sational tone that builds an interpersonal and informal conversation
plications by suggesting an authentic friend-like voice (by adding environment to fit social media users’ socio-psychological needs.
interpersonal indicators and emotionality) in B2B social media com Moreover, the informal and interpersonal interaction can be enhanced
munications (Huotari et al., 2015). Further, our research suggests that a by including emojis and at-mention in B2B brand posts because these
friend speaking with certainty and achievement is more likely to be liked social media features help decrease the social distance between the
and engaged with by the viewers. Because language that shows certainty brand and the audience and provide insights into the emotional posi
endorses the speaker (Pitt et al., 2019) and serves as a risk-reduction tions held by the brand (Mehmet & Clarke, 2016). Importantly, while
heuristic for B2B stakeholders, emotional brand values like trust and constructing an authentic friend-like communication environment, the
security could be built (Lynch & De Chernatony, 2004; Money, 2004), brand is suggested to speak as a friend that is self-assured and
which is the key success factor in B2B marketing (Morgan & Hunt, achievement-focused to fit the ingrained need for trust and security in
1994). It reminds us that, although B2B branding calls for emotional the B2B relationship and the shared value for self-achievement on social
bonding and interpersonal connection, it still needs to consider in media.
dividuals’ risk-related concerns from a business perspective. For
example, for B2B content that aims to build emotional ties with stake 6.3. Limitations and further research
holders by disclosing company/product weaknesses and calling for so
lutions (Sundström et al., 2021), a friend-like verbal tone is suggested. The present study investigates the effects of post linguistic features
However, a voice that speaks with certainty and shows achievement on B2B brand engagement on social media. While this paper presents
goals may help increase the communication effectiveness by diluting many novel findings, it also reveals various research opportunities for
stakeholders’ risk-related concerns and doubts about the company’s future research. The present study mainly addresses the exposure of
capability and competence (Leek et al., 2019). linguistic features in the B2B brand social media post. It would be
interesting to further examine the effectiveness of more specific post
6.2. Managerial implications designs, such as the placement of social media-specific features (e.g.,
emojis, hashtags, at mentions, URLs) in the text and the visual grammar
Our research has implications for B2B marketing practice as it mo of the images, gifs, and videos. Revealing these effects not only enables a
tivates B2B marketers to become more active in updating and replen deeper understanding of how the two information processing routes
ishing social media messages. As social media has inevitably permeated function in B2B brand communication but also provides further guid
our social life, contemporary B2B relationships no longer only operate in ance on tactics that B2B marketers can employ to increase brand
a small offline network or solely rely upon face-to-face meetings at engagement on social media. Additionally, the present study was con
conferences or trade shows (Sundström et al., 2021). Through their so ducted based on Twitter as it has become one of the most adopted social
cial media presence, B2B companies are connecting to a worldwide media platforms for B2B marketing. Recent research suggests that
10
different social media platforms are preferred by people with different expand the metrics to include other engagement behaviours, such as
personalities (Hughes, Rowe, Batey, & Lee, 2012). As a result, users’ replies, mentions, link clicks, links back, and favorites on the page. Such
activities and even language preferences can vary across different social investigation will provide guidance for B2B marketer to better manage
media platforms (Adamopoulos, Ghose, & Todri, 2018; Oz, Zheng, & brand engagement behaviours in a targeted manner. Finally, the present
Chen, 2018; Schweidel & Moe, 2014). Therefore, it would be interesting study mainly addresses the daily communication between B2B brands
to examine whether brand language influences brand engagement on and their social media followers as this has become the core activities
other social media platforms, such as LinkedIn and Facebook, differently performed by social media marketers. It would be interesting for future
from Twitter. Such investigation is undoubtedly meaningful for guiding research to examine the brand engagement behaviours during extensive
the B2B brand’s multi/cross platforms strategy. In the present study, we marketing campaigns and investigate how brand engagement behav
focus on two common brand engagement behaviours, namely, liking and iours on social media influence other conversion metrics, such as email
retweeting, because of Twitter API’s limitation on data access. While sign-up, whitepaper download, and quote request. Such investigation
this operationalization covers both the low and high engagement be will provide further guidance on B2B marketing’s omnichannel strategy
haviours on Twitter, it would be interesting for future research to design.
Appendix A. Sample description
Brand Industry Twitter Handle Followers Following Listed Count Favorites Statuses
Alcoa Aerospace @Alcoa 27,542 5176 664 3041 6660

Chemring Aerospace @Chemring_Group 1187 101 43 22 514
Cobham plc Aerospace @Cobham_plc 7577 806 0 929 2369
Collins Aerospace Aerospace @CollinsAero 42,592 1382 790 8310 9622
GKN plc Aerospace @GKN_plc 3519 159 60 148 568
Martin-Baker Aerospace @MB_EjectEject 8137 391 96 836 1108
Orbital ATK Aerospace @OrbitalATK 108,652 462 1983 2149 7429
Parker Hannifin Aerospace @ParkerHannifin 16,556 4525 481 4379 19,488
Rolls-Royce Aerospace @RollsRoyce 322,824 417 1336 428 3243
Spirit AeroSystems Aerospace @SpiritAero 9668 337 229 1632 2647
Textron Inc. Aerospace @Textron 10,228 113 215 1572 5045
The Boeing Company Aerospace @Boeing 604,323 236 4427 792 5368
United Technologies Aerospace @UTC 45,331 753 685 4015 7200
ABP Foods Agriculture & Food Production @AbpFoods 975 116 3 694 414
Brakes UK Agriculture & Food Production @Brakes_Food 10,711 2608 104 4000 8525
Cargill Agriculture & Food Production @Cargill 57,012 261 933 1940 7551
Cranswick Plc Agriculture & Food Production @CranswickPlc 1865 688 17 804 861
Frontier Agriculture Agriculture & Food Production @FrontierAg 12,479 2964 115 2006 10,098
Glanbia plc Agriculture & Food Production @GlanbiaPlc 5213 651 57 1693 499
IFF Agriculture & Food Production @IFF 5805 2460 93 2444 2599
Kerry Taste & Nutrition Agriculture & Food Production @WeAreKerry 2607 3976 30 804 1685
Meadow Foods Agriculture & Food Production @Meadow_Foods 2783 1236 25 263 1542
Moy Park Agriculture & Food Production @MoyPark 6589 504 33 2672 3640
Openfield Agriculture & Food Production @OpenfieldTM 6830 1780 62 4158 12,044
Smithfield Foods Agriculture & Food Production @SmithfieldFoods 14,291 379 308 2116 7695
Southern Glazer’s Agriculture & Food Production @SGWineSpirits 3337 1056 28 1291 2281
The Mosaic Company Agriculture & Food Production @MosaicCompany 8170 648 160 4 3387
The Scoular Company Agriculture & Food Production @TheScoularCo 436 50 9 70 116
Top Fruit Agriculture & Food Production @topfruitbytes 3603 3359 16 193 3201
US Foods Agriculture & Food Production @USFoods 17,664 2939 278 4254 11,275
Avena Foods Limited Business Software @AvenaFoods 742 679 15 618 340
Capita Business Software @CapitaPlc 6163 1307 129 2977 6654
Cisco Business Software @Cisco 713,167 2911 8436 9259 19,407
Dell EMC Business Software @DellEMC 143,201 3862 2608 5536 24,570
Fidessa Business Software @Fidessa 5506 1922 147 121 2250
IBM Business Software @IBM 610,272 6008 6003 22,037 18,304
Intel Business Software @intel 4,873,838 1371 13,681 9650 15,914
Micro Focus Business Software @MicroFocus 40,026 1528 1356 4506 30,844
MISys Software Business Software @misyssoftware 571 460 25 148 823
NetApp Business Software @NetApp 129,620 3889 1944 12,521 22,818
Oracle Business Software @Oracle 777,554 984 7123 12,182 21,740
Sage Canada Business Software @Sage_Canada 657 114 22 68 311
Salesforce Business Software @salesforce 531,243 151,785 7272 21,671 63,868
SAS Software Business Software @SASsoftware 62,298 608 1780 23,048 46,353
Schneider Electric Business Software @SchneiderElec 59,646 2941 910 16,342 12,115
Sophos Business Software @Sophos 31,390 302 845 13,785 21,006
Symantec Business Software @symantec 205,164 3444 3990 32,260 25,697
Ve Global Business Software @Ve_HeadQuarters 4866 3253 852 5361 23,713
VMware Business Software @VMware 322,803 626 4744 46,587 43,436
Air Products Chemical @airproducts 11,380 551 225 278 3808
Celanese Chemical @celanese 3891 162 124 625 4962
Dow Chemical @DowNewsroom 70,982 1947 993 6949 14,933
DuPont Chemical @DuPont_News 66,690 229 1158 650 23,762
Eastman Chemical Co. Chemical @EastmanChemCo 9042 504 185 1061 1691
F. Ball and Co. Ltd. Chemical @FBallUK 8008 3637 45 5181 13,230
GRI Club Global Chemical @GRIClubGlobal 3627 392 64 79 1520
(continued on next page)
11
(continued )
Huntsman Chemical @Huntsman_Corp 4721 2009 94 300 2780

INEOS Chemical @INEOS 19,485 639 207 3042 4852
LyondellBasell Chemical @LyondellBasell 10,149 1763 180 1438 1762
Nova Laboratories Chemical @NovaLabsUK 163 265 2 211 531
PPG Chemical @PPG 26,489 1019 420 4868 5999
Tennant Company Chemical @TennantCompany 1967 74 34 138 333
Tor Coatings Chemical @TorCoatings 1486 1375 22 248 365
Balfour Beatty Construction @balfourbeatty 69,025 949 570 1317 3968
BAM Nuttall Construction @BAMNuttall 27,430 615 189 2958 5128
Bechtel Corporation Construction @Bechtel 46,835 790 620 2927 7841
Carillion Canada Construction @CarillionCanada 728 202 13 75 880
Clark Construction Group Construction @ClarkBuilds 24,296 764 307 1443 3487
Fluor Corporation Construction @FluorCorp 16,295 678 263 1739 4608
Galliford Try Construction @gallifordtry 12,970 61 147 74 811
Gilbane Building Co Construction @GilbaneBuilding 25,181 2037 412 951 10,726
Interserve Construction @interserve 26,846 278 259 2180 5914
Jacobs Construction @JacobsConnects 38,645 784 748 2356 14,563
KBR Construction @KBRincorporated 5767 494 73 453 1801
Kier Group plc Construction @kiergroup 67,845 602 529 361 7042
Kiewit Construction @kiewit 12,950 101 170 579 2778
Laing O’Rourke Construction @Laing_ORourke 50,850 1560 383 853 4122
Mitie Construction @mitie 21,089 5350 238 2706 8811
Morgan Sindall Group Construction @morgansindall 43,993 231 352 704 4020
PCLConstruction Construction @PCLConstruction 24,745 2142 266 4570 7727
Turner Construction Construction @turner_talk 21,076 188 246 1650 2898
Afren Legal Action Group Energy @saveafrenALAG 967 4854 24 4267 14,985
BG Group Energy @BGGroup 18,665 206 276 130 1289
bp Energy @bp_plc 102,796 301 958 6471 15,210
Centrica Careers Energy @CentricaCareers 720 229 21 315 31,396
Chevron Energy @Chevron 374,244 262 2685 475 15,956
ConocoPhillips Energy @conocophillips 165,144 212 1320 2297 4986
Enterprise Products Energy @EProd_Careers 848 61 119 11 297
Eximius Group Energy @Eximius__Group 606 1915 36 624 1801
ExxonMobil Energy @exxonmobil 325,215 278 2520 361 7076
Hess Corporation Energy @HessCorporation 8326 472 203 422 1334
Marathon Petroleum Energy @MarathonPetroCo 8003 419 157 1102 1749
Petrofac Energy @PetrofacGroup 2681 100 14 424 737
Premier Oil plc Energy @PremierOilplc 5815 115 69 1 568
Sunoco Racing Energy @SunocoRacing 53,287 475 457 7069 17,577
Tullow Oil plc Energy @TullowOilplc 24,087 930 250 33 1109
Valero Energy Energy @ValeroEnergy 8339 95 102 820 828
@aggreko Heavy Industry @Aggreko 5918 574 155 2095 5823
Aptiv Heavy Industry @Aptiv 17,016 531 340 3118 5204
BorgWarner Heavy Industry @BorgWarner 4574 700 60 1383 1976
CaterpillarInc Heavy Industry @CaterpillarInc 137,989 243 1423 1272 12,417
Cosworth Heavy Industry @Cosworth 15,764 346 564 2075 2957
Cummins Inc. Heavy Industry @Cummins 123,884 474 495 3212 3871
Dana Incorporated Heavy Industry @DanaInc_ 2925 270 92 213 1555
J.B. Hunt 360 Heavy Industry @jbhunt360 2026 314 56 834 1861
JCB Heavy Industry @JCBmachines 31,962 835 159 7589 5743
John Deere Heavy Industry @JohnDeere 193,291 223 1297 6059 11,143
Johnson Controls Heavy Industry @johnsoncontrols 27,842 622 411 599 3627
Lear Corporation Heavy Industry @LearCorporation 3837 608 93 4061 2337
Plaxton Heavy Industry @Plaxtoncoach 2410 727 9 376 499
Speedy Services Heavy Industry @wearespeedy 8072 3789 111 2459 14,391
Visteon Corporation Heavy Industry @Visteon 3519 193 126 336 3059
ZF Group Heavy Industry @ZF_Group 8406 1263 171 2757 4382
2 M Automation Industrial Technology @2M_Automation 120 19 5 50 283
Autonomous Solutions Industrial Technology @ASIRobots 1349 228 44 127 767
Emerson Industrial Technology @Emerson_News 18,917 731 368 739 3132
Energid Industrial Technology @energid 635 225 43 123 388
Flowserve Industrial Technology @Flowserve 5481 126 95 409 1536
IMI (now REQ) Industrial Technology @iMarketingInc 6365 2147 292 4980 4096
Mechatronic Solutions Industrial Technology @MechatronicSol 69 24 0 10 108
MKS Instruments Industrial Technology @mksinstruments 878 51 23 39 449
NI (National Instruments) Industrial Technology @NIglobal 27,165 2948 543 9318 8107
Rockwell Automation Industrial Technology @ROKAutomation 39,200 2372 709 11,432 10,230
Sewtec Automation Industrial Technology @Sewtec 333 246 1 190 367
SPX FLOW Industrial Technology @SPXFLOW 1640 793 48 558 1999
AbbVie Medical & Pharmaceutical @abbvie 68,197 498 960 1888 6596
Amgen Medical & Pharmaceutical @Amgen 96,118 350 1575 1456 9895
AstraZeneca Medical & Pharmaceutical @AstraZeneca 229,845 1690 2387 3794 7377
Baxter International Medical & Pharmaceutical @baxter_intl 14,786 2 531 43 1862
Biogen Medical & Pharmaceutical @biogen 36,243 252 555 923 4857
Bristol Myers Squibb Medical & Pharmaceutical @bmsnews 148,807 1114 1768 1115 7421
Eli Lilly and Company Medical & Pharmaceutical @LillyPad 126,705 1344 1463 2026 21,347
(continued on next page)
12
(continued )
FrontierMedicalGroup Medical & Pharmaceutical @FMG_Group 1364 1525 41 1030 4496

GE Healthcare Medical & Pharmaceutical @GEHealthcare 112,641 3175 2220 6900 26,558
Genentech Medical & Pharmaceutical @genentech 118,260 240 1707 1218 7035
Gilead Sciences Medical & Pharmaceutical @GileadSciences 65,857 206 830 969 3451
GSK Medical & Pharmaceutical @GSK 215,856 713 2108 1759 12,091
Merck Medical & Pharmaceutical @Merck 201,762 1000 2041 2072 8604
Pfizer Inc. Medical & Pharmaceutical @pfizer 369,255 2373 3619 525 9859
Renovo.auto Medical & Pharmaceutical @renovo_auto 2035 542 66 669 204
Takeda Medical & Pharmaceutical @TakedaPharma 20,933 268 302 661 2469
Vectura Group Medical & Pharmaceutical @VecturaGroup 511 749 3 297 716
BAE Systems Military Defense @BAESystemsplc 102,676 639 977 2097 5289
Control Risks Military Defense @Control_Risks 16,411 767 394 555 5010
General Dynamics Corporation Military Defense @generaldynamics 2738 251 80 2559 1217
Honeywell Military Defense @honeywell 58,498 324 688 4738 3369
L3Harris Military Defense @L3HarrisTech 16,379 840 447 1520 7798
Lockheed Martin Military Defense @LockheedMartin 444,934 374 4679 8245 16,719
ManTech Military Defense @ManTech 3648 367 119 282 2055
Northrop Grumman Military Defense @northropgrumman 226,348 663 2586 3522 12,988
Oshkosh Corporation Military Defense @oshkoshcorp 918 47 15 318 534
QinetiQ Group Military Defense @QinetiQ 10,963 794 238 1171 4421
Raytheon Technologies Military Defense @RaytheonTech 185,317 273 2439 4633 20,720
SAIC Military Defense @SAICinc 13,004 1695 414 5272 6591
Serco Group Military Defense @SercoGroup 5460 284 46 596 2216
Supacat Military Defense @SupacatLtd 1095 594 6 1758 383
References Buck, B., Minor, K. S., & Lysaker, P. H. (2015). Differential lexical correlates of social
cognition and metacognition in schizophrenia; a study of spontaneously-generated
life narratives. Comprehensive Psychiatry, 58, 138–145.
Adamopoulos, P., Ghose, A., & Todri, V. (2018). The impact of user personality traits on
Burgers, C., Konijn, E. A., Steen, G. J., & Iepsma, M. A. (2015). Making ads less complex,
word of mouth: Text-mining social media platforms. Information Systems Research, 29
yet more creative and persuasive: The effects of conventional metaphors and irony in
(3), 612–640.
print advertising. International Journal of Advertising, 34(3), 515–532.
Agnihotri, R. (2020). Social media, customer engagement, and sales organizations: A
Cartwright, S., Davies, I., & Archer-Brown, C. (2021). Managing relationships on social
research agenda. Industrial Marketing Management, 90, 291–299.
media in business-to-business organisations. Journal of Business Research, 125,
Alboqami, H., Al-Karaghouli, W., Baeshen, Y., Erkan, I., Evans, C., & Ghoneim, A. (2015).
120–134.
Electronic word of mouth in social media: The common characteristics of retweeted
Cartwright, S., Liu, H., & Raddats, C. (2021). Strategic use of social media within
and favourited marketer-generated content posted on twitter. International Journal of
business-to-business (B2B) marketing: A systematic literature review. Industrial
Internet Marketing and Advertising, 9(4), 338–358.
Marketing Management, 97, 35–58.
Alter, A. L., & Oppenheimer, D. M. (2009). Uniting the tribes of fluency to form a
Chiu, C. M., Hsu, M. H., & Wang, E. T. (2006). Understanding knowledge sharing in
metacognitive nation. Personality and Social Psychology Review, 13(3), 219–235.
virtual communities: An integration of social capital and social cognitive theories.
Anand, P., & Sternthal, B. (1990). Ease of message processing as a moderator of
Decision Support Systems, 42(3), 1872–1888.
repetition effects in advertising. Journal of Marketing Research, 27(3), 345–353.
Content Marketing Institute. (2021). 2020 B2B content marketing benchmarks, budgets,
Angst, C. M., & Agarwal, R. (2009). Adoption of electronic health records in the presence
and trends – North America report. Retrieved from https://contentmarketinginstit
of privacy concerns: The elaboration likelihood model and individual persuasion.
ute.com/wp-content/uploads/2019/10/2020_B2B_Research_Final.pdf.
MIS Quarterly, 33(2), 339–370.
Cortez, R. M., Gilliland, D. I., & Johnston, W. J. (2020). Revisiting the theory of business-
Antoniadis, I., Paltsoglou, S., & Patoulidis, V. (2019). Post popularity and reactions in
to-business advertising. Industrial Marketing Management, 89, 642–656.
retail brand pages on Facebook. International Journal of Retail & Distribution
Cox, D. S., & Cox, A. D. (1988). What does familiarity breed? Complexity as a moderator
Management, 47(9), 957–973.
of repetition effects in advertisement evaluation. Journal of Consumer Research, 15
Articulate. (2019). 15 eye-opening B2B social media statistics. Retrieved from: https
(1), 111–116.
://www.articulatemarketing.com/blog/b2b-social-media-statistics.
Cruz, R. E., Leonhardt, J. M., & Pezzuti, T. (2017). Second person pronouns enhance
Backstrom, L. (2013). News feed FYI: A window into news feed. Retrieved from:
consumer involvement and brand attitude. Journal of Interactive Marketing, 39,
https://www.facebook.com/business/news/News-Feed-FYI-A-Window-Into-News-
104–116.
Feed.
Crystal, D. (2006). Language and the internet (2nd ed.). Cambridge: Cambridge University
Barari, M., Ross, M., Thaichon, S., & Surachartkumtonkun, J. (2020). A meta-analysis of
Press.
customer engagement behaviour. International Journal of Consumer Studies. https://
Crystal, D. (2011). Internet linguistics: A student guide. Milton Park, Abingdon. New York:
doi.org/10.1111/ijcs.12609.
Routledge.
Belk, R. W. (2013). Extended self in a digital world. Journal of Consumer Research, 40(3),
Cutler, B. D., & Javalgi, R. G. (1993). Analysis of print ad features: Services versus
477–500.
products. Journal of Advertising Research, 33(2), 62–69.
Berger, J. (2014). Word of mouth and interpersonal communication: A review and
Davis, S. W., Horváth, C., Gretry, A., & Belei, N. (2019). Say what? How the interplay of
directions for future research. Journal of Consumer Psychology, 24(4), 586–607.
tweet readability and brand hedonism affects consumer engagement. Journal of
Berger, J., & Milkman, K. L. (2012). What makes online content viral? Journal of
Business Research, 100, 150–164.
Marketing Research, 49(2), 192–205.
Derks, D., Bos, A. E., & Von Grumbkow, J. (2007). Emoticons and social interaction on
Brandwatch. (2015). B2B Social Media Report. Retrieved from: https://www.
the internet: The importance of social context. Computers in Human Behavior, 23(1),
brandwatch.com/reports/b2b-social-media-report/#download.
842–849.
Brandwatch. (2020). 60 incredible and interesting twitter stats and statistics. Retrieved
Dhaoui, C., & Webster, C. M. (2021). Brand and consumer engagement behaviors on
from https://www.brandwatch.com/blog/twitter-stats-and-statistics/#:~:text
Facebook brand pages: Let’s have a (positive) conversation. International Journal of
=Twitter%20usage%20statistics,That’s%206%2C000%20tweets%20every%20
Research in Marketing, 38(1), 155–175.
second.
eMarketer. (2021). Social networks: Average time spent in the US, by platform,
Brennan, R., & Croft, R. (2012). The use of social media in B2B marketing and branding:
2018–2022 (minutes per day among population). Retrieved from: https://www.
An exploratory study. Journal of Customer Behavior, 11(2), 101–115.
emarketer.com/chart/243761/social-networks-average-time-spent-us-by-platfor
Brown, B. P., Zablah, A. R., Bellenger, D. N., & Johnston, W. J. (2011). When do B2B
m-2018-2022-minutes-per-day-among-population.
brands influence the decision making of organizational buyers? An examination of
Gilliland, D. I., & Johnston, W. J. (1997). Toward a model of business-to-business
the relationship between purchase risk and brand sensitivity. International Journal of
marketing communications effects. Industrial Marketing Management, 26(1), 15–29.
Research in Marketing, 28(3), 194–204.
Gretry, A., Horváth, C., Belei, N., & van Riel, A. C. (2017). “Don’t pretend to be my
Bruhn, M., Schoenmueller, V., & Schäfer, D. B. (2012). Are social media replacing
friend!” when an informal brand communication style backfires on social media.
traditional media in terms of brand equity creation? Management Research Review, 35
Journal of Business Research, 74, 77–89.
(9), 770–790.
13
Grigaliunaite, V., & Pileliene, L. (2016). Emotional or rational? The determination of the McShane, L., Pancer, E., & Poole, M. (2019). The influence of B to B social media message
influence of advertising appeal on advertising effectiveness. Scientific Annals of features on brand engagement: A fluency perspective. Journal of Business-to-Business
Economics and Business, 63(3), 391–414. Marketing, 26(1), 1–18.
Gustafson, B. M., & Pomirleanu, N. (2021). A discursive framework of B2B brand Mehmet, M. I., & Clarke, R. J. (2016). B2B social media semantics: Analysing multimodal
legitimacy. Industrial Marketing Management, 93, 22–31. online meanings in marketing conversations. Industrial Marketing Management, 54,
Halliday, M. A. K. (1976). Functions and universals of language. In G. Kress (Ed.), 92–106.
Halliday: Systems and function in language (pp. 26–31). London: Oxford University Michaelidou, N., Siamagka, N. T., & Christodoulides, G. (2011). Usage, barriers and
Press. measurement of social media marketing: An exploratory investigation of small and
Halliday, M. A. K. (1978). Language as social semiotic. The social interpretation of medium B2B brands. Industrial Marketing Management, 40(7), 1153–1159.
language and meaning. In Open university set book. London: Edward Arnold. Mikk, J. (2008). Sentence length for revealing the cognitive load reversal effect in text
Hootsuite. (2021). 47 facebook stats that matter to marketers in 2021. Retrieved from: comprehension. Educational Studies, 34(2), 119–127.
https://blog.hootsuite.com/facebook-statistics/. Money, R. B. (2004). Word-of-mouth promotion and switching behavior in Japanese and
Hughes, D. J., Rowe, M., Batey, M., & Lee, A. (2012). A tale of two sites: Twitter vs. American business-to-business service clients. Journal of Business Research, 57(3),
Facebook and the personality predictors of social media usage. Computers in Human 297–305.
Behavior, 28(2), 561–569. Morgan, R. M., & Hunt, S. D. (1994). The commitment-trust theory of relationship
Huotari, L., Ulkuniemi, P., Saraniemi, S., & Mäläskä, M. (2015). Analysis of content marketing. Journal of Marketing, 58(3), 20–38.
creation in social media by B2B companies. Journal of Business & Industrial Marketing, Nam, H., Joshi, Y. V., & Kannan, P. K. (2017). Harvesting brand information from social
30(6), 761–770. tags. Journal of Marketing, 81(4), 88–108.
Järvinen, J., & Taiminen, H. (2016). Harnessing marketing automation for B2B content Nguyen, B., Yu, X., Melewar, T. C., & Chen, J. (2015). Brand innovation and social media:
marketing. Industrial Marketing Management, 54, 164–175. Knowledge acquisition from social media, market orientation, and the moderating
Jewitt, C. (2009). The Routledge handbook of multimodal analysis. New York: Routledge. role of social media strategic capability. Industrial Marketing Management, 51, 11–25.
Jünger, J., & Keyling, T. (2019). Facepager. An application for automated data retrieval on Otterbacher, J., Ang, C. S., Litvak, M., & Atkins, D. (2017). Show me you care: Trait
the web (Retrieved from: https://github.com/strohne/Facepager/). empathy, linguistic style, and mimicry on Facebook. ACM Transactions on Internet
Juntunen, M., Ismagilova, E., & Oikarinen, E. L. (2020). B2B brands on twitter: Engaging Technology, 17(1), 1–22.
users with a varying combination of social media content objectives, strategies, and Oz, M., Zheng, P., & Chen, G. M. (2018). Twitter versus Facebook: Comparing incivility,
tactics. Industrial Marketing Management, 89, 630–641. impoliteness, and deliberative attributes. New Media & Society, 20(9), 3400–3419.
Jussila, J., Kärkkäinen, H., & Aramo-Immonen, H. (2014). Social media utilization in Pennebaker, J. W., Boyd, R. L., Jordan, K., & Blackburn, K. (2015). The development and
business-to-business relationships of technology industry firms. Computers in Human psychometric properties of LIWC2015. Austin, TX: University of Texas at Austin.
Behavior, 30, 606–613. https://doi.org/10.15781/T29G6Z.
Kacewicz, E., Pennebaker, J. W., Davis, M., Jeon, M., & Graesser, A. C. (2014). Pronoun Pennebaker, J. W., & King, L. A. (1999). Linguistic styles: Language use as an individual
use reflects standings in social hierarchies. Journal of Language and Social Psychology, difference. Journal of Personality and Social Psychology, 77(6), 1296–1312.
33(2), 125–143. Percy, L. (2012). The role of emotion in processing advertising. In S. Rodgers, &
Kaplan, A. M., & Haenlein, M. (2010). Users of the world, unite! The challenges and E. Thorson (Eds.), Advertising theory (pp. 69–84). New York: Routledge.
opportunities of social media. Business Horizons, 53(1), 59–68. Pitt, C. S., Plangger, K. A., Botha, E., Kietzmann, J., & Pitt, L. (2019). How employees
Karampela, M., Ewelina, L., & McLean, G. (2020). “Just be there”: Social media presence, engage with B2B brands on social media: Word choice and verbal tone. Industrial
interactivity, and responsiveness, and their impact on B2B relationships. European Marketing Management, 81, 130–137.
Journal of Marketing, 54(6), 1281–1303. Quinton, S. (2013). The digital era requires new knowledge to develop relevant CRM
Kaye, L. K., Wall, H. J., & Malone, S. A. (2016). “Turn that frown upside-down”: A strategy: A cry for adopting social media research methods to elicit this new
contextual account of emoticon usage on different virtual platforms. Computers in knowledge. Journal of Strategic Marketing, 21(5), 402–412.
Human Behavior, 60, 463–467. Rich, M. K., & Smith, D. C. (2000). Determining relationship skills of prospective
Khawaja, M. A., Chen, F., & Marcus, N. (2014). Measuring cognitive load using linguistic salespeople. Journal of Business & Industrial Marketing, 15(4), 242–259.
features: Implications for usability evaluation and adaptive interaction design. Ringsberg, T., & Forquer Gupta, S. (2003). The importance of understanding the
International Journal of Human Computer Interaction, 30(5), 343–368. symbolic world of customers in asymmetric business-to-business relationships.
Kim, A. J., & Johnson, K. K. (2016). Power of consumers using social media: Examining Journal of Business & Industrial Marketing, 18(6/7), 607–626.
the influences of brand-related user-generated content on Facebook. Computers in Rodriguez, M. G., Gummadi, K., & Schoelkopf, B. (2014). Quantifying information
Human Behavior, 58, 98–108. overload in social media and its impact on social contagions. In , 8(1). Proceedings of
Kim, J., Kardes, F. R., & Herr, P. M. (1991). Consumer expertise and the vividness effect: the international AAAI conference on web and social media (pp. 170–179).
Implications for judgment and inference. Advances in Consumer Research, 18, 90–93. Rose, S., Fandel, D., Saraeva, A., & Dibley, A. (2021). Sharing is the name of the game:
Krings, W., Palmer, R., & Inversini, A. (2021). Industrial marketing management digital Exploring the role of social media communication practices on B2B customer
media optimization for B2B marketing. Industrial Marketing Management, 93, relationships in the life sciences industry. Industrial Marketing Management, 93,
174–186. 52–62.
Kumar, V., & Mirchandani, R. (2012). Increasing the ROI of social media marketing. MIT Sabate, F., Berbegal-Mirabent, J., Cañabate, A., & Lebherz, P. R. (2014). Factors
Sloan Management Review, 54(1), 55–61. influencing popularity of branded content in Facebook fan pages. European
Labrecque, L. I., Swani, K., & Stephen, A. T. (2020). The impact of pronoun choices on Management Journal, 32(6), 1001–1011.
consumer engagement actions: Exploring top global brands’ social media Schultz, C. D. (2017). Proposing to your fans: Which brand post characteristics drive
communications. Psychology & Marketing, 37(6), 796–814. consumer engagement activities on social media brand pages? Electronic Commerce
Larrimore, L., Jiang, L., Larrimore, J., Markowitz, D., & Gorski, S. (2011). Peer to peer Research and Applications, 26, 23–34.
lending: The relationship between language features, trustworthiness, and Schweidel, D. A., & Moe, W. W. (2014). Listening in on social media: A joint model of
persuasion success. Journal of Applied Communication Research, 39(1), 19–37. sentiment and venue format choice. Journal of Marketing Research, 51(4), 387–402.
Lau, G. T., & Chin, H. W. (2003). Trustworthiness of salespeople in the business-to- Sreejesh, S., Paul, J., Strong, C., & Pius, J. (2020). Consumer response towards social
business market: The five C’s. Journal of Business-to-Business Marketing, 10(3), 1–33. media advertising: Effect of media interactivity, its conditions and the underlying
Lee, D., Hosanagar, K., & Nair, H. S. (2018). Advertising content and consumer mechanism. International Journal of Information Management, 54, 102–155.
engagement on social media: Evidence from Facebook. Management Science, 64(11), Statista. (2019). Most popular mobile social networking apps in the United States as of
5105–5131. September 2019, by average session duration. Retrieved from: https://www.statista.
Leek, S., Canning, L., & Houghton, D. (2016). Revisiting the task media fit model in the com/statistics/579411/top-us-social-networking-apps-ranked-by-session-length/.
era of web 2.0: Twitter use and interaction in the healthcare sector. Industrial Steinhoff, L., Arli, D., Weaven, S., & Kozlenkova, I. V. (2019). Online relationship
Marketing Management, 54, 25–32. marketing. Journal of the Academy of Marketing Science, 47(3), 369–393.
Leek, S., & Christodoulides, G. (2011). A literature review and future agenda for B2B Steuer, J. (1992). Defining virtual reality: Dimensions determining telepresence. Journal
branding: Challenges of branding in a B2B context. Industrial Marketing Management, of Communication, 42(4), 73–93.
40(6), 830–837. Stieglitz, S., & Dang-Xuan, L. (2013). Emotions and information diffusion in social
Leek, S., Houghton, D., & Canning, L. (2019). Twitter and behavioral engagement in the media—Sentiment of microblogs and sharing behavior. Journal of Management
healthcare sector: An examination of product and service companies. Industrial Information Systems, 29(4), 217–248.
Marketing Management, 81, 115–129. Sundström, M., Alm, K. H., Larsson, N., & Dahlin, O. (2021). B2B social media content:
Liu, X. (2020). Analyzing the impact of user-generated content on B2B Firms’ stock Engagement on LinkedIn. Journal of Business & Industrial Marketing, 36(3), 454–468.
performance: Big data analysis with machine learning methods. Industrial Marketing Swani, K., Brown, B. P., & Milne, G. R. (2014). Should tweets differ for B2B and B2C? An
Management, 86, 30–39. analysis of fortune 500 companies’ twitter communications. Industrial Marketing
Luo, J., Toth, Z., Liu, M. J., & Yuan, R. (2021). Social media-related tensions on business- Management, 43, 873–881.
to-business markets–evidence from China. Industrial Marketing Management, 93, Swani, K., & Labrecque, L. I. (2020). Like, comment, or share? Self-presentation vs. brand
293–306. relationships as drivers of social media engagement choices. Marketing Letters, 31,
Lynch, J., & De Chernatony, L. (2004). The power of emotion: Brand communication in 279–298.
business-to-business markets. Journal of Brand Management, 11(5), 403–419. Swani, K., Milne, G. R., & Brown, B. P. (2013). Spreading the word through likes on
Lynch, J., & De Chernatony, L. (2007). Winning hearts and minds: Business-to-business Facebook: Evaluating the message strategy effectiveness of fortune 500 companies.
branding and the role of the salesperson. Journal of Marketing Management, 23(1–2), Journal of Research in Interactive Marketing, 7, 269–294.
123–135. Swani, K., Milne, G. R., Brown, B. P., Assaf, A. G., & Donthu, N. (2017). What messages to
post? Evaluating the popularity of social media communications in business versus
consumer markets. Industrial Marketing Management, 62, 77–87.
14
Tausczik, Y. R., & Pennebaker, J. W. (2010). The psychological meaning of words: LIWC Twitter. (2021b). Getting started guide. Retrieved from: https://help.twitter.co
and computerized text analysis methods. Journal of Language and Social Psychology, m/en/resources/twitter-guide.
29(1), 24–54. Upreti, B. R., Huhtala, J.-P., Tikkanen, H., Malo, P., Marvasti, N., Kaski, S., … Mattila, P.
Thelwall, M., Buckley, K., & Paltoglou, G. (2011). Sentiment in twitter events. Journal of (2021). Online content match-making in B2B markets: Application of neural content
the American Society for Information Science and Technology, 62(2), 406–418. modeling. Industrial Marketing Management, 93, 32–40.
Thelwall, M., Buckley, K., & Paltoglou, G. (2012). Sentiment strength detection for the Van Doorn, J., Lemon, K. N., Mittal, V., Nass, S., Pick, D., Pirner, P., & Verhoef, P. C.
social web. Journal of the American Society for Information Science and Technology, 63 (2010). Customer engagement behavior: Theoretical foundations and research
(1), 163–173. directions. Journal of Service Research, 13(3), 253–266.
Tiwary, N. K., Kumar, R. K., Sarraf, S., Kumar, P., & Rana, N. P. (2021). Impact de Vries, L., Gensler, S., & Leeflang, P. S. (2012). Popularity of brand posts on brand fan
assessment of social media usage in B2B marketing: A review of the literature and a pages: An investigation of the effects of social media marketing. Journal of Interactive
way forward. Journal of Business Research, 131, 121–139. Marketing, 26(2), 83–91.
Trusov, M., Bucklin, R. E., & Pauwels, K. (2009). Effects of word-of-mouth versus Wagner, T. F., Baccarella, C. V., & Voigt, K. I. (2017). Framing social media
traditional marketing: Findings from an internet social networking site. Journal of communication: Investigating the effects of brand post appeals on user interaction.
Marketing, 73(5), 90–102. European Management Journal, 35(5), 606–616.
Tsai, W. H. S., & Men, L. R. (2013). Motivations and antecedents of consumer Zhang, J., & Du, M. (2020). Utilization and effectiveness of social media message
engagement with brand pages on social networking sites. Journal of Interactive strategy: How B2B brands differ from B2C brands. Journal of Business & Industrial
Advertising, 13(2), 76–87. Marketing, 35(4), 721–740.
Twitter. (2021a). Q4 and Fiscal Year 2020 letter to shareholders. Retrieved from: https Zhang, S., Gosselt, J. F., & de Jong, M. D. (2020). How large information technology
://s22.q4cdn.com/826641620/files/doc_financials/2020/q4/FINAL-Q4’2 companies use twitter: Arrangement of corporate accounts and characteristics of
0-TWTR-Shareholder-Letter.pdf. tweets. Journal of Business and Technical Communication, 34(4), 364–392.
15

Industrial Marketing Management: Qi Deng, Yun Wang, Michel Rod, Shaobo Ji

Uploaded by

Copyright:

Available Formats

Industrial Marketing Management: Qi Deng, Yun Wang, Michel Rod, Shaobo Ji

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Industrial Marketing Management: Qi Deng, Yun Wang, Michel Rod, Shaobo Ji

Uploaded by

Copyright:

Available Formats

Industrial Marketing Management 99 (2021) 1–15

Contents lists available at ScienceDirect

Industrial Marketing Management

Fig. 1. Brand Engagement in B2B Social Media: A Conceptual Framework.

Appendix A. Sample description

Alcoa Aerospace @Alcoa 27,542 5176 664 3041 6660

Huntsman Chemical @Huntsman_Corp 4721 2009 94 300 2780

FrontierMedicalGroup Medical & Pharmaceutical @FMG_Group 1364 1525 41 1030 4496

You might also like