Can emojis and B2B mix? The effects of emojis and emoji–text interactions on B2B social media engagement

Business-to-business (B2B) firms are increasingly integrating social media into their marketing strategies because of its growing use by customers and other key stakeholders. In fact, 84% of B2B executives rely on social media content to inform their purchase decisions, and B2B stakeholders (e.g. customers, employees, professionals and civilian social media users) often engage with an organization’s social media posts before interacting directly with the firm (Dean, 2023; Järvinen and Taiminen, 2016; Huotari et al., 2015). Evidence suggests that social media positively influences B2B marketing performance through various B2B sub-domains such as market relations, channels and intermediaries, branding and knowledge management (Wang et al., 2017; LaPlaca and Katrichis, 2009; Young et al., 2015). Beyond sales and customer relationship management (e.g. prospecting, leads generation, customer services and satisfaction) (Cortez et al., 2023; Cao and Weerawardena, 2023), social media enhances B2B firms’ capability of building and maintaining dynamic associations, communications and collaborations among B2B stakeholders (Deng et al., 2021; Huotari et al., 2015). This interconnectedness is a core competitive advantage for B2B firms (Tiwary et al., 2021). Furthermore, it aligns with the Industrial Marketing and Purchasing Group’s network approach, which accounts for the contemporary complexities of the digital and globalized B2B environment (Håkansson and Ford, 2002). By engaging with different stakeholders on social media without time and space restrictions (Huotari et al., 2015; Eid et al., 2020), B2B firms can build a dynamic and multi-actor network that fosters value co-creation, adaptive selling and proactive servicing (Dwivedi et al., 2023; Huotari et al., 2015; Sundström et al., 2021). Furthermore, along with social media interaction and engagement, there is a wealth of user-generated content and behavioral data that can generate market and competitor intelligence and provide insights for both product/market innovation and competitor benchmarking (Corral de Zubielqui and Jones, 2023; Rooderkerk and Pauwels, 2016). Ultimately, social media engagement strengthens B2B firms’ authority and brand awareness, contributing to brand equity and indirectly enhancing marketing, organizational and financial performance (Cao and Weerawardena, 2023; Eid et al., 2020; Liu, 2020; Cortez et al., 2023; Rooderkerk and Pauwels, 2016).

It is, therefore, crucial for B2B marketers to enhance the effectiveness of social media strategies. Evident here is the critical role of social media engagement in B2B social media strategy (Salo, 2017). Defined as the intensity of viewer interactions and involvement with a B2B firm’s social media post and often operationalized via viewer behaviors such as likes, comments and shares (Balaji et al., 2023; Gu et al., 2023), social media engagement determines the quality and quantity of user-generated content and social interactions on social media. Without active audience engagement, none of aforementioned marketing goals can be achieved. However, creating social media content that drives engagement remains the top content marketing challenge for B2B marketers, regardless of company size or marketing budget (Content Marketing Institute, 2022). To date, the overarching relationship between social media content and engagement has primarily garnered significant attention in the B2C sector (Swani et al., 2017). Yet, in contrast, social media engagement in the B2B sector is less understood, and the nuances of how to elicit strong engagement via social media posts remain an understudied area of significant strategic importance for B2B organizations (Cartwright et al., 2021; Salo, 2017). In fact, there is a paucity of research investigating language use in B2B social media posts (Mehmet and Clarke, 2016; Deng et al., 2021; Leek et al., 2019). The limited body of work that has done so, however, suggests that language plays a central role in effective B2B communications. It is, thus, critical to understand the dynamics of how language drives social media engagement within the B2B context (Deng et al., 2021; Leek et al., 2019).

Emojis are iconic visual representations of facial expressions, people, things or activities (e.g. , , , , , , and ) that are rapidly becoming a mainstream part of social media communication. Using X (formerly Twitter) as an example, an analysis of nearly three billion global tweets showed that over 20% of tweets include emojis (Broni, 2021). Furthermore, 92% of online populations use emojis in digital communications, and more notably, 76% use them in professional settings (Daniel, 2021). Paralleling this rise in popularity, emojis are rapidly being embraced by B2B brands and incorporated into their social media posts with an aim of stimulating engagement (Balaji et al., 2023; Mehmet and Clarke, 2016). While B2B practitioners actively experiment with emojis in their social media posts, research examining emojis’ communicative effects in the B2B arena is surprisingly rare (Sundström et al., 2021). In contrast, an emerging surge of B2C studies has examined emojis’ role in enhancing social media engagement (Li et al., 2019; Wang et al., 2023; McShane et al., 2021). While B2C research highlights emojis’ significant role in driving social media engagement, it lacks clarity on whether B2B firms can similarly benefit from emoji use given that the overall communications strategies and nature of social media posts in B2B are distinct from those in B2C organizations (Kwon et al., 2022; Mero et al., 2023; Valenzuela-Gálvez et al., 2023). Unlike B2C marketing, B2B communications prioritize trust, credibility and value creation, which may affect how emojis are perceived by stakeholders (López-López and Giusti, 2020). B2B firms still face significant challenges in creating social media posts that optimally integrate emojis to effectively engage viewers (Balaji et al., 2023; Deng et al., 2021), thus preventing them from leveraging social media to its fullest potential as compared to B2C firms (Tiwary et al., 2021). Therefore, extant research recommends separate examinations of emojis’ effects in the B2B and B2C contexts (Balaji et al., 2023; Deng et al., 2021; Wang et al., 2023).

Drawing on the B2B communication model, fluency theory and extant research on emojis’ communicative effects, this study conducts an in-depth examination of how emoji use in social media posts influences B2B social media engagement. Specifically, we examine: the overall effect of emoji use in B2B social media posts on engagement; and how structural variations in how emojis are integrated with text impact engagement. By analyzing 64,121 tweets from 82 brands across 19 industries, we find an inverted U-shaped relationship between emoji count and B2B social media engagement. That is, while moderate use of emoji can enhance engagement, excessive use reduces it. More interestingly, we find that this relationship is significantly influenced by where emojis are placed in the message and whether they reinforce or substitute textual content.

These findings offer important theoretical implications for B2B social media engagement, communication theories and the role of emojis in multimodal social media content. Firstly, by integrating the B2B communication model and fluency theory, we demonstrate that emojis encompass dual roles in enabling both central and peripheral information processing. When used judiciously, emojis can enhance the emotional tone and visual appeal of a message, acting as peripheral cues that foster interpersonal and emotional relationships between brands and consumers. However, when overused, emojis may impair central processing by lowering the readability and clarity that are essential for processing the technical and information-rich content typical of B2B communication. This can undermine perceptions of professionalism of B2B brands, disrupt cognitive processing of brand messages and ultimately reduce engagement. This extends previous work that treats emojis solely as emotional cues operating through the peripheral route processing (Deng et al., 2021). Secondly, the results also extend the multimodality perspective by revealing the impact of emoji–text interaction on B2B social media engagement. This approach extends previous research that primarily considers multimodal elements as either present or absent (e.g. a brand post has/does not have an image: Deng et al., 2021) to consider how the emojis are used in relation to the textual cues in the post. Finally, the findings also contribute to B2B marketing practice by providing many easy-to-implement tactics regarding effectively using emojis to stimulate engagement on social media.

The remainder of this paper is organized as follows. Section 2 reviews the most relevant literature and introduces our theoretical underpinnings – the B2B communication model and fluency theory. We then develop our specific hypotheses in Section 3 and discuss our chosen methodology in Section 4. In Section 5, we present our empirical findings. In Section 6, we conclude with an extensive discussion of the theoretical and practical implications of the current research, its potential limitations and the opportunities for future research in this area.

With emojis becoming a universal and intuitive language for digital communication, they are frequently used in brand social media posts to enhance message clarity, expressiveness and relatability (Li et al., 2019; Wang et al., 2023). The effectiveness of emojis in B2C social media posts is widely examined, with studies revealing emojis’ ability to evoke positive emotional responses (Li et al., 2019; McShane et al., 2021) and drive social media engagement (Davis et al., 2019; McShane et al., 2021; Ko et al., 2022). However, the application of emojis in B2B social media posts remains underexplored, despite early studies suggesting emojis can play a strategic role in B2B communications by softening formal messaging, enhancing relatability and signaling emotional intelligence in brand communication (Deng et al., 2021; Sundström et al., 2021). For example, emojis have been found to facilitate peripheral route processing by adding visual appeal to B2B posts, potentially increasing viewer engagement (Balaji et al., 2023). Similarly, using emojis in B2B brand tweets has been found to positively impact brand engagement (Deng et al., 2021). On other platforms such as WeChat (a popular social networking platform in east Asia), emojis like handshakes or hugs have demonstrated friendliness, enhancing intimacy in B2B interactions (Weng et al., 2024). Additionally, emojis in business emails have been positively associated with perceptions of likeability and effectiveness (Riordan and Glikson, 2020). A common underlying assumption of these studies is that emojis act as emotional cues in B2B messages and mainly drive emotional responses and perceptions via their effects on the peripheral route of information processing.

Despite these few pioneer B2B studies on emojis, many unresolved questions and issues remain regarding the effects of emojis in B2B social media engagement (Sundström et al., 2021). For example, given that the pioneer studies adopt the view of emojis as emotional cues that may facilitate viewers’ peripheral route of information processing (Balaji et al., 2023; Deng et al., 2021), it remains unclear whether and how emojis influence viewers’ cognitive information processing and, in turn, their social media engagement behaviors. Consideration of this potential impact of emojis on central route processing is particularly critical for B2B social media content, which is typically more informational and complex, so more dependent on cognitive processing than B2C content (Deng et al., 2021; Swani et al., 2014; Swani et al., 2017; Gu et al., 2023). Moreover, existing studies primarily focus on whether emojis are present or absent in social media posts, without considering specific aspects of emoji use, such as the number of emojis in the message, the placement of the emoji(s) relative to the text and the emoji(s)’ communicative role in the message. Nevertheless, extant research indicates that message elements do not act in isolation but interact with each other to influence social media engagement (Balaji et al., 2023). This research gap leads to little clarity around if, when, why and how emojis influence B2B social media engagement, highlighting the need for a more systematic understanding of emojis’ effects in B2B social media posts. The current study aims to address these research gaps. We now turn to the theoretical foundation of the current study.

The current research draws on the B2B communication model to understand the role of emojis in B2B social media communications. Initially developed by Gilliland and Johnston (1997) and rooted in the elaboration likelihood model, this model explains attitude formation and change via dual process persuasion: central route processing (cognitive, rational) and or peripheral route processing (emotional, heuristic) (Petty and Cacioppo, 1984; Angst and Agarwal, 2009). The B2B communication model was recently adapted by Cortez et al. (2020) to reflect the rising influence of social media and emotional appeals in industrial communications, underscoring that B2B communication engages both cognitive responses – elicited through rational, central route processing – and affective responses, triggered by emotional, peripheral route processing. It highlights that both routes are significant; both cognitive and emotional responses can yield beneficial outcomes, such as increased social media engagement (Cortez et al., 2020). Here, peripheral route processing relies on heuristic and emotional cues, while central route processing emphasizes argument quality and informativeness – accessed through timeliness, relevance, comprehensiveness and accuracy – to form attitudes and evaluations (Cheung and Thadani, 2012; Gilliland and Johnston, 1997; Petty and Cacioppo, 1984). Compared to peripheral route processing, central route processing is linked to higher levels of elaboration, typified by devoting more attention and cognitive resources to the task. Assessments of weak argument quality or informativeness have been shown to lead to less successful persuasion attempts (Petty and Cacioppo, 1984).

A dominant perspective in these dual-process theories is that the peripheral and central route processing represent two ends of a spectrum where the extent of elaboration depends on the situational involvement of the message receiver (Gu et al., 2023; Petty and Cacioppo, 1984). However, extant work on dual process theories in the area of digital persuasive messages argues that affective and cognitive processing systems, although independent, likely work in an interactive way (SanJosé-Cabezudo et al., 2009). Recent work finds that persuasive messages necessarily elicit neither peripheral route processing nor central route processing, but rather that both routes may be activated in the online persuasion process (El Hedhli and Zourrig, 2023). This is consistent with recent applications of the B2B communication model, which suggests that the processing of digital messages is a non-linear and a co-creative process in part because of the multimodal nature of digital messages that blend textual language with other elements such as images, emojis and hyperlinks (Deng et al., 2021; Mehmet and Clarke, 2016). B2B audiences have been shown to respond to messages in a nuanced way, blending rational and emotional elements rather than leaning entirely toward one route or another (Cortez et al., 2020; Deng et al., 2021). The current research applies the B2B communication model to understand how emojis impact B2B social media engagement, and it does so by adopting the more recent perspective that peripheral and central route processing may operate interactively in digital persuasion attempts.

A growing body of research on B2B social media engagement is drawing on the B2B communication model, or related dual-process theories of persuasion, to explore how brand post features influence social media engagement (e.g. linguistic message features: Deng et al., 2021; visual information cues: Gu et al., 2023; functional and emotional appeals; Swani et al., 2017). A critical insight from this research is that brand post features can drive social media engagement to the extent that they facilitate central and/or peripheral route processing (Deng et al., 2021; Gu et al., 2023). Deng et al. (2021), in particular, adopt the perspective that certain linguistic features in B2B social media messages may stimulate both peripheral and central route processing. This is consistent with recent developments in dual process theories in online persuasion (SanJosé-Cabezudo et al., 2009; El Hedhli and Zourrig, 2023). Furthermore, one of the tenets of dual-process theories is that the same variable can play multiple roles in persuasion attempts, serving as a heuristic and/or emotional cue for peripheral processing and as an argument for central processing (Petty and Caccioppo, 1984). Pairing this tenet with the perspective that peripheral and central route processing can co-occur in digital persuasion attempts, we posit that B2B social media messages with emojis can affect both peripheral and central route processing to impact B2B social media engagement.

From a peripheral route perspective, research on emojis in B2B social media highlights that a central function of emojis is their capability to enrich the emotionality and affective signaling of digital communications and to nurture relationships (Balaji et al., 2023; Deng et al., 2021; McShane et al., 2019; Swani et al., 2014). The power of emotions in B2B social media engagement has been increasingly recognized, with both emotional cues and interpersonal cues being identified as critical to B2B social media engagement (Deng et al., 2021; Mehmet and Clarke, 2016; Swani et al., 2017). In fact, emotional appeals have been shown to significantly enhance B2B message liking, as well as the perceived user-friendliness and persuasiveness of the message (Swani et al., 2017; Deng et al., 2021; Kaye et al., 2016). Taken together, we thus expect that emojis are likely to enhance social media engagement to the extent that they enrich the emotionality of brand posts and facilitate peripheral route processing.

From a central route perspective, extant literature on B2B social media engagement suggests that viewers of B2B social media content are likely to engage in central route processing because such messages typically contain more technical and complex offerings and/or often contain informational and functional messages that viewers are motivated to process and understand (Swani et al., 2014; Swani et al., 2017; Gu et al., 2023). We anticipate that emojis will influence central route processing through their effects on message comprehensibility, a key dimension of argument quality (Petty and Cacioppo, 1984). Unlike traditional visuals, such as images or videos, which are typically separate from text, emojis are more versatile visual cues that are frequently intertwined with text. For instance, in the CDW social media posts, emojis serve as visual reinforcements (e.g. “The supercomputers have joined the fight to find a treatment. ”) or as substitutes for words (e.g. “#Smallbiz is the latest hot for #ransomware attacks.”). We argue that these unique characteristics of emoji–text multimodality can influence message comprehensibility (linked to central route processing), leading to corresponding effects on social media engagement. We now turn to the literature on processing fluency to explore this idea further.

Processing fluency refers to the ease with which information is processed and understood (Alter and Oppenheimer, 2009). Of particular interest for the current work is that of orthographic fluency, which refers to “ease with which people are able to translate written information into comprehensible language” (Alter and Oppenheimer, 2009, p. 225). Extant literature on orthographic fluency finds that messages interspersed with symbols can reduce message comprehensibility by making the content more difficult for message viewers to interpret message meaning than those without symbols (e.g. G@dget$@nd GizmØ$’ vs Gadgets and Gizmos). In the B2B social media literature, such effects have been shown when multiple words are written without the correct spacing (e.g. socialmediaengagement vs social media engagement) and particularly in the case of hashtags that similarly merge words (McShane et al., 2019). This has important implications for central route processing in that it suggests that messages presented in ways that are easy to process (i.e. high processing fluency) will facilitate message comprehension, thus facilitating central route processing. Notably, extant research also finds that processing fluency can positively impact B2B social media content engagement (McShane et al., 2019).

Taken together then, and drawing on the B2B communications model and fluency theory, we thus first propose that emojis, as emotional cues, can influence peripheral route processing, thereby impacting social media engagement. Second, we propose that the ways in which emojis are integrated with textual content can influence central route processing, thereby influencing social media engagement.

In this section, we develop specific hypotheses regarding how emojis will impact B2B social media engagement. To do so, we consider emojis from two perspectives. First, we consider emojis as independent visual cues with emotional appeal that impact B2B social media engagement via peripheral route processing. Second, we consider emojis as linguistic features that can be uniquely intertwined with text in ways that influence argument quality and, thus, impact B2B social media engagement via central route processing. Together, we examine: how the number of emojis in a B2B message (i.e. emoji count) might impact engagement; how emoji–text interactions (i.e. the location of emojis relative to the text and whether the emoji serves to reinforce the text or rather substitute for text) impact engagement and moderate the effect of emoji count (Figure 1).

B2B social media posts often vary in the number of emojis included (e.g. “Share a pic of your favorite CiscoLive ! Hats-off to #NationalHatDay” – three emojis vs “Wake up to how we named Java! ” – one emoji). Such variation may significantly influence engagement by altering the emotional and visual salience of the post, thereby activating peripheral route processing. Peripheral processing is typically triggered by emotional appeals that elicit favorable affect, leading message viewers to evaluate the message more positively without extensive cognitive elaboration (Gilliland and Johnston, 1997; Petty and Cacioppo, 1984). Prior research shows that messages with emotional appeals are more persuasive and increase the popularity of B2B brand posts relative to B2C posts (Swani et al., 2017). Along similar lines, images frequently serve as peripheral cues that can facilitate persuasion by either engendering affect that is then transferred to the message and/or drawing the message viewers’ attention, which then engenders more positive message evaluations (Miniard et al., 1991; Petty and Cacioppo, 1984). In fact, emotional appeals presented as visual content have been shown to have a stronger effect on B2B social media engagement than non-visual emotional appeals (Gu et al., 2023).

Emojis, as emotionally expressive and visually engaging symbols, are effective tools for infusing emotion into B2B brand posts (Balaji et al., 2023; Deng et al., 2021; McShane et al., 2019; Swani et al., 2014) and are, therefore, likely to increase engagement. However, unlike B2C communication, which often allows for informal and emotionally intensive language, B2B communication is characterized by expectations of professionalism, clarity and information-rich content (Swani et al., 2014). In such environments, excessive emoji use may undermine message credibility or distract from core business content (López-López and Giusti, 2020). As such, we anticipate that, at low to moderate levels, the use of emojis may enhance emotional tone, improve perceived interpersonal warmth, draw attention and increase engagement by offering subtle cues that facilitate peripheral message processing. However, we anticipate that excessive use of emojis may reduce perceived professionalism, compromise the perceived seriousness of the brand, introduce visual clutter and/or create a cognitive burden for B2B audiences (López-López and Giusti, 2020). Thus, rather than a linear positive relationship, we expect the effect of emoji count on engagement to follow an inverted U-shaped pattern, whereby a small number of emojis may enhance engagement, but too many may hinder it. Such a relationship aligns with prior findings that excessive peripheral cues can backfire when they overwhelm the message or reduce source credibility (Petty and Cacioppo, 1984). Therefore, we predict:

In addition to emoji count, we anticipate that the location of emojis in B2B social media posts plays a critical role in shaping engagement outcomes. Emojis can be placed outside the main body of text (before or after the textual content of the message) or embedded within the text (either all together or interspersed throughout the text when there are multiple emojis). While both location types serve as visual cues, their impact on message processing is likely to differ. B2B social media content is typically more technical, information-dense and goal-driven than its B2C counterpart, leading audiences to rely more on central route processing when evaluating such messages (Cartwright et al., 2021; Swani et al., 2014, 2017; Gu et al., 2023). In B2B contexts that prioritize informational clarity, professional tone and message efficiency, we expect that the way in which emojis are placed relative to the text can influence central route processing through their differential effects on message comprehensibility and professionalism, two key attributes that shape argument quality and, by extension, user engagement (Swani et al., 2014; López-López and Giusti, 2020; Petty and Cacioppo, 1984).

While emojis can convey simple emotions and ideas, they are inherently simplistic visual symbols that may not serve as adequate substitutes for the detailed, functional and objective information that is expected by B2B audiences (Gu et al., 2023). When integrated within the text (versus before or after textual content), emojis interrupt the syntactic structure of the message, forcing viewers to navigate around the icons to piece together the segmented textual information that enables them to understand the informational content. This interruption disrupts the natural flow of reading, increases cognitive load and decreases processing fluency, particularly in B2B posts where the textual content is often dense and information-heavy (Bashirzadeh et al., 2022). For B2B audiences, who often require efficient information processing for professional purposes, this disruption can negatively impact message comprehensibility and central route processing, which may reduce their engagement with the post. In contrast, emojis placed outside the text allow readers to process the textual content sequentially and uninterrupted, facilitating easier extraction of meaning. This distinction is crucial for central route processing, as ease of comprehension directly impacts assessments of argument quality.

Furthermore, research indicates that translating textual and visual content relies on distinct brain areas (Bashirzadeh et al., 2022; Kaye et al., 2016). In the case of emojis outside the text, readers can easily ignore the emojis and focus solely on the uninterrupted string of textual information, therefore simplifying cognitive processing. In contrast, when emojis appear within the text, message viewers must either switch between processing textual and visual elements or cognitively skip over them to continue extracting meaning from the message. In both cases, the emojis divert viewers’ cognitive resources away from understanding the core message. The cognitive effort required to integrate emojis with surrounding text is amplified in the B2B context, where the content is often designed to deliver technical, analytical or functional information and audiences often engage with content for professional purposes and value message clarity, brevity and logic (Swani et al., 2017). This added cognitive effort may reduce overall engagement, as viewers might find that B2B posts with embedded emojis (i.e. emojis placed within the text) appear more fragmented and less professional and are harder to process. Thus, we predict:

Moreover, emoji location may influence how emoji count is processed. When emojis are placed outside the text, they remain visually distinct without fragmenting the message regardless of their counts, allowing viewers to process textual information more efficiently. As a result, the inverted U-shaped relationship between emoji count and engagement is likely preserved, as emojis do not interfere with the primary content. When emojis appear within the text, translating the content into meaning becomes more difficult because the message is segmented by these non-informational visual cues that must either be translated into meaning using a different part of the brain or skipped over to continue extracting informational textual content. In this case, each additional emoji may increasingly disrupt message processing fluency and comprehension, which is critical for central route processing in B2B audiences (Petty and Cacioppo, 1984; McShane et al., 2019). Therefore, the cumulative effect of more emojis may feel overwhelming or unprofessional, accelerating the diminishing returns and weakening the peak of engagement. As such, emoji location may serve as a boundary condition that moderates the curvilinear relationship between emoji count and engagement. Thus, we predict:

Beyond emoji count and location, the functional role of emojis may significantly shape how B2B audiences respond to social media posts. Drawing from research on multimodal communication, messages can combine textual and visual elements to either reinforce meaning or substitute one modality for another (Cohn, 2016). Both approaches are evident in B2B social media communication. For example, emojis may be used to reinforce the message’s tone or content, as seen in posts like “Share a pic of your favorite CiscoLive hat ,” where the emoji enhances the visual appeal of the message but is less relevant to constructing message meaning. Alternatively, emojis can be used to substitute for text, conveying the core idea with minimal written explanation, such as in the post “Share a pic of your favorite CiscoLive ,” where the emoji alone captures the intended message.

However, in B2B contexts, where audiences prioritize clarity, precision and professionalism to meet the informational needs (Swani et al., 2014; Swani et al., 2017), the functional role of the emojis (i.e. either as text substitutions or reinforcements) may significantly influence B2B social media engagement because of their differential effects on message comprehension and, by extension, central route processing. When emojis are used to reinforce the text, message viewers can extract informational content purely from the text without needing to understand emoji meaning. In this case, emojis can add visual and emotional appeal without impairing central route processing. In contrast, in the case of emojis used as text substitutions, message viewers must process both visual and text cues to achieve message comprehension, thereby requiring the activation of two distinct areas of the brain to interpret the message meaning (Bashirzadeh et al., 2022; Kaye et al., 2016). Furthermore, substitutive emojis have been noted as requiring multimodal decoding (Orazi et al., 2023). That is, to achieve message comprehension, a substitutive emoji must first be processed visually and then phonologically, thus necessitating greater cognitive resources than reinforcement emojis to comprehend the message (Orazi et al., 2023). The limited work examining the downstream effects of the emoji role shows that reading a message with emojis as substitutions increases reading times relative to those using emojis as reinforcements (Cohn et al., 2018), which is consistent with this argument. Recent neuroscience research finds that multimodal integration, that is, visual information from the emojis and linguistic information from words, can lead to higher semantic processing difficulty (Pfeifer et al., 2022).

Moreover, when emojis are used as substitutions for words, they inherently reduce the textual content of the message and allow for a broad search of meaning (Pfeifer et al., 2022), thus potentially creating ambiguity or leaving critical information open to interpretation. For example, replacing a word like “target” with in a B2B social media post may confuse readers who expect precise terminology. In contrast, emojis used as reinforcements (e.g. “target ”) complement the textual content, preserving clarity while adding visual emphasis. In B2B contexts, such confusion can undermine the sense of certainty conveyed by the message, which may further reduce B2B social media engagement (Deng et al., 2021). Thus, we predict:

We expect emoji role (reinforcement versus substitution) to moderate the positive effects of emoji count predicted in H1. When emojis are used as reinforcements, they complement the text by emphasizing key points without reducing the amount of information conveyed. In this case, multiple emojis serve as non-disruptive visual cues that enhance the emotional tone and visual interest of the message. As the number of emojis increases (up to an optimal point), their cumulative impact may strengthen engagement without undermining message comprehensibility. However, when emojis are used as substitutions, each additional emoji further reduces the textual content, requiring the message viewers to engage more in multimodal decoding to translate the message into meaning (Bashirzadeh et al., 2022; Kaye et al., 2016; Orazi et al., 2023). Emojis used as substitutions requires audiences to decipher the meaning of each emoji while simultaneously extracting meaning from the surrounding text to understand the message. As emoji count increases, this process becomes increasingly demanding, making the message harder to comprehend and mitigating their effectiveness in driving engagement (Gu et al., 2023; Pancer et al., 2019).

Moreover, professionalism and credibility are critical in B2B social media communication (Deng et al., 2021; López-López and Giusti, 2020). When emojis are used as reinforcement, they may signal creativity and approachability without compromising the professional tone, as the core informational content remains intact. However, substituting multiple words with emojis may make the message appear overly casual or unprofessional, particularly in B2B contexts where precise language is usually expected (López-López and Giusti, 2020). This mismatch between audience expectations and message tone can reduce the positive effect of additional emojis on engagement. This suggests that the inverted U-shaped relationship between emoji count and engagement (as proposed in H1) will be flattened when emojis are used as substitutions rather than reinforcements. Therefore, we predict:

To test our hypotheses, we conducted a field study where we created a data set using the top 100 global B2B brands identified and ranked by MERIT based on their digital engagement, employment sentiment and market performances (MERIT, 2017). These B2B brands were identified as digitally savvy market leaders, thus suitable for the current study. We collected our data from X (Twitter), as it has become one of the most important social media platforms for B2B marketing (Leek et al., 2019; McShane et al., 2019). According to a recent report, 71% of B2B marketers used X (Twitter) for content marketing (Content Marketing Institute, 2022). Besides, as a public social media platform, X (Twitter) provides a context where B2B social media engagement can be observed unobtrusively.

We used the X (Twitter) application program interface v2 to collect the previous tweets of these brands that were posted between January 1, 2020 and May 31, 2022. This period covers 29 months in total, thus mitigating any potential seasonal fluctuations of brand engagement. We conducted data collection at the end of July 2022 to leave a two-month gap to avoid potential changes in engagement after data being recorded. For each tweet, we collected the full text, engagement counts (i.e. the number of likes, replies, retweets and quotes), posting date and time, media types and other elements incorporated in the tweets (e.g. hashtag and user mentions). When scraping the tweets, we excluded brand replies and retweets because these tweets are not original brand posts. For brands with multiple X (Twitter) accounts, we scraped tweets from the account with the most followers. We excluded brands that do not have X (Twitter) accounts. We also excluded eight brands, that is, Microsoft, Google, Amazon, Facebook, Expedia, FedEx, UPS and HP, that may not be considered B2B brands to their social media audience. Our data set includes 64,547 tweets from 82 brands representing 19 industries, such as Construction, Chemical, Aerospace, Business Services, Military Defense, Technology and Energy (see  Appendix for the sample description).

Dependent variable: To operationalize brand social media engagement, we used four common X (Twitter) engagement metrics, that is, likes, replies, retweets and quotes (i.e. retweets with comments). These metrics have become the prevalent objectives in social media marketing for both B2B and B2C companies and have been extensively used in previous literature as important B2B social media engagement measures (Davis et al., 2019; Deng et al., 2021; Leek et al., 2019; McShane et al., 2019; Silva et al., 2020). Specifically, we followed previous research and used the sum of the brand posts’ likes, replies, retweets and quotes as the measure of social media engagement (Aydin, 2020; Cruz et al., 2017; Khan and Dongping, 2017; Tafesse and Wien, 2018; Demmers et al., 2020; Singh et al., 2023).

Independent variables: We included a series of variables to measure the use and features of emojis. To calculate emoji count (i.e. the number of emojis contained in a tweet), we developed a Python script based on a Python library named Emoji (Kim and Wurster, 2023). This provided an objective approach for counting the emojis in each tweet. The results showed that 11,949 tweets (18.5% of our sample) contained at least one emoji, of which 7,181 (60.1%) tweets contained a single emoji and 4,768 (39.9%) tweets contained multiple emojis. Of the tweets that contained emojis, the average emoji count was 1.82, with the maximum emoji count of 121. To measure emoji location, we created a dummy variable to determine whether the emojis were located inside the tweet’s text (i.e. emojis are in the middle of the tweet or spread throughout the tweet) or outside the tweet’s text (i.e. emojis are in the beginning or the end of the tweet). Notably, when emojis are placed at the end of one sentence but not at the end of the entire tweet, we classify them as “inside the tweet’s text”. To determine the emoji role, two coders who were unaware of our hypotheses were recruited and coded the 11,949 tweets containing emojis. Before coding, the coders were provided operational definitions and examples of various emoji roles. Then, the two coders independently coded the tweets. For tweets with a single emoji, the emoji role was coded either as “word reinforcement” or “word substitution.” For tweets with multiple emojis, the emoji role was coded into “word reinforcement,” “word substitution” or “both” (i.e. some emojis in the tweet were used as word reinforcement and others as word substitutions). The overall reliability of the coding was 0.929, with a reliability of 0.961 for tweets with single emojis and 0.881 for tweets with multiple emojis. Consensus was reached through discussions between the two coders. Given that our focus was on comparing emojis used as word reinforcements to those used as word substitutions, we removed the 426 tweets that were coded “both.” This was necessary to create a unified coding of the emoji role (either reinforcement or substitution) for all brand posts, regardless of whether they had single or multiple emojis. We then created a dummy variable based on the coding results to measure whether the emojis were used as word substitutions (vs word reinforcement).

Control variables: We included a variety of control variables to account for their potential influence. Previous research has shown that emotional words and the readability of brand posts can influence consumer engagement (Deng et al., 2021; McShane et al., 2019; Pancer et al., 2019). So we included two variables to control these effects. We used the 2022 Linguistic Inquiry and Word Count to capture the percentage of emotional words in brand posts. Linguistic Inquiry and Word Count is computational linguistic software that can be used to assess a text’s linguistic style in approximately 90 predefined linguistic categories (Pennebaker et al., 2015). It has been widely adopted in social research. We used the Dale Chall Score of the tweets to capture the readability (Chall and Dale, 1995). The Dale Chall Score is one of the most well-established readability measures in linguistics and has been widely used in marketing research (McShane et al., 2019; Pancer et al., 2019). It evaluates the text readability by considering the use of familiar (vs unfamiliar) words and average sentence length, and the higher the Dale Chall Score, the lower the readability. To control the emotion carried by emojis per se, we included a dummy variable to measure whether a tweet contains face emojis. Previous research has shown that emojis depicting faces (i.e. “face” emojis) are likely to be interpreted as conveying emotion, whereas those that depict objects, activities or concepts (i.e. “non-face” emojis) instead tend to be perceived as conveying semantic information and do not convey emotion (Cao et al., 2024; Orazi et al., 2023). As such, controlling for whether the emoji is a “face” versus “non-face” emoji serves as an effective proxy by which to account for the emotional content of the emoji. Similar to the approach used for counting emojis, a Python script was developed to retrieve the Unicode-designated group name associated with each emoji (Unicode, 2025). These group names were subsequently analyzed to classify each emoji as either a face emoji or a non-face emoji. Based on this classification, a binary (dummy) variable was then created to indicate whether a given tweet contains any face emojis. We also included two variables to control the effects of hashtags and mentions, as previous research has found their significant effects on social media engagement (Deng et al., 2021; McShane et al., 2019). Media elements, such as gifs, photos and videos, have been found to increase consumer engagement with brand posts (Deng et al., 2021; McShane et al., 2019). We, thus, included one dummy variable to measure whether a brand tweet includes a gif, photo or video. Previous research has found that brand is a strong influencer of consumer engagement (Leek et al., 2019; Swani et al., 2017). In the current study, our data was collected from 82 brands. We created and included dummy controls for brands. By adding these dummy variables, we inherently controlled for the effects of sectors, unique brand audience profiles and brand follower sizes. Finally, we included dummy controls for tweet dates. Previous research found that social media users engage more with brands during weekdays (Deng et al., 2021; McShane et al., 2019). Therefore, we included a dummy variable to control the effect of posting during weekdays (vs weekends). We also included controls for the tweet posting years and months to minimize other potential influences raised by posting time and historical X (Twitter) updates. While the X (Twitter) application program interface does not allow access to the historical follower size of brands, these dummy variables of posting year and month, together with brand dummy variables, also help mitigate the potential influences of historical brand follower size changes (Deng et al., 2021). The measures and descriptive statistics of variables are reported in Table 1.

To test our hypotheses, we ran a series of multiple linear regressions using engagement as the dependent variable. As the measure of engagement is positively skewed, we followed previous research (McShane et al., 2019; Deng et al., 2021) and used its natural logarithmic transformation, i.e. Ln(Engagement + 1), as our dependent variable. Here, we added 1 to the measures to avoid taking logs of 0. We used IBM SPSS Statistics Version 28 to perform the statistical estimations. We reported the results of the analysis in the next section.

To test the hypotheses, we ran a series of models. We first ran a model using emoji count as an independent variable while including all the control variables. The standardized estimation results are presented in Table 2 (Model 1). The estimation model (Model 1) is significant (F = 656.275 and p < 0.001) and explains the variance of the dependent variable well (R² = 0.511). We then ran a Model 2 where a variable, the emoji count squared, was added to analyze the inverted U-shaped relationship between emoji count and engagement. The estimation model (Model 2) is significant (F = 650.454 and p < 0.001) and explains the variance of the dependent variable well (R² = 0.511). The results of Model 2 showed that the coefficient of emoji count is positive and significant (β = 0.048 and p < 0.001), and the coefficient of emoji count squared is negative and significant (β = −0.024 and p < 0.001), indicating that emoji count has an inverted U-shaped relationship with engagement. Thus, H1 is supported.

To test the remaining hypotheses, we ran another model using a data set consisting only of the brand tweets that contained emoji(s). In this model, we added emoji location and role as additional independent variables. We also included the interaction terms to test the moderating effects of emoji location and emoji role on the relationship between emoji count and engagement. The standardized estimation results are presented in Table 2 (Model 3). The estimation model is significant (F = 240.911 and p < 0.001) and explains the variance of the dependent variable well (R² = 0.657). The results of Model 3 revealed a positive and significant coefficient of emoji count and a negative and significant coefficient of emoji count squared, suggesting an inverted U-shaped relationship between emoji count and engagement, which provides additional support for H1. There was no significant effect of emoji location on engagement (β = 0.016 and p > 0.05), suggesting that emoji location does not directly affect B2B social media engagement. Thus, H2a is not supported. However, interestingly, emoji location moderates the relationship between emoji count and engagement. Specifically, the interaction between emoji location and emoji count is negative and significant (β = −0.107 and p < 0.001), and the interaction between emoji location and emoji count squared is positive and significant (β = 0.079 and p < 0.001). These results suggest that placing emoji(s) within the tweet text (vs outside of the text) weakens the inverted U-shaped relationship between emoji count and engagement (Kang et al., 2021; Zhong et al., 2025). In other words, the inverted U-shaped relationship becomes less pronounced with a flatter curve and a less distinct peak, indicating that adding more emojis has a more limited effect on engagement when emojis are placed inside the text. Thus, H2b is supported. The results of H2a and H2b together suggest that emoji location does not shape engagement directly; instead, it influences engagement indirectly by altering the relationship strength between emoji count and engagement. One possible explanation is that audiences may not be consciously sensitive to emoji location alone unless it interacts with other features (e.g. emoji count).

The results showed a significant effect of emoji role on engagement (β = 0.022 and p < 0.05), suggesting that using emojis to substitute words in brand tweets can lead to higher engagement than using emojis to reinforce words. This result is contrary to the H3a. However, like emoji location, emoji role also moderates the relationship between emoji count and engagement. Specifically, the interaction between emoji role and emoji count is negative and significant (β = −0.045 and p < 0.05), and the interaction between emoji role and emoji count squared is positive and significant (β = 0.05 and p < 0.05). These results suggest that, when emojis are used to replace words in the text, the inverted U-shaped relationship between emoji count and engagement is weakened (Kang et al., 2021; Zhong et al., 2025). Thus, H3b is supported. This finding is interesting, as it suggests that while substitution emojis weaken inverted U-shaped effect of emoji count, they can lead to more engagement than reinforcement emojis. One possible explanation for this can be drawn from the extant literature on emojis and playfulness. In a B2C context using lab experiments, McShane et al. (2021) find that the path from emoji use to engagement is mediated by perceived playfulness. In other words, more playful emoji use enhances engagement. As such, it is possible that, although substitution emojis are likely to reduce fluency, they may also be perceived as more playful and, as such, may strengthen peripheral route processing in such a way that offsets the effects of the reduced fluency on central route processing. Table 3 summarizes the results of the hypothesis testing.

Besides the main effects, our results revealed that many control variables influence social media engagement (Table 3). Of particular note was the result that both the use of emotional words and the readability of brand tweets positively influence social media engagement (i.e. the lower the Dale Chall Score, the higher the readability). This result provides additional support for our general argument that B2B posts can affect social media engagement by influencing both central and peripheral route processing. That is, B2B brand audiences on social media prefer brand tweets that are both emotionally charged and easy to process and comprehend.

In this paper, we examine the effects of emoji count, emoji location and emoji role on B2B social media engagement and find that such variations in emoji use influence B2B brand social media engagement. Our results, yielded through a robust analysis of 64,121 tweets from 82 top B2B brands, show that the relationship between emoji count and engagement follows an inverted U-shaped pattern: engagement increases with moderate emoji use but declines when emoji use becomes excessive. Furthermore, this curvilinear relationship is significantly moderated by both the location and functional role of emojis. Specifically, both placing emoji(s) within the text (vs outside of the text) and using emojis as word substitutions (versus reinforcements) can weaken the effects of emoji count on B2B social media engagement. These findings support the argument that emojis may impact B2B social media engagement via both central route by affecting message clarity, readability and professionalism and peripheral route processing by enhancing emotional tone and visual appeal. Together, these results challenge the assumption that professional communication must avoid informal visual elements, revealing that strategic emoji use, even when unconventional, can be compatible with and even beneficial for B2B brand communication.

Prior research in B2C settings has consistently found that emojis can enhance social media engagement through emotional expression, warmth, positive mood, vividness and visual appeal (Bai et al., 2019; Boutet et al., 2021; Gretry et al., 2017; McShane et al., 2021). In this body of work, emojis are predominantly framed as playful and affective cues that appeal to consumers through peripheral processing (Bai et al., 2019; McShane et al., 2021) with little concern for considerations that become particularly relevant in the B2B context such as message professionalism, expertise, credibility, informational clarity and message comprehension (López-López and Giusti, 2020). In contrast, in the current work, we adopt a B2B perspective to critically examine how emojis might impact social media engagement in this context. This approach highlights the critical importance of considering the dual impact of emojis on both peripheral and central route processing to better understand the role of emoji in B2B social media. Our findings suggest that the dynamics of emoji use in a B2B context operate under more complex dynamics. We find that although emojis may certainly enrich engagement via peripheral route processing, their use must also be counterbalanced to account for their impact on central route processing because of the unique properties of B2B communication. Specifically, our findings indicate that B2B communication demands a balance between emotional relatability and cognitive clarity and that emojis can be strategically used to help achieve that balance.

Our results clearly highlight this need for balance, showing that although emojis can indeed increase engagement in B2B posts, the effect follows an inverted U-shaped pattern, whereby too many emojis can diminish fluency and undermine professionalism, which are particularly critical in a B2B context. This stands in contrast to findings in the B2C literature, showing that more emojis lead to greater engagement (McShane et al., 2021). Furthermore, although substitution emojis are generally considered less effective in B2C research because of ambiguity (Pfeifer et al., 2022), we find that in B2B posts, they may paradoxically increase engagement, possibly because of their perceived creativity or playfulness. These results suggest potentially nuanced trade-offs between affective (e.g. playfulness) and cognitive (e.g. message clarity) considerations that require further tailored investigation. More broadly, these findings underscore the need for context-sensitive frameworks to understand the complex dynamics of visual and affective cues in digital brand communication. It also makes clear that what has been found to work in a B2C context, may not translate seamlessly to B2B.

This research makes several theoretical contributions to the emerging literature on digital communication in B2B contexts. First, it addresses a significant gap regarding emojis’ communicative effects in B2B social media posts by highlighting their impact on engagement. Whereas prior research has largely focused on textual strategies or the mere presence of visual cues (e.g. images and links) (Balaji et al., 2023; Deng et al., 2021), our study shows that, when used strategically, emojis can shape how B2B audiences respond to branded content. This insight advances our understanding of emoji–text multimodality in B2B brand posts and the corresponding impact on engagement. As social media becomes increasingly integral to B2B branding and stakeholder engagement (Cortez et al., 2023; Andersson and Wikström, 2017; Tiwary et al., 2021), these findings offer timely insights into how B2B firms can leverage the linguistic tools (both textual and visual) to enhance message effectiveness (Balaji et al., 2023; Zhang and Du, 2020). Flowing from the previous point, the current work advances theory development on multimodality in B2B social media communication by shifting attention beyond binary cue (e.g. emoji, image and embedded links) presence and absence to consider how the ways in which these cues appear relative to the text may differentially impact engagement. Gu et al. (2023) note a fundamental shift in B2B social media strategy from text-centric to image-centric. Implicit here is the idea that B2B social media messages are now largely multimodal in that they use both textual and visual cues as one semantic entry to communicate a message (Cohn, 2016; Mehmet and Clarke, 2016). The dominance of multimodal social media messages highlights the importance of developing a rich understanding of how visual and textual cues impact social media engagement. However, to date, extant research has primarily considered only the presence or absence of such cues and whether multimodal (e.g. text-plus-image) is more effective than unimodal (e.g. text-only) format (Deng et al., 2021; Gu et al., 2023; Singh et al., 2023). Absent from this literature, however, is consideration for how different configurations of multimodal elements, such as their placement or functional integration, affect message processing and engagement. The current work, thus, contributes to this gap by adopting a multimodality perspective to show that emoji–text interactions (quantity, location and role of emojis) all meaningfully impact B2B social media engagement. These findings, although limited to emoji–text integration in tweets, suggest that future research on multimodal B2B social media communication should move beyond presence/absence dichotomies to account for structural and semantic integration between modes.

Second, this research enhances our theoretical understanding of how emojis influence message processing more generally. By using the B2B communication model to examine emojis’ potential impact on engagement, we consider how emojis likely affect engagement via both peripheral and central route processing. The limited B2B research on emojis has primarily relied on the elaboration likelihood model, solely viewing emojis as emotional cues that facilitate peripheral processing (Balaji et al., 2023; Deng et al., 2021). Even within the extensive body of B2C social media research, significantly more attention has been oriented toward understanding the emotional appeals of emojis (Wang et al., 2023). While acknowledging the positive role of emojis as emotional cues in aiding peripheral processing, this research underscores the involvement of emojis in dual processing pathways. In particular, drawing on recent work on persuasion in digital communications, this research adopts the perspective that peripheral and central route processing can co-occur (SanJosé-Cabezudo et al., 2009). This approach enables us to more fully consider the impact of emoji on processing and, in turn, B2B social media engagement. Specifically, it enables us to consider emojis as not only visual cues that impact peripheral route processing but also emoji–text multimodal cues that impact fluency and professionalism, something typically tied to central route processing (Orazi et al., 2023; Wu et al., 2022). The findings suggest that adopting a co-occurrence perspective on dual process theories may elicit significant insights on B2B social media strategy.

Finally, this research adds to a growing body of research examining the effects of emotional cues in B2B social media strategy. Extant research finds that B2B firms use emotional cues extensively and that emotional cues in B2B social media messages elicit more likes than in B2C posts (Swani et al., 2014; Swani et al., 2017). Emojis, as visual emotional cues, are typically assumed to function independently of textual content (Deng et al., 2021). Our findings challenge this assumption by demonstrating that emoji–text interactions significantly affect B2B social media engagement. This nuanced view suggests that emotional cues (e.g. emojis) and informational cues (e.g. text) in B2B messages do not operate in isolation; instead, their interaction shapes both processing fluency and engagement. This insight adds theoretical depth to the growing body of research on emotion in B2B messaging by moving beyond additive models to consider interactive effects across cue types.

This research also offers several actionable insights for B2B marketers.

First, this research clearly demonstrates that using emojis in B2B social media posts can enhance engagement, challenging the conventional view that emojis are too informal for professional communication. Even in the traditionally formal B2B space, emojis can play a critical role in capturing attention and conveying emotion, thereby eliciting social media engagement. Therefore, B2B brands should not shy away from incorporating emojis into their posts. However, our analysis also reveals that this effect follows an inverted U-shaped curve: while moderate emoji use increases engagement, excessive use can undermine message clarity and appear unprofessional, thus decreasing engagement. Thus, it is important to find the right balance. B2B brands should experiment with different emoji counts to optimize engagement without diluting the message or appearing unprofessional.

Second, while B2B brands are encouraged to leverage emojis in their posts to foster a sense of connection and stimulate engagement, our findings highlight that the positioning of these emojis is crucial. Specifically, placing emojis outside the main body of the text (e.g. at the beginning or end of a tweet) is more effective than embedding them within the text because it facilitates central route processing and message comprehension. This may suggest that, in B2B contexts, emojis are most influential when they serve as attention-grabbing elements or summary highlights rather than being intertwined with written content. B2B marketers should consider this when designing posts to ensure that emojis do not disrupt the flow of the message.

Third, interestingly, contrary to our expectations, emojis used as substitutions for words (rather than as reinforcements) lead to higher engagement. This unexpected finding suggests that B2B audiences may value the creativity and playfulness associated with using emojis as substitutions, possibly because it adds interpretive depth and novelty to professional messaging. However, our findings also highlight that while substitution emojis can be effective, their overuse, especially when combined with a high emoji count, may backfire. These findings suggest a nuanced effect of emoji role. B2B brands can benefit from strategically using substitution emojis to convey complex ideas creatively or to inject a playful tone that resonates with professional audiences. For instance, replacing industry jargon with a reasonable number of relevant emoji(s) that do not jeopardize message clarity may succeed in making posts more creative and approachable while still being contextually relevant. At the same time, B2B marketers should be judicious in their use of substitution emojis, ensuring that they add value and are not overused.

Finally, these insights are increasingly relevant in an era of AI-generated brand communication. As tools like ChatGPT and other large language models become integral to automated content creation, guiding AI systems to apply emojis thoughtfully will be essential. B2B marketers should incorporate these findings into AI prompting workflows, instructing content generators to consider emoji count, location and communicative role to ensure that automated brand posts remain aligned with brand voice and audience expectations. By doing so, B2B organizations can better generate multimodal posts that effectively use emojis to optimize stakeholder engagement.

The current research takes important initial steps toward understanding the communicative effects of emojis on B2B social media engagement. However, several limitations should be noted to inform future research in this area.

First, this study only examines brand posts published on X (formerly Twitter), rather than a diverse array of social media platforms (e.g. Instagram, Facebook and LinkedIn). While this platform offers a text-centered format ideal for examining emoji–text interactions, caution is recommended when generalizing the findings to other platforms. Given the diverse uses, users and formats of different social media platforms, future research is needed to assess whether the found effects hold across different platforms. Besides, it is worth noting that the current study’s design was limited in its capacity to investigate potential sector-level differences in emoji effectiveness. Specifically, by including a dummy variable for each brand, the current study only controlled for potential differences across sectors rather than digging into potential differences in emoji effectiveness across sectors. Future research could address this limitation by examining whether emojis are more or less effective on engagement across different sectors.

Second, the current study focuses on emojis quite generally without differentiating emotional levels or connotations contained by individual emojis. While this approach allows for generalizability across emoji use, it does not account for the diverse emotional and semantic properties of specific emojis. Future research could examine the emotional levels or contents of individual emojis to provide deeper insights into how specific emojis differentially impact social media engagement in B2B contexts. Such analysis would require advanced analytical tools capable of automatically detecting and classifying the emotional meanings of individual emojis (Kaye et al., 2016). Furthermore, the current study does not consider message viewers’ expertise with emojis. Recent research suggests that viewers’ knowledge of emojis can influence their perceptions of processing fluency of messages that use emojis (Wu et al., 2022). Thus, future research can dig deeper into how different types of emojis (common emojis vs non-common emojis) may differentially impact engagement, as it may be easier for emoji novices to interpret the meanings of common emojis than non-common ones.

Third, while this study focuses on the emoji–text interactions, it does not consider how other elements may similarly interact with emojis to impact engagement. For instance, future research examining how emojis interact with other visual elements, such as images or videos, could yield important insights into broader multimodal communication strategies. Additionally, research could consider how the effectiveness of emojis may be moderated by factors external to the message. For instance, there is certain B2C literature indicating that hedonic brands can use emojis for engagement more successfully than utilitarian brands (Das et al., 2019). Building on this, future research could examine how factors such as customers’ perceptions of the B2B brand or the purpose of the message (e.g. information-sharing vs relationship-building: Leek et al., 2019) are more conducive to emoji integration. Similarly, the effectiveness of emojis as an engagement tool in B2B social media may vary depending on the relationship between the message receiver and the B2B organizations. Future research should address these limitations by extending this research to develop a more fulsome understanding of emoji use in B2B social media communications.

Finally, given its cross-sectional nature, the current study is limited in providing empirical clarity on causality between emoji use and B2B social media engagement. However, the temporal sequence in our research context, that is, brand posts are published first, and only then can followers interact with them, is consistent with the logic of “causal ordering,” where the independent variable (post content) precedes the dependent variable (engagement) in time (Finkel, 1995), thus inherently suggesting a causal relationship. Going forward, it would be highly valuable to conduct experimental studies to more rigorously establish causal relationships. Moreover, the current study focuses on observable behavioral outcomes, namely, social media engagement metrics such as likes, comments and retweets, as the dependent variable. While this approach is consistent with prior research and suitable for archival data, it does not capture the underlying psychological processes that may mediate emoji effects, such as viewers’ perceptions of emotional tone, professionalism, clarity or playfulness. Future research could use experimental or survey-based methods to examine specific mediating mechanisms and validate how emoji use influences both message interpretation and engagement behavior in B2B digital communication. Such investigation will further reveal the proposed psychological processes underlying the effects and provide stronger validation of causality.