Shaping fan engagement behaviour through digital platform features: the role of social capital – an experimental study Open Access

In today's digital transformation era, sports industry managers encounter a variety of physical and digital platforms that extend beyond mere content distribution, functioning instead as strategic tools for establishing and sustaining relationships with fans. Such platforms enable fan engagement behaviour (FEB), which encompasses voluntary, non-transactional actions directed toward sport organisations or affiliated actors within a sport organisation's ecosystem (Alexander et al., 2018; McDonald et al., 2022; Stegmann et al., 2023b; Yoshida et al., 2023).

Previous sport management research has identified various forms of FEB, such as word-of-mouth communication, referrals (Asada and Ko, 2016; Yoshida et al., 2014), prosocial behaviours (Thompson et al., 2016), and management cooperation (Biscaia et al., 2016; Huettermann et al., 2019; Uhrich, 2021). Additionally, researchers have concluded that FEB occurs on specific platforms (e.g. sports stadiums and social media; cf. Buser et al., 2022; Uhrich, 2014). However, McDonald et al. (2022, p. 296) highlight a significant gap, “engagement is a context-dependent construct, and how and why different forms of engagement manifest across these interdependent and often nested touchpoints are largely unknown”. Consequently, there is limited understanding of the reasons behind the occurrence of specific FEB types on certain engagement platforms. Furthermore, although scholars widely accept the multidimensional nature of FEB (Jaakkola and Alexander, 2014; Stegmann et al., 2024; Yoshida et al., 2023), most sport management studies still rely on unidimensional measures (Gordon et al., 2025; Yoshida et al., 2023), limiting comprehensive insights into this phenomenon.

Fan engagement theories suggest that FEB arises as fans draw on resources available through social connections on engagement platforms (Buser et al., 2022; Woratschek et al., 2014). In this context, social capital theory offers a potential perspective, positing that individual actions and resource integration depend on social interactions and network structures (Coleman, 1988). Social capital typically differentiates between bridging social capital (resources accessed through heterogeneous, weak ties) and bonding social capital (resources from homogeneous, strong ties; Bourdieu, 1986; Putnam, 2000). Consequently, access to different social resources on a platform may explain the manifestation of specific FEB types. For example, limited access to intense social interactions on a platform could reduce cooperative and helpful engagement behaviours (Stegmann et al., 2024; Uhrich, 2021).

Additionally, Storbacka et al. (2016) emphasise that platform owners possess significant agency in designing features, potentially influencing the types of resources accessible to users. Thus, platform owners (e.g. sport club managers) might be able to influence and assert control over how or which forms of resources individuals can access using a specific platform. Therefore, they may have the power to intentionally design platforms to foster bridging or bonding social capital, potentially affecting FEB. For instance, platform features such as open social feeds, which encourage bridging social capital, might enhance FEBs through word-of-mouth, connecting diverse fans (Asada and Ko, 2016). Conversely, other forms (e.g., private chat features) facilitating bonding social capital could foster more intimate engagement behaviours, such as providing service feedback or peer support (Biscaia et al., 2016).

Despite considerable research on fan engagement, the role of platform design on social capital access and FEB remains unexplored. However, understanding this relationship would be critical for sport management practitioners seeking to design digital platforms that stimulate desired fan behaviours. Thus, this study aims to investigate how specific platform features selectively grant access to social capital forms, thereby shaping FEB.

To address these gaps, this research employs a scenario-based experimental design manipulating digital platform features. Specifically, this study examines the impact on four distinct FEB dimensions using a 2 (bridging social capital) × 2 (bonding social capital) factorial between-subject design while controlling key covariates. By doing so, the study contributes to the fan engagement literature by (1) providing explanatory approaches why certain FEB forms emerge on specific digital platforms, (2) identifying the role of sport club managers on fan engagement, and (3) differentiating the role of platform features in influencing specific FEB forms. The findings offer actionable insights for sports managers to enhance platform designs that cultivate targeted engagement behaviours.

Engagement behaviour is understood as customers' resource integration that goes beyond transactions (McDonald et al., 2022). It can be conceptualised as a psychological state of cognitive, emotional, and behavioural responses (Brodie et al., 2011), a disposition to engage (Storbacka et al., 2016), or as voluntary, extra-role behaviours (van Doorn et al., 2010). Yoshida et al. (2023, p. 3) define FEB as “[fans'] voluntary contribution to the success and welfare of a sport team through value-adding behaviours, going beyond the mere consumption of sport products such as ticket purchase and television viewing.”

FEB is recognised as a multidimensional construct in sport management (McDonald et al., 2022). Yoshida et al. (2014) initially proposed a three-dimensional model—management cooperation, prosocial behaviour, and performance tolerance—which was later expanded into a six-dimensional model including fan learning, resource integration, ritualistic behaviour, flow experience, management cooperation, and fan knowledge feedback considering it a higher-order construct with multiple sub-dimensions (Yoshida et al., 2023). These dimensions yield varying outcomes; for instance, management cooperation is associated with purchase intention, whereas prosocial behaviours are not (Yoshida et al., 2014). McDonald et al. (2022) further advanced this work by developing a broader typology of FEB behaviours, including blogging, tailgating, word-of-mouth, and basking in reflected glory. Digital FEB is often defined as reading, posting, liking, and commenting, or as consuming, contributing and creating (Cao and Matsuoka, 2024; Lee and Na, 2024).

Others adopted a more straightforward, unidimensional perspective (Gordon et al., 2025). As McDonald et al. (2022) articulate, acknowledging FEB's inherent diversity through a multidimensional model provides a superior analytical lens for capturing its complexity and outcomes, as it prevents the loss of critical information and enables the consideration of contradictions among sub-dimensions. However, current models often conflate behavioural, cognitive, and emotional elements, exclusively focus on online fan engagement, do not consider prosocial behaviours or insufficiently differ between fan engagement and value co-creation (McDonald et al., 2022; Yoshida et al., 2023).

To address these criticisms, we adopt a model using a value co-creation lens (Conduit and Chen, 2017; Woratschek et al., 2014), which considers engagement behaviours as voluntary, non-transactional forms of value co-creation. Jaakkola and Alexander (2014) propose four distinct types of engagement behaviour: augmenting, co-developing, influencing, and mobilising behaviours. In a sport sense, augmenting FEB includes creative contributions such as chanting and posting content (Hedlund et al., 2018; Stieler et al., 2017). Co-developing FEB involves collaborative efforts such as providing feedback and management cooperation (Huettermann et al., 2019; Stegmann et al., 2023a). Influencing FEB covers word-of-mouth and recommendations (Asada and Ko, 2016), while mobilising FEB refers to activities such as rallying support, defending the club, organising events or showing visible solidarity (Bradford and Sherry, 2015; Ströbel et al., 2021).

While non-transactional in nature, these behaviours occur within broader consumer relationships. Like customers in commercial markets, fans demonstrate brand- and community-related behaviours that reinforce transactional consumption. FEB thus complements sport consumption. This view aligns with studies linking it to behavioural loyalty, purchase intention, and attendance (Fehrer et al., 2018; Yoshida et al., 2023). Accordingly, we conceptualise FEB as participatory behaviour that is related to but transcends transactional sport consumption.

FEB unfolds embedded in layered social contexts that shape fans' ability and motivation to integrate resources (Buser et al., 2022). Socio-cultural settings, platform features, and interpersonal dynamics encourage or constrain the scope and expression of FEB (Alexander et al., 2018; McDonald et al., 2022). In sport ecosystems, fans, sponsors, and organisers co-create value across physical and digital platforms (Horbel et al., 2016; Grohs et al., 2020). In this view, platforms provided by sports clubs or event organisers actively facilitate resource integration.

Digital environments have transformed FEB by facilitating synchronous and asynchronous engagement across geographical boundaries (Fenton et al., 2023; Santos et al., 2019; Stegmann et al., 2023b; Uhrich et al., 2024). Additionally, FEB is shaped both by internal factors (e.g. individual motivation, prior experiences) and external enablers (e.g. platform design and social dynamics;, cf. Storbacka et al., 2016). Strong emotional attachments and team identity may drive many fans (Mastromartino et al., 2019; Yoshida et al., 2021), however, FEB might not manifest if the surrounding context fails to support or stimulate such actions (Storbacka et al., 2016). Engagement platforms thus play an active role in translating motivation into FEB. Accessible platforms with explicit norms and interactive features can convert latent dispositions into actual FEB (Storbacka et al., 2016; Buser et al., 2022). Sport clubs, as platform operators, can orchestrate FEB through deliberate platform design, enabling or constraining engagement by adjusting platform features, establishing norms, or defining boundaries (Blasco-Arcas et al., 2020). Recent qualitative research highlights how digital platforms influence fan interactions and enhance access to resources necessary for engagement (Fenton et al., 2023).

As described in (sport) sociology and theories of value co-creation, resources can be accessed and integrated through social structures within social networks (Buser et al., 2022; Woratschek et al., 2014). In seminal sociological works, the idea of embedded resources in social networks was subsumed under the notion of social capital (Bourdieu, 1986; Putnam, 2000). Social capital refers to resources embedded within durable, institutionalised networks of relationships that facilitate mutual benefit through social interactions (Bourdieu, 1986; Putnam, 2000). Social capital is typically categorised into two types: bridging and bonding. Bridging social capital arises from heterogeneous, weak ties that promote information exchange and diverse interactions (Putnam, 2000) while bonding social capital stems from close-knit networks that offer emotional support and reinforce shared identities (Putnam, 2000; Skinner et al., 2008).

Within sport contexts, social capital shapes fan interactions. Bridging social capital helps spread information broadly, enabling word-of-mouth, referrals, and expanding fan networks, thus reinforcing fan identification and enhancing overall consumer happiness (Mastromartino et al., 2022; Yoshida et al., 2021). Bonding social capital, developed within tight fan communities, strengthens emotional bonds and team loyalty, leading to higher fan satisfaction and improved brand equity (Biscaia et al., 2016; Mastromartino et al., 2019). Digital platforms further amplify these dynamics by providing new avenues for resource exchange and interaction (Fenton et al., 2023).

Summarising, FEB is a function of resource integration enabled by access to bridging and bonding social capital, which can be facilitated by (digital) engagement platforms as social contexts within sport ecosystems. By creating environments that encourage information sharing, collaboration, and shared identities, platform operators can influence how fans mobilise resources for FEB (Blasco-Arcas et al., 2020; Storbacka et al., 2016). This interplay among individual behaviours, social contexts, and resources forms the theoretical foundation for advancing research on FEB.

Fans need to access, mobilise, and integrate social capital to pursue FEB. Digital platform features significantly shape the social capital accessible to fans (Blasco-Arcas et al., 2020). Bridging social capital emerges from platform features connecting fans with diverse networks while bonding social capital arises from features fostering closer and more intense interactions (Coleman, 1988; Granovetter, 1973; Putnam, 2000). Prior research shows that increased social capital is associated with increased engagement behaviours (Collins and Heere, 2018; Mastromartino et al., 2019), highlighting the role of community membership and interactions in motivating fan activities. Our interest lies in investigating which platform features, enabling bridging and bonding social capital, differentially influence FEB's four dimensions—augmenting, co-developing, influencing, and mobilising—each capturing distinct facets of fan engagement.

Bridging social capital provides access to diverse, loosely connected networks facilitating information dissemination (Phua, 2012). “Strength in weak ties” emphasises that loosely connected networks efficiently distribute new information (Granovetter, 1973). Bridging social capital is relevant in integrating external individuals into fan communities and disseminating innovative ideas and content (Collins and Heere, 2018; Walseth, 2008). Considering FEB's multidimensional nature, bridging social capital particularly enhances behaviours benefiting from expansive networks. For instance, augmenting behaviours—such as creative content sharing, social media posts, and stadium chants (Hedlund et al., 2018; Stieler et al., 2017)—thrive on broad connectivity and rapid information flow. Similarly, influencing behaviours like word-of-mouth and referrals strongly depend on widespread network interactions (Asada and Ko, 2016; Phua, 2012). Conversely, co-developing and mobilising behaviours demand stronger, trust-based interactions characteristic of bonding social capital, suggesting bridging capital alone may be insufficient to stimulate these behaviours (Fenton et al., 2023).

Digital platform features facilitating broad connectivity—such as open social feeds and friendship requests—strengthen access to bridging social capital by facilitating diverse exchanges. Such platforms continually expose fans to varied content, which may stimulate augmenting behaviours like content creation and sharing. Additionally, influencing behaviours could benefit as fans broadly disseminate recommendations and spread word-of-mouth messages across extensive networks (Collins and Heere, 2018; Mastromartino and Zhang, 2020). However, these loosely connected interactions likely offer insufficient depth and trust needed for co-developing and mobilising behaviours. Thus, we hypothesise:

Second, bonding social capital arises from close relationships that emerge from dense and reciprocal interactions (Collins and Heere, 2018). Such close, stable ties often enhance feelings of familial belonging—often described as fans considering their sports club community “like family” (Fenton et al., 2023) and significantly influencing emotional and functional community involvement (Skinner et al., 2008). Bonding social capital could encourage engagement behaviours demanding strong interpersonal trust, such as co-developing behaviours (feedback provision, collaborative decision-making; cf. Huettermann et al., 2019; Stegmann et al., 2023a) and mobilising actions (collective support or public brand advocacy; cf. Bradford and Sherry, 2015; Thompson et al., 2016). In contrast, these strong relational ties are less likely to stimulate more superficial activities such as augmenting or influencing behaviours, which tend to depend on broader, less intimate networks.

Digital platforms fostering bonding social capital provide features supporting intimate, personalised interactions—such as private messaging, exclusive forums, and dedicated fan spaces. These features create environments where fans can reciprocally interact and feel closely connected, enhancing their emotional investment and commitment (Collins and Heere, 2018; Putnam, 2000). Such digital environments could facilitate co-developing and mobilising behaviours such as providing constructive feedback or contributing to joint initiatives and mobilising actions. Conversely, these features, such as private chats, may have a limited impact on engagement activities that depend on broader, less intensive interactions, such as posting creative content or influencing broader audiences. Consequently, we hypothesise:

Finally, we anticipate interactive effects between access to bridging and bonding social capital. Literature suggests that these forms of social capital function synergistically rather than independently (Tacon, 2012). Studies indicate that the benefits derived from one form can reinforce the other (Collins and Heere, 2018; Putnam, 2000). The engagement process often triggers a virtuous cycle, where fans utilising social capital further strengthen their community ties, increasing their willingness for continued engagement (Mastromartino and Zhang, 2020).

Platform features can play a crucial role in facilitating this synergy. Combining features promoting broad connectivity (e.g. social feeds, friendship requests) with intimate interactions (e.g. private messaging, personalised offerings), a digital platform is expected to enhance all FEB dimensions by allowing expansive information exchange alongside trust-building interactions. Such designs enable the simultaneous spread of novel information and the development of deeper relational bonds, amplifying augmenting, co-developing, influencing, and mobilising behaviours. Thus, we propose that the combination of access to bridging and bonding through platform features produces a synergistic and intensified enhancement of FEB:

Following the recommendations for experimental research in sport management (O'Reilly, 2011), in February 2022, we conducted an online scenario-based experiment via Qualtrics using a 2 × 2 between-subject factorial design to examine how digital platform features influence perceptions of bridging and bonding social capital and, in turn, affect the four dimensions of FEB. Participants first selected their favourite professional tier-1 sport club from five major sports (American football, ice hockey, soccer/football, basketball, baseball), with the option to manually enter their choice if not listed. This personalisation aimed to increase the ecological validity of the scenario by increasing connection to the fictitious platform and associated brand community. Participants were then randomly assigned to one of four experimental conditions, each presenting distinct platforms varying in features related to facilitating access to bridging and/or bonding social capital (Appendix A, Table A1). After exposure to the scenario, participants completed measures for confounding variables, manipulation checks, dependent variables, and sociodemographic questions. Embedded attention checks ensured response quality. The ethics committee of the Faculty of Human Sciences (University of Bern) has assessed the study as ethically unobjectionable (Nr. 2021-06-00006).

The experimental stimulus comprised a detailed description of a fictitious digital engagement platform for a sport club (Appendix A). Platform features were systematically manipulated along bridging and bonding social capital. The stimulus was developed within our research group and validated by independent sport management and sociology scholars.

To operationalise bridging social capital, we varied the diversity of actor groups, the diversity of posts in the newsfeed, and the openness of friendship connections. In the high bridging condition, participants were given access to fans, players, coaching staff, and sponsors, a newsfeed featuring posts from each group, with friendship requests open to any member. In contrast, the low bridging condition restricted access solely to fans, offered a newsfeed limited to posts from the club and fellow fans, and confined friendship requests to fans. This manipulation aligns with theoretical definitions of bridging capital as involving weak, diverse, outward-reaching ties across heterogeneous social groups (Putnam, 2000; Granovetter, 1973). Enabling interaction with diverse stakeholders simulates such weak ties, while restricting engagement to fan–fan interactions represent a more inward-facing network.

Bonding social capital was manipulated by altering the depth, intensity and reciprocity of interactions. In the high bonding condition, the platform included direct chat, personalised content (e.g. behind-the-scenes insights, player videos), customised offers, and a participatory, interactive fan feedback tool (“Wall of Ideas”). The low bonding condition removed or replaced these with impersonal alternatives, such as generic contact forms and non-tailored content. This reflects the conceptualisation of bonding capital as comprising emotionally close, trust-based ties developed through frequent, reciprocal interactions within homogeneous groups (Putnam, 2000). Platform features fostering personalised and interactive exchanges were used to simulate the digital affordances that facilitate strong-tie relationships. In contrast, generic, one-way tools diminished opportunities for trust and intimacy. The manipulation thus focused on the quality and emotional resonance of interaction, not simply its volume or frequency.

All constructs were measured using 7-point Likert scales, with detailed scales, items, and reliability statistics provided in Appendix B (Table B1). The dependent variable was operationalised across four dimensions—augmenting, co-developing, influencing, and mobilising—using a 16-item scale adapted from Roy et al. (2018). Originally developed for the hotel industry, items were tailored to the sport club context. For example, “I post photographs of my stay at this hotel” was adapted to “I would post photographs of my experiences with my favourite club on this platform.” The adapted scale comprised 4 items for augmenting, 3 for co-developing, 3 for influencing, and 6 for mobilising FEB. Roy et al.’s (2018) scale was chosen for its grounding in value co-creation theory and its emphasis on voluntary, non-transactional behaviours—conceptually analogous to FEB in sport settings (cf. McDonald et al., 2022).

To account for potential confounds, we included several covariates. First, team identification—reflecting fans' psychological attachment and sense of community membership with their chosen sport club known to influence FEB (Yoshida et al., 2023)—was measured using the scale developed by Woratschek et al. (2020). Second, engagement disposition, which captures an individual's tendency to engage, was included based on prior research in researching FEB in sport (Schönberner and Woratschek, 2023) and we adapted Fehrer et al.’s (2018) scale to sport. Third, behavioural loyalty, indicative of past or future consumption behaviours that signal strong brand commitment (Bauer et al., 2008), was measured to control for its potential influence on FEB. Although the role of loyalty as an antecedent or consequence of FEB is debated (Brodie et al., 2011; Cao and Matsuoka, 2024), recent studies suggest that loyalty may enhance engagement (Fehrer et al., 2018; Yoshida et al., 2021). Thus, we used Bauer et al.’s (2008) scale on past behavioural loyalty. Finally, trust in digital technologies was included due to its relevance in digital settings. Higher trust levels typically predict greater engagement (Yin et al., 2023) and were measured using a four-item scale adapted from Slade et al. (2015).

Before data collection, we conducted a pretest with 28 participants to assess whether the manipulation checks succeeded. The results show successful manipulation checks for both access to bridging social capital (t(26) = −3.08, p < 0.001) and access to bonding social capital (t(26) = −2.32, p < 0.05). In the main study, we initially recruited 393 participants through sport-related platforms and university networks, with no compensation provided. Data cleaning procedures involved excluding incomplete responses (n = 77), surveys completed in less than 150 s (n = 12), excessively long responses (>10,000 s; n = 4), and instances of straight-lining (where respondents provided zero variance in their responses for covariates (n = 2) or dependent variables (n = 13). In addition, 59 participants who failed attention checks—by endorsing non-existent platform features such as the option to order stadium food or reserve parking spaces—were removed. The final sample comprised N = 226 (94 male, 129 female, 3 non-binary; age: M = 27.03, SD = 7.78), somewhat below our a priori power calculation recommending N = 279. This limitation in power will be considered when interpreting the results. Nevertheless, participants were evenly distributed across the four experimental conditions (see Table 1), and subsequent ANOVA tests confirmed that randomisation was successful, with no significant differences emerging in terms of demographics or covariates (team identification, behavioural loyalty, engagement disposition, and trust in technology), as detailed in Appendix C (Tables C1 to C4).

Next, we applied confirmatory factor analysis using “lavaan” in R to evaluate the theoretical four-factor structure of FEB. The analysis yields a marginally acceptable fit (χ²(103) = 276.00, p < 0.001; CFI = 0.92; TLI = 0.90; RMSEA = 0.086; SRMR = 0.06). Composite reliabilities for each FEB dimension ranged from 0.74 to 0.89, and most average variances extracted exceeded 0.50 (with only augmenting slightly below at 0.44), which can be considered acceptable (Fornell and Larcker, 1981). Reliability analyses with McDonald’s (1999) showed robust internal consistency (ωh between 0.90–0.95) and acceptable item loadings predominantly between 0.50–0.80, indicating coherent latent constructs (for details, see Appendix B, Table B1). Further assumptions for subsequent MANCOVA were tested: linearity between covariates and dependent variables was assessed through scatterplots, Box's M test (Box's M = 41.072, p = 0.110) indicated homogeneity of covariance matrices across groups, and Levene's tests supported equality of variances. Multicollinearity diagnostics revealed correlations below 0.80, variance inflation factor values below 3.31, and tolerances above 0.30, all within acceptable thresholds (Appendix C, Tables C5 to C8).

Finally, manipulation checks confirmed successful experimental conditions evoking differential perceptions of social capital. Participants in the high bridging condition reported significantly higher perceived bridging social capital (M = 4.98, SD = 1.56) compared to the low bridging condition (M = 3.50, SD = 1.97; t(224) = 6.26, p <. 001). Similarly, perceived bonding social capital was significantly higher in the high bonding condition (M = 4.64, SD = 1.45) versus low bonding (M = 3.81, SD = 1.67; t(224) = 3.98, p < 0.001). These results demonstrate that the platform's bridging and bonding features successfully influenced participant perceptions as intended.

We conducted a MANCOVA with bridging (low vs. high) and bonding social capital (low vs. high) as between-subject factors, controlling for four covariates: team identification, behavioural loyalty, engagement disposition and trust in technology. The four dependent variables captured the facets of FEB: augmenting, co-developing, influencing, and mobilising. The multivariate tests (Wilks Lambda) showed no significant overall effect of bridging social capital (F(4, 215) = 1.920, p = 0.108, η² = 0.034), bonding social capital (F(4, 215) = 1.174, p = 0.323, η² = 0.021) and the interaction effect (F(4, 215) = 1.243, p = 0.294, η² = 0.023). Engagement disposition as a covariate was significant at the multivariate level (F(4, 215) = 2.758, p = 0.029, η² = 0.049), suggesting that it explains variance across all FEB dimensions.

Despite the non-significant multivariate results for bridging and bonding, follow-up univariate analyses revealed distinct patterns for each FEB dimension (see Table 2). Regarding access to bridging social capital through respective platform features, there are significant effects on augmenting FEB (F(1, 218) = 5.542, p = 0.019, η² = 0.025), co-developing FEB (F(1, 218) = 5.134, p = 0.024, η² = 0.023) and influencing FEB (F(1, 218) = 5.340, p = 0.022, η² = 0.024). These platform features, however, do not significantly influence mobilising FEB (F(1, 218) = 3.282, p = 0.071, η² = 0.015). Accordingly, bridging social capital had a significant positive effect on three of the four dimensions of FEB (augmenting, co-developing, and influencing).

Regarding access to bonding social capital through platform features, there only mobilising FEB (F(1, 218) = 4.045, p = 0.046, η² = 0.018) is significantly influenced by high levels of bonding social capital. Contrastingly, augmenting FEB (F(1, 218) = 1.637, p = 0.202, η² = 0.007), co-developing FEB (F(1, 218) = 2.615, p = 0.107, η² = 0.012) and influencing FEB (F(1, 218) = 0.969, p = 0.326, η² = 0.004) are not significantly influenced. Thus, only one dimension of FEB is significantly and positively influenced through access to high bonding social capital through platform features of the digital platforms we investigated.

Moreover, the interaction effect of access to high levels of bridging and bonding social capital is significant concerning mobilising FEB (F(1, 218) = 4.083, p = 0.045, η² = 0.018). The effect shows that mobilising FEB is lower in the high bridging/high bonding experimental condition (M = 4.13, SD = 1.42) than in the low bridging/high bonding (M = 4.25, SD = 1.35) or high bridging/low bonding (M = 4.24, SD = 1.24) conditions. There are no significant effects regarding augmenting FEB (F(1, 218) = 0.714, p = 0.399, η² = 0.003), co-developing FEB (F(1, 218) = 0.509, p = 0.476, η² = 0.002) and influencing FEB (F(1, 218) = 1.228, p = 0.269, η² = 0.006).

Finally, several covariates, particularly engagement disposition, also explained variance in the four dependent variables. For instance, engagement disposition was a significant predictor for augmenting FEB (F(1, 218) = 7.891, p = 0.005, η² = 0.035), influencing FEB (F(1,218) = 4.090, p = 0.044, η² = 0.018), and mobilising FEB (F(1, 218) = 9.572, p = 0.002, η² = 0.042).

This study advances conceptual understanding of how digital platform features, by facilitating access to bridging and bonding social capital, shape four distinct forms of FEB: augmenting, co-developing, influencing, and mobilising. In doing so, it responds to recent calls for more detailed insights into FEB mechanisms in sport (McDonald et al., 2022) and complements theoretical propositions regarding the role of social capital in driving FEB.

While the central finding, that more personalised and inclusive content enhances engagement, may appear intuitive, its theoretical significance lies in how engagement emerges from the underlying structure of social interactions. Seemingly routine actions such as posting, liking, or messaging are digital traces of deeper processes: network expansion, trust formation, and identity reinforcement. From a social capital perspective, “better” content implies that it is structurally more effective in cultivating communities rather than merely being more entertaining. By linking FEB to patterns of bridging and bonding social capital, this study provides explanatory depth to otherwise familiar platform functionalities.

However, these findings must be interpreted with caution given the study's design and statistical limitations. Most notably, the sample size fell short of the a priori power threshold (N = 226 vs. required N = 279), likely limiting sensitivity to detect subtle effects, especially in interactions. The MANCOVA did not yield significant multivariate results for bridging, bonding, or their interaction. However, exploratory univariate analyses revealed theoretically aligned patterns across FEB dimensions. These results should thus be regarded as preliminary indicators rather than confirmatory evidence.

First, bridging social capital was associated with higher levels of three FEB types: augmenting (e.g. posting or blogging; Hedlund et al., 2018), co-developing (Stegmann et al., 2023a), and influencing (e.g. positive word-of-mouth; Asada and Ko, 2016). Platform features that enable open friendship connections, diverse newsfeeds, and content sharing may support these engagement forms by broadening fans' exposure to diverse information and perspectives (Collins and Heere, 2018). Interestingly, access to bridging social capital also enhanced co-developing behaviours despite the assumption that collaboration requires closer ties. This suggests that weak, heterogeneous connections may stimulate ideation and participatory behaviour by introducing novel viewpoints (Phua, 2012). As expected, bridging capital was not associated with mobilising FEB, which typically depends on deep trust and group cohesion (Thompson et al., 2016).

Second, bonding social capital was positively associated with mobilising FEB, reinforcing the idea that close-knit, trust-based relationships facilitate collaborative actions (Putnam, 2000; Thompson et al., 2016). Features such as private messaging, tailored content, and feedback mechanisms likely foster the intimacy needed for fan activism or volunteering (Bradford and Sherry, 2015). In contrast, bonding capital showed no significant association with augmenting, co-developing, or influencing behaviours, which appear more reliant on the breadth than the depth of social ties.

Third, and unexpectedly, mobilising FEB decreased when both bridging and bonding capital were simultaneously high. Although a synergistic effect was hypothesised, this interaction suggests potential interference or tension between social capital types. Possible explanations include cognitive overload, role conflict, or loyalty fragmentation (Maier et al., 2015). Rather than reinforcing one another, the concurrent activation of both social capital types may, in some cases, inhibit FEB. While this finding contradicts initial assumptions, it aligns with McDonald et al.’s (2022) view that more FEB is not always better.

In sum, despite the absence of significant multivariate effects, the univariate results, although based on small effect sizes (η² = 0.018–0.035), offer an exploratory map of how specific social capital structures could relate to particular FEB types. Given the study's exploratory nature, these findings should be interpreted cautiously. Still, they support the view that platform features facilitating different types of social capital can serve different FEB types. Bridging capital appears particularly relevant for augmenting, co-developing, and influencing behaviours, whereas bonding capital supports mobilising FEB.

In addition, the unexpected interaction effect adds a layer of complexity to understanding the mechanisms of when and how digital platform features foster FEB. Overall, our findings still reinforce FEB's context-dependency (Grohs et al., 2020; Horbel et al., 2016) and highlight platform owners' crucial role in orchestrating engagement (Blasco-Arcas et al., 2020; Storbacka et al., 2016), which extends qualitative findings (e.g. Fenton et al., 2023). They also reinforce the need to view FEB as a multidimensional construct, with platform features affecting each engagement type differently (McDonald et al., 2022).

This study provides several exploratory but actionable insights for sport club managers aiming to facilitate distinct types of FEB via digital platforms. These implications, however, should be interpreted cautiously given the exploratory nature of the study and the small effect sizes.

First, managers may consider aligning platform functionalities with specific engagement goals. For instance, to encourage content creation (augmenting FEB) or peer-to-peer referrals (influencing FEB), clubs could implement features that promote bridging social capital, such as open social feeds, hashtag-based content aggregation, and public comment functions. Social media platforms, which provide fans with access to interactive feeds, exemplify this approach. To support co-developing FEB, collaborative practices such as idea co-creation or feedback, clubs might integrate participatory mechanisms across stakeholder interfaces, such as appointing supporter liaison officers or using platforms like Discord or fan tokens to collect community input. In contrast, bonding-oriented features such as private messaging, closed groups, and personalised push notifications are more effective for mobilising FEB (e.g. fan activism or volunteering). Clubs may consider creating closed groups with key fan leaders to regularly connect with them and foster mobilising FEB among these key fans.

Second, clubs should avoid overloading platforms with both bonding and bridging social capital features simultaneously. Our findings suggest that co-activation may inhibit mobilising FEB, possibly due to conflicting demands on fans' attention. A segmented strategy, designing distinct spaces or user segments for different engagement types, may prove more effective. For example, a general club app could facilitate broader (bridging-based) FEB, while exclusive “inner circle” communities hosted on platforms like Discord could strengthen bonding-based FEB.

Third, clubs should establish FEB-specific performance metrics to evaluate FEB outcomes. Useful indicators might include the number of user-generated posts, engagement ratios (likes/comments per post), referral conversions for fan club sign-ups, participation in feedback tools, idea submissions, poll responses, volunteer enrolments, and repeated engagement levels.

Fourth, individual fan dispositions should be considered. Factors such as loyalty, trust in digital tools, and willingness to engage shape the effectiveness of platform features. Clubs should clearly communicate privacy policies, security assurances, and offer personalisation controls to build digital trust. Loyal and digitally engaged fans could also be recruited as peer advocates or brand ambassadors, amplifying engagement through social proof.

This study has several limitations that offer avenues for future research. First, while our scenario-based experiment provided internal validity, it may not capture the complexity of real-world digital fan engagement, which is shaped by dynamic, multi-platform use, live match contexts, and social interactions. Future research should replicate our findings in naturalistic, longitudinal, or multi-platform settings, particularly during live events, to enhance generalisability.

Second, our reliance on self-reported behavioural intentions introduces potential biases, such as social desirability and discrepancies between intention and behaviour (Rhodes and Dickau, 2012). Incorporating objective behavioural data in future research, such as posting frequency or referrals, could improve external validity.

Third, our sample fell short of the planned power threshold (N = 226 vs. 279), which potentially limited the ability to detect subtle or interaction effects. Replication with larger, more diverse samples is needed to assess the robustness of our results.

Fourth, the observed negative interaction between bridging and bonding social capital for mobilising FEB suggests that combinations of social capital structures may not always be synergistic. Future studies could apply mixed-method approaches, including interviews, netnography, or network analysis, to explore how digital feature configurations interact with user dispositions.

Fifth, while our study provides exploratory insights into the platform features that align with different FEB types, future research should continue to examine causal effects of specific features (e.g. gamification, personalisation, second-screen, AR/VR). Analysing moderators such as fan culture or club norms may shape when and for whom these affordances are effective.

Finally, this study focused on positive FEB. Future research should also consider negative engagement forms (e.g. brand sabotage; Stieler et al., 2017) and empirically validate FEB metrics (e.g. referrals, loyalty participation) in real-world settings, linking them to transactional behaviours such as attendance or purchase intentions.

The supplementary material for this article can be found online.