1 Introduction

The ever-evolving realm of social media has been fertile ground for the emergence of a new digital profession. Social media influencers (SMIs) are personalities who, through content creation, have built a solid followers base on whom they exert a certain degree of influence (Lou & Yuan, 2019). Marketers are recognising the importance of collaborating with SMIs. In influencer marketing strategy, brands entrust influencers with the creation and sharing of branded contents that includes endorsement or product mention (Lou & Yuan, 2019) to change opinions and behaviours of followers toward products or services (Farivar et al., 2021). Reaching a market value of $21.1 billion in 2023 (Statista, 2023), this strategy has advantages over traditional celebrity collaborations. For instance, SMIs are perceived as more credible and similar by consumers due to their genuine positive attitude toward brands (Schouten et al., 2020).

The SMIs literature has explored how influencers can guide followers towards positive attitudes and behaviours regarding endorsed brands (e.g., Lee & Watkins, 2016) and their own profile (e.g., Casaló et al., 2020). There are two main categories of influential factors. On one hand, most studies focused on the role of influencer characteristics such as the parasocial relationship, which refers to an apparent friendship, the status of opinion leader, credibility, trustworthiness, expertise, physical and social attractiveness (e.g., Farivar et al., 2021; Lee & Watkins, 2016; Masuda et al., 2022; Sokolova & Kefi, 2020). On the other hand, followers engage with the content created by influencers evaluating characteristics such as information quality, design quality, originality, and uniqueness of the content (e.g., Casaló et al., 2020; Cheung et al., 2022; Kim, 2022).

From the analysis of the literature, especially in the research area of content characteristics to which this study aims to contribute, the following gaps have emerged. Firstly, SMIs scholars tend to draw more heavily from other fields than from marketing knowledge to explain the phenomenon under analysis (Zeithaml et al., 2020). In this regard, most studies utilised psychological perspectives established from traditional celebrity studies (Cheung et al., 2022), describing and subsequently analysing the activity of SMIs as a “means of customer persuasion” (Barta et al., 2023, p. 4) aimed at prompting them to engage in desired behaviours (Farivar et al., 2021). As a result, the influencer is considered more as a persuader than as a service provider for both followers and brands, and marketing perspectives such as e-service and e-service quality (Santos, 2003) are overlooked. Secondly, the role of satisfaction needs further exploration (Kim, 2022). While most studies have utilised opinion leadership (e.g., Casaló et al., 2020) and parasocial relationships (e.g., Lou & Kim, 2019) as mediators between content characteristics and follower behaviour, these constructs do not conceptually separate the follower-influencer relationship from the concurrent use and evaluation process of the service they provide. Thirdly, studies have yielded contrasting results regarding which characteristics followers consider in the decision-making process. For instance, entertainment and information were not often simultaneously significant (Lou & Kim, 2019; Lou & Yuan, 2019). Moreover, research focusing on YouTube channels (Kim, 2022; Sokolova & Perez, 2021) utilised some characteristics that are not well-aligned with the specific context under analysis.

This study aims to expand the exploration of the influencer category from an e-service perspective by examining the role of user evaluation of content characteristics in shaping their satisfaction and behaviours. The theoretical model was tested using structural equation modelling applied to data collected from followers of YouTube influencers.

YouTube is a social media platform categorised as a video content community (Kaplan & Haenlein, 2010). Youtubers are expert creators of long-form video content that they share on their channels, serving as specialised media addressing specific topics relevant to the channel owner and its followers (Sokolova & Perez, 2021). The choice of this context was driven by two main reasons. Firstly, the professionalism exhibited by youtubers is akin to a service rather than an amateur pursuit. The intense competition among channels and the demanding nature of users have prompted youtubers to pay increasing attention to the quality of content and additional services offered, such as merchandising and subscription benefits. Indeed, to differentiate their production and attract viewers, they have invested in collaborators and equipment to support pre- and post-production stages such as decisions on topic and format mix, video script, set design and editing (France et al., 2021). Secondly, YouTube is one of the main platforms used in influencer marketing campaigns, incorporating commercial messages like sponsored content, affiliate marketing, and product placement (Schwemmer & Ziewiecki, 2018). Indeed, the type of content shared on this platform is more suitable than other format to promote the characteristics of goods or services (Ladhari et al., 2020).

This study contributes to the literature in three ways. First, conceptually, it integrates literature on influencer marketing, e-service, and e-service quality to provide a new perspective on SMIs. Beyond being influencers, these personalities act as service providers for social media users through content creation and for brands through endorsement as digital communication partners. Second, it demonstrates how satisfaction with the service provided by influencers and user loyalty toward the channel, acting as serial mediators, explain the path from evaluating content characteristics to behavioural engagement, purchase intention of sponsored brands, and purchase intention of influencers’ brands. In particular, the latter is a behaviour not previously explored in the SMIs literature. Third, the study identifies characteristics of a YouTube channel and its content, revealing that both information and entertainment quality are important in the follower evaluation process. Moreover, the originality and interactivity of the content also explain satisfaction and behaviours.

Additionally, this study offers guidance for content creators to enhance the quality of their digital environment and insights for marketing practitioner to choose appropriate influencer partners.

2 Theoretical background and hypothesis

Influencers’ production has professionalised with the evolving landscape of diversified social media platforms. Most of the literature, which distinguishes influencers for their power of influence on users (Barta et al., 2023), fails to capture the complexity of their role. The activities of SMIs can be more aptly described as a system of services, the characteristics of which are evaluated by customers for decision-making purposes.

According to Edvardsson et al. (2005), service is a perspective that concerns the co-creation of value rather than a category of market offerings. The interactive, processual, experiential, and relational nature of service forms the basis for its representation. In the context of the web, Taherdoost et al. (2012) define e-services as electronic-based services that provide interactive and content-centered services over electronic networks. Given that, it is appropriate to describe SMIs as e-service providers for two types of customers. Firstly, they are content creators who generate and distribute digital content to consumers in the interactive social media landscape. Secondly, they serve as primary communication partners for brands across social media. In both these capacities, the relationship with the customer and the co-creation of content are central.

In their decision-making process, customers evaluate e-service quality based on its characteristics. Santos (2003) defined e-service quality as the consumers’ overall perceptions, evaluations, and judgments of services provided through electronic media. Establishing e-service quality is critical to developing enduring relationships with customers in the virtual space, differentiating service offerings, and gaining a competitive edge in terms of user satisfaction and loyalty (Rita et al., 2019; Santos, 2003). Although researchers have studied e-service quality in various contexts like e-commerce and online banking (Parasuraman et al., 2005; Teeroovengadum, 2022), this perspective has received limited attention in the context of social media. Given that, the e-service quality approach can be employed to assess the effects of SMIs’ service quality factors (i.e., content characteristics) on the satisfaction and behavioural intentions of their followers.

An emerging branch of SMIs literature has analysed how content characteristics impact follower behaviours (see Table 1), such as purchase intention toward sponsored brands (e.g., Lou & Yuan, 2019) and the intention to recommending and continue to follow the influencer (e.g., Casaló et al., 2020). In the context of YouTube fitness channels, information quality, social interaction among followers, and the presence of relevant content explain satisfaction and channel loyalty (Kim, 2022). Additionally, entertainment improves the attitude towards the YouTube channel (Sokolova & Perez, 2021). Studies have also explored the antecedents and consequences of the opinion leadership status and parasocial relationship of Instagram and TikTok SMIs. Content attributes, such as originality, uniqueness, humour, information, entertainment, design quality, and interactivity, can be linked to favourable follower behaviours (Barta et al., 2023; Casaló et al., 2020; Cheung et al., 2022; Fakhreddin & Foroudi, 2022; Jegham & Bouzaabia, 2022; Ki & Kim, 2019; Lou & Kim, 2019). Moreover, SMIs must maintain consistent information and image cues across the various social media they use to enhance positive responses from followers (Hsieh, 2023).

Table 1 Review of SMI content characteristics literature
Full size table

2.1 The impact of channel characteristics on user satisfaction

The literature review has identified five constructs that potentially represent the characteristics of a YouTube channel and its content: content originality (Casaló et al., 2020), information quality (Kim, 2022), entertainment quality (Acikgoz & Burnaz, 2021), interactivity (Xiao et al., 2018), and video quality (Ki & Kim, 2019).

Originality refers to the level of novelty and differentiation perceived by individuals regarding a specific content compared to other alternatives and indicates the extent to which it is perceived as creative, innovative, unusual, and sophisticated by the user (Acar et al., 2017). Perceived originality positively influences the generation of word-of-mouth (Moldovan et al., 2011). Studies have also shown that the originality of contents on Instagram and TikTok positively impacts the perception of influencer opinion leadership (Barta et al., 2023; Casaló et al., 2020; Jegham & Bouzaabia, 2022). Based on these considerations, we state that:

H1a

Content originality positively impacts YouTube user satisfaction.

Information quality refers to the completeness, usefulness, accuracy, and currency of oral and textual information on a topic provided by influencers to their community (Kim, 2022; Kim & Kim, 2017). Specifically, users with niche interests that demand expertise are more inclined to assess the information quality of the channel (Kim, 2022).

Studies have verified that the information quality of a website or YouTube and Instagram influencer contents, positively influences users’ trust (Ho & Lee, 2015; Lou & Yuan, 2019), users’ satisfaction (Kim, 2022), influencer opinion leadership (Ki & Kim, 2019), user’s engagement towards sponsored brands (Gupta et al., 2023), and users’ loyalty (Kim & Niehm, 2009). Based on these considerations, we state that:

H1b

Information quality positively impacts YouTube user satisfaction.

The entertainment refers to the pleasure, excitement, and enjoyment gained during an experience (Jung et al., 2021). In the context of SMIs, entertainment quality pertains to users’ evaluations of whether they find the content entertaining, enjoyable, and exciting (Acikgoz & Burnaz, 2021).

Studies have demonstrated that entertainment value has a positive impact on users’ positive attitudes toward the youtuber and their content (Acikgoz & Burnaz, 2021; Sokolova & Perez, 2021), user engagement with Instagram sponsored brands (Gupta et al., 2023) and the development of parasocial relationships (Lou & Kim, 2019). Moreover, entertainment perceived by users during e-shopping influences their loyalty toward the e-commerce (Kim & Niehm, 2009). Finally, the perceived entertainment while using augmented reality positively impacts users’ satisfaction with such applications (Jung et al., 2021). Based on these considerations, we state that:

H1c

Entertainment quality positively impacts YouTube user satisfaction.

Previous studies have considered social interaction among subscribers as a characteristic of a youtuber’s channel, demonstrating a positive influence on user satisfaction (Kim, 2022) and attitudes toward videos (Sokolova & Perez, 2021). Except for specific communities, YouTube is not a social media that encourages conversation between subscribers, who usually use the comments section to leave their feedback in the hope of an answer from the influencer. For this reason, we replaced the concept of social interaction with a dimension more representative of the context, namely interactivity. In general terms, interactivity is defined as the extent to which users can participate in editing the content of a media in real-time and as the media’s ability to enable user-to-user communication (Burgoon et al., 2002). Based on the research of Xiao et al. (2018), we refer to this dimension in a twofold guise: interactivity as interpersonal communication (i.e., the degree to which the youtuber attempts to communicate and interact with its subscribers); secondly, to the degree to which the youtuber involves subscribers in the co-creation of content by receiving requests and suggestions from them. On Instagram, studies indicated that interactivity positively influences users’ perceptions of influencer authenticity, trust, and opinion leadership (Jun & Yi, 2020; Ki & Kim, 2019).Based on these considerations, we state that:

H1d

Interactivity positively impacts YouTube user satisfaction.

A crucial factor contributing to user satisfaction is the aesthetic experience, involving the pleasurable sensations derived from sensory experiences (Jung et al., 2021). Given that YouTube videos are visual media, followers assess video quality, conceptualised as the extent to which a follower perceives a content to be graphically or aesthetically appealing (Ki & Kim, 2019).

Studies have demonstrated that dimensions related to the graphic quality of content on Instagram and TikTok positively impact parasocial relationships, wishful identification (Cheung et al., 2022), and influencer taste leadership (Ki & Kim, 2019). Additionally, website design is identified as one of the determinants of user satisfaction (Rita et al., 2019). Thus, we assert that:

H1e

Video quality positively impacts YouTube user satisfaction.

2.2 The impact of user satisfaction on channel loyalty

Customer is satisfied when a service meets or exceeds expectations providing a sense of pleasure in achieving needs, desires, or goals (Oliver, 1999). In Kim (2022), satisfaction towards YouTube channel is defined as the overall emotional state of subscribers resulting from their cumulative experiences with the channel.

Customer loyalty towards a service could be defined as a profound dedication to repeatedly choose a favoured service irrespective of situational influences or marketing initiatives (Oliver, 1999), resulting in recommending these services to others and speaking positively about them (Zeithaml et al., 1996). In the context under analysis, channel loyalty represents the followers’ intention to continue consuming the youtuber’s content and recommend channel to others. Previous studies have shown that user satisfaction leads to increased loyalty to e-services (e.g., Ho & Lee, 2015; Rita et al., 2019; Teeroovengadum, 2022) and YouTube channels (Kim, 2022). Therefore, we conclude that:

H2

User satisfaction positively impacts channel loyalty.

2.3 The impact of channel loyalty on purchase intentions and behavioural engagement

Purchase intention can be defined as the consumers’ intention to purchase a product or service based on their subjective judgment from their general evaluations (Blackwell et al., 2001). This study explores followers’ intent to purchase two distinct types of products and services. On one hand, the purchase intention of sponsored brands involves third-party brands recommended by the SMI as part of influencer marketing collaborations. On the other hand, the purchase intention of influencers’ brands pertains to products or services offered by the influencer’s own brand, such as merchandise and channel subscriptions that grant access to paid content or other benefits.

While previous studies have delved into the purchase intention of sponsored brands, the exploration of influencers’ brands is novel. For instance, purchase intention of sponsored brands is influenced by parasocial relationships, credibility (Masuda et al., 2022; Sokolova & Kefi, 2020), opinion leadership (Ki & Kim, 2019), and the identification with the influencer (Hsieh, 2023). Furthermore, purchase intention can be a consequence of user loyalty. Within the context of e-services (Balakrishnan & Griffiths, 2018; Ho & Lee, 2015), online communities (Kim & Lee, 2019), and celebrities (Ma et al., 2022), studies demonstrated that a loyal consumer exhibits a higher intention to purchase products or services. Based on these considerations, we state that:

H3

Channel loyalty positively influences user purchase intention of sponsored brands.

H4

Channel loyalty positively influences user purchase intention of influencers’ brands.

Behavioural engagement refers to the level of energy, dedication, and time spent interacting with a social profile, expressed by the consumer reacting, commenting, and sharing a post (Onofrei et al., 2022). If users believe that the content posted is of high quality, they will be more likely to show their approval or support other people on social media (Onofrei et al., 2022). Based on these considerations, we state that:

H5

Channel loyalty positively influences user behavioural engagement.

The proposed model (see Fig. 1) aims to verify which characteristics of a YouTube channel influence the behavioural intentions of the followers through the serial mediation of user satisfaction and channel loyalty.

Fig. 1
figure 1

Research model

Full size image

3 Methodology

3.1 Data description

For the empirical evaluation of the theoretical model, an online survey was utilised. The survey was targeted toward Italian YouTube users who followed at least one channel in which a youtuber makes recommendations of products or services. The respondents were asked to think about and answer the questions based on a YouTube channel or youtuber who typically makes product or service recommendations. Out of a total of 420 respondents, only 304 passed the filter questions set. Additionally, six interviews were removed because incomplete. The final sample size was 298 users, characterised as follows: age 18–24 (40.9%), 25–34 (51.0%), over 35 (8.1%); male (59.1%); uses YouTube 3–4 times a week (26.5%), uses it once or several times a day (63.4%); 68.5% of them have already purchased a product recommended by a youtuber. The YouTube channels chosen by the respondents for the survey belong to the following categories: fun & humour (14.0%), fashion and beauty (11.0%), education (11.0%), gaming (11.0%), food (11.0%), others (42.0%).

3.2 Measures

The constructs and items shown in Table 2 were adapted from previous studies and evaluated on a five-point Likert scale ranging from strongly disagree (1) to strongly agree (5).

The literature review identified five constructs potentially representative of the characteristics of a YouTube channel and its content. Specifically, content originality by Casaló et al. (2020), information quality by Kim (2022), entertainment quality by Acikgoz and Burnaz (2021), interactivity by Xiao et al. (2018), and video quality by Ki and Kim (2019). Finally, other measures are user satisfaction and channel loyalty by Ho and Lee (2015), purchase intentions by Chetioui et al. (2020), and behavioural engagement by Onofrei et al. (2022).

4 Results

4.1 Measurement model and structural model

To validate the measurement scales, the standardized items loadings on the relative constructs were first assessed (all greater than 0.5) and then the internal consistency of each construct was examined, using composite reliability (found to be greater than 0.7 for all constructs) and Cronbach’s alpha (ranging from 0.786 to 0.905). Finally, the convergent validity of each construct was examined, with the average variance extracted (AVE) found to be greater than 0.50 for all constructs. Although one dimension did not fit within the overall model, i.e. behavioural engagement, it was decided to still include it in the structural equation model analysis. Further investigation is necessary to understand the implications of this finding and its impact on the results (Table 2).

Table 2 Items and scale properties for each construct
Full size table

For the analysis of the research model (Fig. 1), structural equation modelling (SEM) was employed using the maximum likelihood estimator and applying the bootstrapping procedure with 5,000 random samples (AMOS 27). The fit indices indicate a good fit between the model and the data: χ²/df = 1.754; CFI = 0.936; RMSEA = 0.050; SRMR = 0.070.

Hypotheses H1(a-e) assumed that the evaluation of the attributes of a YouTube channel and its content positively influences subscribers’ satisfaction with the channel. The empirical results found that user satisfaction is significantly and positively influenced by 4 of the 5 identified attributes, namely: content originality (β = 0.124; p < 0.05), information quality (β = 0.366; p < 0.01), entertainment quality (β = 0.359; p < 0.01) and interactivity (β = 0.153; p < 0.05). The effect of information quality and entertainment on satisfaction is similar, as well as being much stronger than the remaining two characteristics. Consequently, H1a, H1b, H1c, H1d are accepted, while H1e is rejected. In particular, the impact of the video quality is negative but not statistically significant. Regarding the latter, the results could suggest that the measures do not accurately capture the graphic quality of the content.

Hypotheses H2, H3, H4, and H5 was all accepted. The effect of satisfaction is positive and statistically significant on channel loyalty (β = 0.792; p < 0.01). Finally, the effect of channel loyalty is positive and statistically significant on purchase intention of sponsored brands (β = 0.290; p < 0.01), purchase intention of influencers’ brands (β = 0.330; p < 0.01), and behavioural engagement (β = 0.479; p < 0.01). The results are presented in Fig. 2.

Fig. 2
figure 2

Results of the structural model

Full size image

4.2 Model comparison

In order to assess whether the serial mediation of satisfaction and loyalty is either full, as in the research model, or partial, a comparison between nested models was conducted using a formal delta chi-square test and by comparing their fit indices.

For the comparison, four models were considered, each involving the serial mediation of satisfaction and loyalty in the relationship between content characteristics and behaviours: (i) the “Research Model” with total mediation; (ii) the “1st Partial Mediation Model” with additional direct relationships between characteristics and loyalty; (iii) the “2nd Partial Mediation Model” with additional relationships between characteristics and purchase intention of sponsored brands, purchase intention of influencers’ brands, and behavioural engagement; (iv) the “3rd Partial Mediation Model” with additional relationships between satisfaction and purchase intention of sponsored brands, purchase intention of influencers’ brands, and behavioural engagement. Each model was tested on the entire sample (n = 298) using SEM with maximum likelihood estimator and a bootstrapping procedure with 5,000 random samples (AMOS 27). The results of the model comparisons are reported in Table 3.

The Δχ2 test between the “Research Model” and the “1st Partial Mediation Model” is significant (p < 0.05), indicating that the restrictions significantly reduce the model fit; therefore, the “1st Partial Mediation Model” performs better than the “Research Model”. This result, along with the fit indices, suggests that satisfaction partially mediates the relationship between content characteristics and channel loyalty.

Similarly, the Δχ2 between the “1st Partial Mediation Model” and the “2nd Partial Mediation Model” is significant (p < 0.05), indicating that the restrictions significantly reduce the model fit; therefore, the “2nd Partial Mediation Model” performs better than the “1st Partial Mediation Model”. This result, along with the fit indices, suggests that satisfaction and loyalty partially mediate the relationship between content characteristics and purchase intention of sponsored brands, purchase intention of influencers’ brands, and behavioural engagement.

However, the Δχ2 test between the “2nd Partial Mediation Model” and the “3rd Partial Mediation Model” is not significant (p > 0.05), indicating that the restrictions do not significantly reduce the model fit; therefore, the “2nd Partial Mediation Model” has a similar performance to the “3rd Partial Mediation Model”, but it is more parsimonious and therefore the better alternative. This result, along with the fit indices, suggests that channel loyalty fully mediates the relationship between satisfaction and purchase intention of sponsored brands, purchase intention of influencers’ brands, and behavioural engagement.

In conclusion, the comparison results demonstrate that the “2nd Partial Mediation Model” performs the best. This suggests that the serial mediation of satisfaction and loyalty in the relationship between the evaluation of content characteristics and follower purchase intention of sponsored brands, purchase intention of influencers’ brands, and behavioural engagement is partial rather than complete. This result could be explained by the absence of certain mediators, discussed in the conclusion sections, which could be further explored in future research.

Table 3 Results of model comparison
Full size table

5 Conclusion

Building on influencer marketing and e-service literature, this research depicts SMIs work as service whose characteristics are evaluated by followers to determine their behaviours on social media. The results reveal that information quality, entertainment quality, interactivity, and content originality of a YouTube channel influence user satisfaction with the service provided by SMIs. Furthermore, satisfaction predicts follower loyalty to the channel, which, in turn, positively influences other behaviours towards endorsed brands (purchase intention of sponsored brands) and the influencer’s profile (purchase intention of the influencer’s brands and behavioural engagement).

This work has yielded significant theoretical contributions. First, it introduces the e-service perspective as a new lens for comprehending the SMIs phenomenon, effectively bridging theory and practice of the field. Previous literature has often portrayed influencers’ activities as a persuasion process aimed at their followers (Masuda et al., 2022). Beyond their role as influencers, these individuals operate as service providers for both users and brands. On one hand, they are recognised as content creators, actively involved in crafting and sharing different social media content formats based on their interests and community demands. On the other hand, they act as communication partners for brands, contributing to influencer marketing strategies. Second, this research underscores the role of satisfaction in the context of SMIs. Previous literature has frequently utilised constructs like opinion leadership (e.g., Casaló et al., 2020) and parasocial relationships (e.g., Lou & Kim, 2019) as mediators between content characteristics and behaviours, not conceptually separating the role of influencers characteristics from content factors. Our findings demonstrate that satisfaction and loyalty are suitable for describing the journey that leads followers from assessing the quality characteristics of the creators’ digital environment to their behaviours. Notably, the results confirm the primary effect of satisfaction in determining follower loyalty to the channel (Kim, 2022). Consequently, a loyal follower is more likely to engage with the channel (i.e., behavioural engagement), purchase brands sponsored by SMIs, and purchase products or services from influencers’ brands. The latter dimension, capturing the additional services the influencer offers through YouTube, has not been considered in the literature before.

Third, this study identifies which content characteristics are appropriate to the specific context of a YouTube channel, revealing which of these impact followers’ satisfaction and behaviours. Previous literature has primarily focused on analysing the Instagram context, sometimes producing conflicting results regarding the informative and entertaining dimensions (Lou & Yuan, 2019). Moreover, studies analysing YouTube channels (Kim, 2022; Sokolova & Perez, 2021) did not use characteristics such as content originality and interactivity. Our results demonstrate that both information and entertainment quality are important factors explaining follower satisfaction and choices. Furthermore, for the first time to our knowledge in SMIs domain, we found a significant effect of content originality and interactivity on satisfaction.

Our findings suggest valuable implications for practitioners. Concerning youtubers, their primary focus should be on investing in content quality to craft an immersive experience that satisfies subscribers. Firstly, the quality of information and entertainment carries the most weight in user evaluation. To achieve this, youtubers need to invest time in understanding their topics and developing their unique entertainment style. Consequently, effective video scripting becomes crucial. Starting with a compelling introduction that outlines the topic’s relevance, the content should follow a logical flow, presenting information concisely to capture attention and convey the intended message in an enjoyable and pleasing manner. Secondly, creator need to create a virtual environment in which they actively interact with their users. By leveraging the interactive features of the social media, they should make users feel part of a community involved in the process of new content creation. Thirdly, to differentiate themselves from competitors, they should be able to continuously innovate their productions by providing content that can be perceived as innovative, creative, and original in comparison to other channels. In this regard, they should create the right mix of content to meet the various needs of users, combining old successful formats with new proposals.

Furthermore, satisfied followers are more likely to support the creator by continuing to follow, recommend to others, and engage with the channel. These behaviours enhance the youtuber’s visibility, attract new subscribers, and open avenues for revenue expansion through merchandise and paid content subscriptions. Additionally, brands recognise them as valuable partners for influencer marketing strategies.

Regarding brands, firstly, the study confirms that the number of reactions can serve as KPI to assess influencers as potential partners. Secondly, beyond evaluating influencers based solely on personal attributes, brands should also assess them based on the qualitative and quantitative characteristics of the digital environment they create. In fact, when users positively value the quality of a channel, they are more inclined to purchase products or services recommended by influencers in branded content.

This study has some limitations that can guide future research directions. Firstly, the sample size is limited and considers only Italian users. Consequently, our findings should be generalised with caution. Secondly, the results indicate a non-significant effect of video quality on satisfaction. This outcome may be attributed to an inappropriate operationalisation process. Additionally, other dimensions representing YouTube content characteristics, such as content duration and audio features, should be identified. Thirdly, the model did not consider contextual variables. Future research could explore the potential moderation effects of user characteristics (e.g., age, gender, and education level) or channel type (e.g., number of followers, content category). Lastly, the model considers only satisfaction and loyalty as serial mediators. In the comparison between the research model with full mediation and the models with partial mediation, partial mediation prevails. Consequently, the absence of additional mediators is probable. Future research should consider other dimensions of relationship marketing, such as trust and commitment, as well as the e-service quality construct, as potential parallel or serial mediators. Moreover, to assess the influence of both influencer and content evaluation on follower behaviours, future research could compare satisfaction with attitudes towards the influencer (e.g., opinion leadership, parasocial relationship).