Continued engagement intention with social media influencers: the role of experience Open Access

Social media influencers (SMIs) are widely recognized as online opinion leaders (Geng et al., 2024). However, their potential role in enhancing brand performance remains under-researched (Sánchez-Fernández and Jiménez-Castillo, 2021). The global value of SMI marketing is estimated at US$1.64 billion and is expected to grow in the coming years (Statista, 2023). Brands are increasingly turning to SMIs to engage consumers (Aw et al., 2022) by posting branded content, opinions, recommendations, and experiences to drive post-engagement through views, likes, comments, shares, visibility, virality, and persuasion (Sheng et al., 2024). For instance, PewDiePie garnered over 322 million views on YouTube in a single month (Rohde and Mau, 2021). Social media creates opportunities for interactions among SMIs, brands, and consumers for entertainment, hedonic experiences, and economic rewards (Jahn and Kunz, 2012). However, creating and maintaining a long-term engaging experience is challenging for many brands (Giakoumaki and Krepapa, 2020). Research on technology acceptance suggests that understanding users’ engagement behavior is more important than their adoption because it exemplifies long-term use beyond initial adoption (Kim et al., 2013). Thus, brands need to create continued engagement between users and SMI by enhancing their experiences to ensure the long-term usage of their products beyond the initial adoption.

Consumer engagement (CE) refers to the interaction between a consumer as an engagement subject and a focal engagement object (Hollebeek et al., 2014; Hollebeek and Macky, 2019), such as SMI. The cognitive and emotional absorption experienced by consumers through CE with a focal object such as SMIs (Brodie et al., 2013) is anticipated to influence outcomes such as ongoing search behavior and repurchase intentions (Cheung et al., 2021), word-of-mouth (WOM) (De Oliveira Santini et al., 2020; Wang and Lee, 2020), sharing intention, and recommendations. Recognizing the interactive nature of CE has become more crucial (Sheng et al., 2024), especially when users experience fatigue when repeatedly exposed to substantial volumes of information SMIs and brands (Bright et al., 2015), often opting to disengage when overwhelmed by excessive content in their social media feeds (Bright and Logan, 2018). Typically, 51% of users scroll past posts from SMIs in their feeds (Carter, 2023). Instagram engagement rates have decreased from 1.67% in 2020 to 1.18% in 2022 (Amoyo, 2022). Thus, there is a distinct lack of understanding of what drives consumers’ continued engagement with SMIs (Dolan et al., 2019; Pezzuti et al., 2021), and few studies have investigated the drivers that affect continued engagement activities (Vander Shee et al., 2020). Moreover, as firms increasingly leverage influencers for promotional activities, it is imperative to comprehend the dynamics of CE with SMIs (Pradhan et al., 2023). However, the results of extant research provide contrasting positive, negative, insignificant, and significant findings between CE and its outcomes (Beckers et al., 2018), emphasizing a lack of consistency in the results (De Oliveira Santini et al., 2020). For example, some studies indicate a strong relationship between CE and WOM (Halaszovich and Nel, 2017), whereas other studies observe a weak relationship (Badrinarayanan et al., 2015). Research has been inconsistent regarding the relationship between content-related antecedents (e.g. hedonic and informative content) and CE behavior (Kefi and Maar, 2020; Kulikovskaja et al., 2023; Hollebeek et al., 2014). To address this knowledge deficit, we contribute to the CE literature in this study by developing and testing a conceptual model that explores and explains the relationships between the factors that influence continued engagement intention (CEI), a form of behavioral intention (Hepola et al., 2020) in the SMI context. Various theories have examined continued engagement intentions, such as the theory of planned behavior (Madden et al., 1992) and the unified theory of acceptance and use of technology (Venkatesh et al., 2003).

In this study, we adopted a theory-integration approach by integrating the flow theory and social exchange theory to advance the understanding of CEI in the SMI context. Furthermore, several studies have focused on homophily attitudes in the context of social media. For example, Ladhari et al. (2020) focused on the three dimensions of homophily (i.e. attitude, values, and appearance) of SMIs. In the luxury fashion context, Lee and Watkins (2016) revealed a strong positive effect of parasocial interaction determined by attitude on luxury brand perceptions and purchase intention.

Luxury brands have been slower to adopt social media channels (Hughes et al., 2019), given their desire for exclusivity (Chandon et al., 2016). Despite this initial reluctance, such brands now regularly use social media (Kim and Ko, 2012), where platforms such as Instagram offer opportunities for visual extensions of benefits to luxury fashion brands (Filieri et al., 2023). Luxury fashion brands, such as Louis Vuitton, Burberry, and Calvin Klein, live stream content from the catwalk via social channels (Plangger et al., 2021), which has allowed these brands to strengthen consumer–brand engagement (Dhaoui, 2014).

However, there is a limited understanding of luxury brands and the effectiveness of marketing efforts on social media platforms (Oc et al., 2023), and CEI research in the luxury domain is also limited (Giakoumaki and Krepapa, 2020). Moreover, there is a commonly held belief that the luxury sector has not experienced substantial advantages from social media, given the distinctive traits of luxury brands, such as exclusivity, scarcity, and prestige, which run counter to social media’s emphasis on widespread accessibility (Zha et al., 2023). Furthermore, personalized physical interactions and services offered in brick-and-mortar stores, integral to the luxury shopping experience, are difficult to replicate online (Hoang et al., 2022). This study examines the factors that drive continued engagement between SMIs and consumers to address these knowledge gaps and to understand the factors that elicit positive consumer responses in the context of luxury goods on Instagram.

Many brands prefer Instagram as their marketing communication platform owing to its 1.3 billion monthly active users, with 79% of marketers considering Instagram an important part of their campaigns (Santora, 2022). Instagram provides the context for this research because it is the preferred social media platform for luxury fashion brands and consumers (Santora, 2022). In the luxury space, Chiara Ferragni, an Italian blogger and fashion designer, has 27.7 million Instagram followers and 755,000 X/Twitter followers (DePino, 2023). However, studies on the use of SMIs by luxury brands remain scarce despite the number of luxury brands that now use social media (Athwal et al., 2019). To address these shortcomings, we develop and test a conceptual framework that examines consumers’ experiences with SMIs and how these experiences influence CEI. We also examine the influence of homophily attitudes toward SMIs on the relationship between these experiences and CEI and investigate the impact of CEI on consumer responses.

Accordingly, this study contributes significantly to the literature in several key respects. First, we focus on the challenges brands face in creating CEI and the drivers of CEI (Yuan and Lou, 2020). Second, we contribute to the growing but scarce literature on SMI and luxury branding marketing by responding to calls for further research on how SMIs can foster CEI (Hughes et al., 2019). Third, we examined the relationship between CEI and consumer responses, including purchase intention, intention to share, and adoption of the SMI recommendations. SMI-related consumer behavior and the mechanisms enabling positive behavioral response formation can be considered complex phenomena that cannot be explained by a single theory-driven model (Basha et al., 2022). Therefore, we developed and tested a holistic conceptual framework by drawing on a theory-integration approach (Cheung et al., 2022) that combines social exchange theory (SET) and flow theory.

Scholars have been focusing on the concept and drivers of CE for several years (Pradhan et al., 2023; Wang and Huang, 2023), which can be considered a form of behavioral intention (Hepola et al., 2020). Thus, CEI can be conceptualized as a set of measurable consumer intentions in response to content posted on social media by SMIs (Barger et al., 2016). CEI is an important factor in enhancing the interactions between consumers and brands in a social media context. CEI contributes to enhanced consumer outcomes such as self-brand connection, heightened empowerment, value co-creation (Giakoumaki and Krepapa, 2020), brand attachment and loyalty (Hollebeek et al., 2014), and consumer–brand relationships (Pansari and Kumar, 2017). Table 1 summarizes the selected studies that focus on engagement and behavioral outcomes from a social media perspective.

However, SMI–brand partnerships also have disadvantages. For instance, SMIs and brands are subject to “cancel culture” (Jones et al., 2022). Cancel culture refers to the act of “canceling” someone, which involves rejecting, ignoring, and publicly opposing their views or actions. It often entails depriving individuals of time and attention and even impeding their ability to sustain a livelihood (Goldsbrough, 2020). SMIs and their branded content can be scrutinized in detail. Thus, any missteps in terms of an SMI’s conduct or posts can lead to a drop in followers and their being “canceled,” which means the SMI would lose lucrative brand-partnership opportunities (Wei et al., 2022). The lucrative nature of these partnerships has prompted several SMIs to resort to unethical methods to amass followers and exert their influence (Tsapatsoulis et al., 2019). This includes tactics such as spoofing or stealing content from more popular profiles or deceiving followers who believe they are engaging with legitimate accounts (Marwick and Boyd, 2011). Furthermore, if their followers realize that commercial motivations drive an SMI’s involvement with a brand, it can have negative effects on the SMI’s trustworthiness and reputation (Singh et al., 2020). Consequently, any negative publicity for an SMI can put a brand at risk of being “guilty by association,” leaving it vulnerable to any criticism that the SMI may face (Pantano, 2021). Brands want to focus on developing long-term interactions beyond initial ones (Chen et al., 2020). Therefore, we posit that CEI is a better indicator of engagement than a conventional approach (Chen et al., 2020), where CEI allows brands to create, maintain, and develop long-term relationships with consumers and, thus, build long-term competitive advantage (See, 2018) as opposed to singular engagement platforms (Roy, 2016). There are value-added benefits to brands and SMIs that affect CEI (Chae et al., 2002; De Moor et al., 2010; Kim and Han, 2011; Pihlström, 2007). These value-added benefits could be social, utilitarian, or hedonically driven and could serve user needs (Revels et al., 2010) and encourage CEI (Lee and Kim, 2010). For example, SMIs, by holding Gucci handbags or wearing Prada on their social media posts, can illustrate a hedonic-driven lifestyle. Consequently, CEI has emerged as a pivotal metric for marketers, offering insights into whether consumers will maintain their relationship with a brand (Kwon et al., 2014).

Social media has changed how brands create and distribute content and how they communicate with and among consumers (Tsai and Men, 2013). Brands are aware of the potential benefits of cooperating with SMIs (Munnukka et al., 2019), including enhanced brand attitudes and perceptions and increased purchase intention (Lee and Watkins, 2016) and engagement (Sheng et al., 2024). This study focuses on consumers who follow SMIs (i.e. SMI followers) to better understand the relationship between consumers’ SMI experiences and their CEI for a specific SMI. Following Prentice et al. (2019), we conceptualize the SMI experience as a higher-order construct consisting of affective, intellectual and sensory experience dimensions evoked in consumers’ interactions with an SMI. In the context of Instagram, SMI experiences are evoked by a range of sensory experiences through SMI-created texts, visuals, colors, sounds, and videos. Sensory brand cues have the capacity to be presented in profoundly emotional terms, thereby potentially evoking more intense affective responses (Tafesse, 2016), that is, affective experiences. Intellectual experiences occur when SMIs provoke thoughts, stimulate curiosity, and influence problem-solving abilities, that is, cognitive processing (Prentice et al., 2019). Brand posts that leverage media-rich and interactive capabilities to foster a comprehensive brand experience are more likely to attract heightened CE (Tafesse, 2016). Homophily attitudes toward SMIs can be formed based on whether consumers perceive that SMIs think, behave, and are like them in many respects (Ladhari et al., 2020). Thus, SMIs seek to develop and maintain meaningful consumer interactions by enhancing consumers’ experiences (see Table 2) and homophily attitudes toward the SMI.

The experiences and homophily attitudes toward an SMI will determine the extent of engagement consumers are likely to have with an SMI, potentially leading to purchase intention, sharing, and recommendations. We use SET in parallel with flow theory to provide a conceptual underpinning for examining the role of SMI experiences and homophily attitudes toward SMI in cultivating CEI and consumer responses.

SET represents an interdependent relationship between the SMI and the follower, based on the principles of reciprocity and rewarding actions initiated by the SMI (Cortez and Johnston, 2020). Users commence and maintain relationships with positive exchange interactions (e.g. positive experiences and CEI), resulting in expected rewards such as financial benefits, friendships, and emotional satisfaction interactions (Lambe et al., 2001). For example, 61% of consumers have engaged with brands on social media in exchange for an incentive (Wiley, 2022). Users engage and interact with SMI, believing they will benefit from sharing and exchange (Shiau and Luo, 2012). Usually, if the SMI provides a benefit to a user (e.g. a positive experience), the user will reply in kind by engaging in positive reciprocating actions (Whitener et al., 1998), such as CEI (Cropanzano et al., 2017), and eventually enhance positive consumer responses.

Flow theory can help explain the factors that influence consumer experience (Novak et al., 2000). Flow is a cognitive state that enhances a consumer’s experience (Csikszentmihalyi, 1990) because of the heightened level of concentration, energy, and focus when engaging in a very specific activity (Novak et al., 2000), such as interacting with SMIs. This leads to high enjoyment, involvement, and concentration, making an activity interesting, gratifying, and compelling (Hyun et al., 2022). For example, a positive experience with SMI allows followers to experience flow, which leads to CEI and positive consumer responses. The SMI experience is so enjoyable that followers lose track of time and become cognitively locked into watching the SMI (Kim et al., 2020), resulting in CEI. Thus, the flow theory is appropriate for examining the impact of SMI experiences on consumer responses. In line with this discussion, we propose our conceptual model (Figure 1).

Consumer experiences with SMIs on various platforms constitute intricate interactions involving SMI-generated texts, visuals, colors, sounds, and videos, collectively shaping sensory experiences (Rose et al., 2012). These experiences, both tangible (e.g. features of a luxury handbag) and intangible (e.g. the emotional resonance of owning a luxury brand handbag; Ong et al., 2018), contribute to overall consumer perception. Consumers actively engage in the cognitive, affective, and sensory processing of information derived from these brand-post interactions, forming lasting impressions in their memory (Rose et al., 2012). Consequently, SMIs serve as pivotal conduits for brand interactions, potentially influencing the determination of positive consumer experiences (Buzeta et al., 2020). A flow experience in social media is characterized by a consumer’s cognitive state featuring a seamless sequence of responses, intrinsic enjoyment, loss of self-consciousness, and self-reinforcement during network navigation (Hoffman and Novak, 1996). Thus, a flow experience can explain how consumers will likely spend more time than they intend on SMI-related content.

A positive sensory flow experience triggered by SMIs will likely evoke favorable emotions, feelings, and sentiments that constitute affective experiences. Simultaneously, intellectual flow experiences unfold as SMIs prompt thoughtful consideration, stimulate curiosity, and influence consumers’ problem-solving abilities through cognitive processing (Prentice et al., 2019). Thus, a positive flow experience can be important in discerning continued engagement with SMIs. Based on flow theory, we posit that a positive SMI experience fosters CEI with SMIs, thereby shaping ongoing interactions. This leads to the following hypothesis:

Purchase intention is a direct indicator of eventual purchasing actions (Adelaar et al., 2003) and is an “individual’s conscious plan to make an effort to purchase a brand” (Spears and Singh, 2004, p. 56). In the context of SET, purchase intention denotes the intention to act favorably in response to information stimuli related to a brand or product (Kim and Johnson, 2016). However, recent studies have found that potential consumers are inclined to purchase a product promoted by SMI (Kemp, 2023). Given that flow is a cognitive condition that is intrinsically enjoyable and creates a very favorable experience (Novak et al., 2000), it leads to CEI and the intention to purchase (Hossain et al., 2018; Liu and Shiue, 2014). Therefore, individuals with a high CEI are more inclined to purchase. This suggests that followers who are deeply connected or exhibit a sense of attachment to an SMI are more likely to engage in purchasing behavior than those with lower attachment (Sokolova and Kefi, 2020). We posit that consumers exposed to luxury brands promoted by SMIs are more likely to purchase them owing to continued engagement with a brand’s SMI. Therefore, we hypothesize as follows:

Moreover, SMI marketing relies heavily on the spread of branded content being shared with SMI followers’ network of friends and family on social media platforms. The ability to share content, gain higher brand exposure, and acquire more brand followers is the central tenet of SMI marketing. Sharing intention refers to the retransmission of information generated by others, where content is shared in its original form beyond the reach of the current followers of the original poster, thereby developing a new level of exposure and new SMI followers (Kwak et al., 2010). For example, the unique features of Instagram facilitate users’ sharing of information with others by tapping on the heart symbol to show “love” for the posts made by other users or “sharing” by commenting on the post or sharing the post with other users privately by tapping the send icon. Sharing intention manifests when the SMI’s followers perceive that the posts are worth sharing with friends and family, resulting in their ability to view the posts (Chen and Lee, 2014). Therefore, we hypothesize that:

The adoption of an SMI recommendations is the process whereby the SMI’s followers intentionally engage with information use (Cheung et al., 2008) and consider the brand(s) recommended by the SMI based on the post (Rietveld et al., 2020). The information presented to SMI followers increases their brand knowledge, making purchasing decisions easier (Filieri et al., 2015). However, excessive repetition can be detrimental when followers perceive sameness in SMI content. Notably, 47% of consumers worldwide experience fatigue from repetitive SMI content, leading them to be less inclined to follow SMI recommendations (Concannon, 2023). Thus, followers who are deeply connected to the influencer look for recommendations compared with those with lower levels of attachment (Sokolova and Kefi, 2020). If an SMI’s followers believe the posts are reliable, they are more likely to adopt the SMI’s recommendations (Filieri et al., 2015). Therefore, we hypothesize that the more SMI followers engage with an SMI, the more likely they will be to follow the SMI’s recommendations.

Homophily is “the degree to which people who interact are similar in beliefs, education, social status, and the like” (Eyal and Rubin, 2003, p. 80). The concept of homophily suggests that individuals prefer associating with others they perceive as sharing similar values and status (Lazarsfeld and Merton, 1954). This tendency influences how people seek opinions, respond to content, and react to content generated by SMIs (Chu and Kim, 2011; Ismagilova et al., 2020). Furthermore, consumers are more likely to be influenced by recommendations from SMIs and purchase the products or brands (Ladhari et al., 2020) that share similarities with them (Ismagilova et al., 2020). Homophily attitude reflects the extent to which a person perceives that another person shares his or her attitudes (thinks, behaves, is similar, is like) (Ladhari et al., 2020). Perceiving a more homophily attitude (McGuire, 1985) creates feelings of connection, affection, and passion toward the SMI. Thus, understanding consumer homophily attitudes toward SMIs is crucial for comprehending the impact of SMI experiences on CEIs and the resulting consumer responses. We propose that if consumers have a favorable homophily attitude toward the SMI, they are more likely to intend to engage with it continually, which would then lead to positive consumer responses. Therefore, we hypothesize that:

The influence of CEI on consumer responses, such as purchase intention, sharing intention, and adoption of the SMI recommendations, is based on flow theory and SET, whereby CEI mediates the relationship between positive experiences and consumer responses. We propose that if the SMI experience is positive, consumers will likely continue to engage with SMI. This will lead to positive consumer responses, such as the intention to share, purchase, and adopt the SMI’s recommendations. This leads to the following hypotheses:

After obtaining institutional ethics approval, we tested the proposed hypotheses using data from an online survey of 604 Australian respondents. We first discuss the measurement items and the development of the survey instrument. Next, we discuss the sampling and data analysis procedures.

We adopted multi-item scales from marketing literature as measures for the constructs of the study (see Table 3). SMI experience is a multidimensional construct that is reflective-formative in nature (Jarvis et al., 2003). We drew upon items from Prentice et al. (2019) to measure the three SMI experience dimensions—affective, intellectual, and sensory. These dimensions explain how consumers are likely to spend more time than they intended to with the SMI-related content to evoke favorable emotions, feelings, and sentiments, constituting affective experiences. CEI was measured as the degree of intended future engagement with the SMI’s posts (Högberg et al., 2019). The items for CEI were adapted from Högberg et al. (2019). We define purchase intention as an individual’s conscious plan to make an effort to purchase a brand. The items for purchase intention were adapted from Fritz et al. (2017). The adoption of recommendation is when the SMI’s followers intentionally engage with information use (Cheung et al., 2008) and consider the brand(s) recommended by the SMI based on the post (Rietveld et al., 2020). The measures were adapted from Filieri et al. (2015). The measures of sharing intention, which is the retransmission of information generated by others, where the content is shared in its original form beyond the reach of the current followers, were adopted from Chen and Lee (2014). Homophily attitude has been characterized as the extent to which a person perceives that another person shares his or her attitudes toward the SMI (Ladhari et al., 2020). To measure homophily attitude toward the SMI, we adopted items from Ladhari et al. (2020). All measurement items used a 7-point Likert scale (1 = strongly disagree, 7 = strongly agree). Several control variables, such as gender, income, educational qualifications, and frequency of Instagram access, were also included to account for respondent heterogeneity (Carlson et al., 2021). We also incorporated a three-item marker variable, socially desirable responses, to test for potential common method variance (Donavan et al., 2004).

Three screening questions qualified respondents for the survey: those less than 18 years old, those not having an Instagram account, and those not interested in luxury brands were screened out. To test respondent attentiveness, we included a check question based on the SMI profile shown. Those who failed to answer the check question correctly were excluded.

Following an approach similar to that of Delbaere et al. (2020) and Belanche et al. (2021), we focused on popular fashion influencers on Instagram. Respondents were randomly shown one of two images of Instagram SMIs (one male and one female with a similar number of followers). For our SMIs, we selected Instagram accounts in the fashion industry with many followers who were growing in popularity (Casaló et al., 2020). The Instagram SMI profiles noted the luxury brands they endorsed and the number of followers and featured a styled photo with several comments posted by their followers (see Table 4).

The survey instrument was pre-tested using 60 respondents to ensure the interpretability of the questions, survey flow, validity, and integrity of the survey instrument. The full survey was administered once the survey instrument was deemed sound.

We administered an online survey using an online panel from a reputable commercial market research company. The survey yielded 604 complete Australian responses with the following characteristics: 66% female (34% male) respondents, with an average annual income of AU$50,001 to 100,000, and an average age of 27–36 years. In terms of the highest level of education, 32.1% held a bachelor’s degree, whereas 6.5% held a postgraduate certificate, 11.1% a master’s degree, and 2.5% a PhD/doctorate). In terms of the frequency of Instagram usage, 77.6% of respondents indicated that they logged in to their respective Instagram accounts five times or more per week, with 22.4% logging in less than five times a week (see Table 5).

We followed the data analysis procedure used in previous studies with multidimensional reflective-formative constructs in the path model (Merz et al., 2018). First, we analyzed the reliability and validity of the measurements for all constructs in the proposed path model (MacKenzie et al., 2011). Next, we examined whether evidence of common method bias affected the responses, statistically tested the significance of the hypothesized structural model relationships, and examined the mediating effect of continued engagement intention on consumer responses. Finally, we assessed the moderation effect of homophily attitudes on the relationship between SMI experience and CEI.

We conducted partial least squares-structural equation modeling (PLS-SEM) using SmartPLS 3.3.3 (Ringle et al., 2015) to evaluate the measurement model and estimate and assess the structural model relationships, including the moderating and mediating effects. First, PLS-SEM offers more reliable estimations with greater precision when the path model includes both reflective and formative constructs with many measurement items (Hair et al., 2017). Next, an analysis of normality indicated that most variables in this study were not normally distributed. PLS-SEM does not make any assumptions regarding the normality of the data and can produce good results when the data distributions are non-normal (Hair et al., 2017). Finally, PLS-SEM is a causal-predictive approach to SEM that produces useful estimates when a study aims to predict dependent variables (Carlson et al., 2021). As this study aimed to predict purchase intention, adoption of the SMI recommendations, and sharing intention of consumers from their experience with the SMI, we used PLS-SEM because this method would offer important estimates regarding the predictive quality of the model.

The measurement model of SMI experience follows the Type II reflective-formative hierarchical component modeling approach using the repeated-indicator approach (Hair et al., 2017). Following Hair et al.’s (2017) hierarchical guidelines for assessing the measurement model of SMI experience, we first assessed the reliability and validity of the lower-order constructs (i.e. affective, intellectual, and sensory experience) and subsequently examined the higher-order construct (i.e. SMI experience). The results indicate the reliability of the lower-order constructs (see Tables 6 and 7): the item loadings for all reflective measures of the lower-order constructs exceeded 0.708, and the Cronbach’s alpha, rho_A, and composite reliability scores exceeded 0.70. Moreover, all the average variance extracted (AVE) scores exceeded 0.50, and all the heterotrait–monotrait (HTMT) ratio values were below 0.90 (see Table 6), indicating that convergent validity and discriminant validity had been achieved (Hair et al., 2019).

For the formative higher-order construct, SMI experience, the outer weights of the lower-order constructs were meaningful in size (0.365, 0.364, and 0.373, for affective, intellectual, and sensory experience, respectively) and significant (p < 0.001; see Table 6). These results indicate that all lower-order constructs are significant and contribute almost equally to the formation of the higher-order construct (Hair et al., 2017). Moreover, the variance inflation factor (VIF) values for intellectual and sensory experience were less than 3, and for affective experience, less than 5, indicating that the measurement model did not suffer from multicollinearity (Diamantopoulos and Winklhofer, 2001). Hence, this analysis supports the reliable and valid measurement of SMI experience as a higher-order formative construct.

The reliability and validity of the variables, CEI, purchase intention, adoption of the SMI recommendations, sharing intention (see Table 7), and homophily attitude toward SMI (see Table 8), were assessed following the guidelines of Hair et al. (2019). The outer loadings of all corresponding reflective indicators exceeded the minimum threshold of 0.708; all the Cronbach’s alpha, rho_A, and composite reliability scores exceeded 0.70, and all the AVE scores exceeded 0.50, indicating the reliability and convergent validity of the measures. The reliability and validity analysis results showed that both reliability and convergent validity were achieved for the moderating variable (Cronbach’s alpha = 0.913, rho_A = 0.914, and composite reliability = 0.945), which was greater than the minimum threshold of 0.70, while the AVE = 0.852 was greater than 0.50. All the HTMT ratios were less than or equal to 0.90, indicating that the constructs were empirically distinct from each other (Henseler et al., 2015).

We employed several techniques to test for potential common method variance. First, following the partialling out a marker variable technique, we examined the changes in path coefficients and R² values of the endogenous constructs after including the marker variable in the path model (Podsakoff et al., 2003). The results (see Table 9) indicate that the path coefficients and R² values both changed very slightly owing to the inclusion of the marker variable, indicating that estimates are not affected by common method variance (Podsakoff et al., 2003). Second, we estimated the correlations between the marker variable and all other constructs in the path model (Lindell and Whitney, 2001). All correlation coefficients were less than 0.30 (see Table 9), which implies that the constructs in the path model were not significantly related to the marker variable, and common method variance was not a concern in this study (Lindell and Whitney, 2001). Finally, we examined the VIF values of all constructs. The low VIF scores (<5) also support the notion that the study was not seriously affected by common method variance (Kock, 2015).

We began the assessment of the structural model using the SmartPLS bootstrapping procedure with 5,000 bootstrapping samples and the bias-corrected and accelerated bootstrap (BCa) method (Hair et al., 2017) to examine the standardized path coefficient (β), t-values, p-values, and f² effect size. The results presented in Table 10 show that SMI experience has a significant and positive influence on CEI, with a large f² effect size (β = 0.768; t = 31.093; p = 0.000; f² effect size = 1.434). Similarly, the results also indicate that CEI has a significant and positive influence on purchase intention (β = 0.773; t = 33.082; p = 0.000; f² effect size = 1.486), adoption of SMI recommendations (β = 0.779; t = 34.840; p = 0.000; f² effect size = 1.545) and sharing intention (β = 0.808; t = 45.118; p = 0.000; f² effect size = 1.880). Thus, these results support H1, H2, H3, and H4.

Next, we examined the variance in the endogenous variables explained by the model (R²), which is a key criterion for assessing the quality of the structural model in PLS-SEM (Hair et al., 2019). Henseler et al. (2015) described an R² value of more than 0.67 as substantial, more than 0.33 as moderate, and more than 0.19 as weak. Our results indicate that the model has moderate explanatory power for all the endogenous variables because it accounts for 59% (R² adjusted = 0.588) of the variance of CEI, 60% (R² adjusted = 0.597) of the variance of purchase intention, 61% (R² adjusted = 0.606) of the variance of adoption of SMI recommendations, and 65% (R² adjusted = 0.652) of the variance of sharing intention. Moreover, because all R² values were less than 0.90, it can be concluded that the model did not suffer from overfitting (Hair et al., 2019). Furthermore, we applied the PLS_predict procedure with 10 folds and 10 repetitions to evaluate the predictive performance of the model (Hair et al., 2019). The results show that the Q²_predict values for the indicators of all endogenous variables are larger than 0 (minimum Q²_predict = 0.359; maximum Q²_predict = 0.513). Moreover, by comparing the root mean squared error values produced in the PLS-SEM analysis with those produced by the naive LM benchmark model, we found that in most cases, PLS-SEM produced smaller values than the naive LM estimations. Hence, from these results, we conclude that the model offers excellent predictive performance (Hair et al., 2019).

To test the moderation effect (Hair et al., 2019), we followed the PLS-SEM moderation analysis procedure by examining the significance of the interaction term (i.e. SMI experience × homophily attitude toward SMI). As presented in Table 11, the interaction term had a significant negative effect on CEI (effect = −0.066; t = 2.115; p = 0.034), meaning that with a more favorable homophily attitude toward SMI, the effect of SMI experience on CEI decreased. Conversely, with a less favorable homophily attitude toward SMI, the effect of SMI experience on CEI increased. Overall, these results empirically support H5 and affirm that homophily attitude toward SMI has a significant negative effect on the relationship between SMI experience and CEI.

As part of the mediation analysis, we evaluated the significance of both direct and indirect effects using the PLS bootstrapping procedure (Hair et al., 2017). The results presented in Table 12 show that the direct effect of SMI experience on purchase intention (0.256; t = 5.28; p = 0.000) and the indirect effect of SMI experience on purchase intention through CEI (0.443; t = 12.176; p = 0.000) were both significant. Therefore, this result empirically supports the notion that CEI partially mediates the relationship between SMI experience and purchase intention. Moreover, because both effects are positive, we can conclude that mediation is complementary (Zhao et al., 2010). Similarly, we found that both the direct effect of SMI experience on sharing intention (0.268; t = 5.887; p = 0.000) and the indirect effect of SMI experience on sharing intention through CEI (0.462; t = 13.743; p = 0.000) were positive and significant. Conversely, both the direct effect of SMI experience on the adoption of the SMI recommendations (0.294; t = 5.63; p = 0.000) and the indirect effect of SMI experience on the adoption of the SMI recommendations through CEI (0.425; t = 10.725; p = 0.000) were also positive and significant. Therefore, the results empirically support that CEI represents complementary mediation of the relationship from SMI experience to purchase intention, sharing intention, and the adoption of the SMI recommendations.

Despite the increasing role that SMIs play in working with luxury brands to generate CE, the factors influencing CEI in the SMI context and related outcomes remain under-researched. A brand–SMI collaboration may not always be successful owing to many reasons, such as the prevalence of cancel culture (Jones et al., 2022), SMI misconduct (Wei et al., 2022), unethical behavior (Tsapatsoulis et al., 2019), negative motivations (Singh et al., 2020), or influencer fatigue (Bright et al., 2015). However, our study shows CEI can be a key factor in facilitating long-term relationships between SMIs and consumers, in which engagement is behavior-based and stretches beyond brand purchase. Furthermore, our results suggest that CEI is beyond intention—consumers are engaged and will drive future behavior (Kim et al., 2013).

Moving beyond the theory of planned behavior (Madden et al., 1992) and the unified theory of acceptance and use of technology (Venkatesh et al., 2003), and drawing on SET and flow theory, we developed and tested a new conceptual framework and examined the relationship between consumers’ experiences with an SMI and their CEI with a luxury brand. We also investigated the moderating role of homophily attitudes toward SMIs. Our study shows that brands should focus on long-term interactions beyond initial interactions (Chen et al., 2020), which may create value for consumers and subsequently drive involvement and loyalty (See, 2018). We also examined the mediating role of CEI on the relationship between SMI experience and consumer responses, such as adoption of SMI recommendations, purchase intention, and sharing intention. Our findings suggest that consumers’ SMI experience has positive effects on their CEI with a brand. This result suggests that because consumers have a positive experience with SMI, they forge a stronger relationship with that SMI and are less likely to defect from that relationship (Kwon et al., 2014).

Moreover, CEI has a positive mediation effect on the relationship between the SMI experience, their adoption of the SMI recommendations, and their sharing and purchase intention. CEI plays a critical role in consumer responses, such as the intention to purchase or share and the adoption of the SMI recommendations. Our data also suggest that the homophily attitude toward the SMI moderates the relationship between the SMI experience and CEI. Interestingly, the strength of the influence of SMI experience on CEI decreases when a consumer possesses a more positive homophily attitude toward an SMI. Therefore, consumers may demonstrate some level of CEI owing to their more positive homophily attitude toward an SMI, irrespective of their SMI experience. Collectively, our findings show that creating positive SMI experiences benefits brands that seek continued engagement with their consumers. This research addresses calls for research by adding to the scarce literature on SMI marketing on how SMIs can foster CEI and the resulting outcomes.

SMIs have become integral to brands’ online marketing strategies that attempt to enhance consumers’ CEI. This study provides two important managerial implications in the context of luxury brands. First, we present some notable implications for practitioners planning to implement SMI marketing strategies. Brand managers must focus on developing long-term relationships that could be positively affected by consumers’ positive experiences with SMIs. Marketers can attempt to ensure that all the information and content provided by SMIs are relevant to consumers. Thus, a key objective for marketing practitioners should be working in partnership with SMIs to create positive SMI experiences (e.g. intellectual, affective, and sensory experiences) to foster CEI. Second, SMIs should focus on brands that fit their lifestyles and values (Kim and Kim, 2020) to increase CEI, enhance consumers’ homophily attitudes toward them (Belanche et al., 2021) and forge a stronger relationship with their followers. In the context of luxury brands, SMIs that focus on the newest trends, high-fashion styles, and desirable lifestyles are likely to exert a positive influence on consumer homophily attitudes to impress, delight, and attract audiences (Yan et al., 2023), leading to CEI, which, in turn, can have significant positive effects on consumer responses. For example, Balenciaga collaborated with Fortnite, allowing players to acquire digital outfits mirroring real-life Balenciaga fashion through a virtual boutique (Duggal, 2022).

This study is cross-sectional in its design, and it comes with several limitations, which, in turn, present opportunities for future research. First, this cross-sectional study was limited to the use of Instagram in Australia. Future studies could replicate this study in other contexts using different social media platforms, such as Facebook, X (formerly known as Twitter), or TikTok, in other countries and longitudinally to assess the generalizability of the model. As this study focuses on Australian Instagram users interested in luxury brands, this can limit the generalizability of the findings to other cultural contexts or social media users with different interests. The choice of social media platforms may indicate different experiences with SMIs and consumer responses. This study analyzed followers’ perceptions of and responses to two SMIs in the luxury market. Thus, caution should be exercised when generalizing these results. Future studies could examine SMIs who present different characteristics and audiences with different profiles to ensure the generalizability of the results. Furthermore, the focus of the study was limited to luxury brands because they have become a widespread phenomenon in social media marketing. In addition, the study did not consider respondents’ nationalities and cultural backgrounds (Nikolinakou and Phua, 2024). Lastly, from a theoretical perspective, future studies could examine the consequences of CEI on other consumer responses, such as brand commitment and brand loyalty.

This research has been supported by a Faculty Grant from the Newcastle Business School, University of Newcastle, Australia.