Discussion

the interview technique. The questions used for this interview guide were based on questions for quantitative research on PSI, e.g. Labrecque (2014). This could have led to questions that were too direct or guiding, since questions used for surveys are usually more closed. The questions used for the interview guide were adapted to fit qualitative interviews, yet maybe further adaptation is needed.

Although the respondents show signs of feelings of intimacy and emotional bonds towards the influencers across all three cases, none of the respondents express any illusionary experience with the influencer. In relation to the illusionary experience the respondents express an awareness of the realistic nature of the relationship between them and the influencer. The respondents acknowledge this relationship as being more commercial and transactional than an illusionary experience. In relation to this, Uzunoğlu and Kip (2014) identify the commercialization of bloggers as a possible threat to relationships between influencers and their followers. This in turn leads to the suggestion that the commercial nature of the influencer, specifically the ones used for this study, could disrupt PSI. The respondents’ awareness of the fact that being an influencer is a job, could make it more difficult for the respondents to experience an illusionary experience. The explanation for this could be found in the study conducted by Liljander, Gummerus and Söderlund (2015) who suggest that readers lose interests in blogs that are open about their sponsorship and thus can be considered commercial. Further, interest in a blog can be seen as a prerequisite for having an illusionary experience and thus it appears less likely that an illusionary experience will take place in a relationship that is perceived as commercial, such as the studied cases in this thesis.

Additionally, in contradiction to theoretical propositions that suggest PSI is developed over time by

frequent interactions (e.g. Eyal and Rubin, 2003), the respondents in the case of Isabella Löwengrip

did not exhibit any PSI even though they interacted with the influencer regularly during the last ten

years. Some researchers however argue that PSI can also develop through single interactions

(Labrecque, 2014; Hartmann and Goldmhoorn, 2011). Thus, the finding in the case of Isabella

Löwengrip seems to add understanding and support for the notion that PSI is not necessarily related

to the longitude of the relationship.

7.2 Reflections on trust

Across all three cases the respondents express a cross-situational trust in the influencer, separate from the connection to sponsored content and PSI. This might be considered an unanticipated finding in the sense that trust is related to sponsored content and PSI in the theoretical framework.

Labrecque (2014) suggests PSI can build trust, yet since there is no clear evidence of PSI in the conducted study, the general trustworthiness that the respondents exhibit could be attributed to other reasons. Across the cases the respondents mention that being trustworthy relates to being transparent by openly communicating. The explanation for the connection between openness in communication and trust could be found in alternative conceptualizations of trust. Drawing on concepts from relationship marketing, several antecedents of trust are identified by Morgan and Hunt (1994). One of the identified antecedents is communication, which Morgan and Hunt (1994) further explain as the alignment of perceptions and sharing of meaningful information between two parties. Across the three cases this relates to the influencers expressing both positive and negative sides of their lives. Thus, the notions of Morgan and Hunt (1994) support the finding that openness in communication is closely related to trust towards the influencer. In order to gain a better understanding of trust building mechanisms, any similar future research should take into consideration the direct effect openness in communication can have on trust, particularly in cases where there is no PSI.

7.3 Reflections on engagement

The respondents’ generally low engagement level presents a noteworthy finding across all three

cases. Due to the screening and sample selection, it was anticipated that the respondents would

engage to a greater extent. According to Muntinga, Moorman and Smit (2011) the motivations for

consuming are entertainment, information and remuneration, where information also includes

inspiration as a sub-motivation. This matches the findings of the interviews, which show that the

respondents are motivated by information, entertainment and inspiration. Further, Muntinga,

Moorman and Smit (2011) state that other motivations than the ones expressed by the respondents

lead to higher levels of engagement, namely personal identity, integration and social interaction and

empowerment. Even though these motivations were not part of the theoretical framework, the open

ended questions in the interview would have allowed the respondents to express these motivations,

yet they are not mentioned by the respondents. The fact that the respondents are not driven by these

motivations could explain why the respondents did not engage on higher levels of engagement

across the cases in this study. However, the reason could also be that consuming is the most common form of online engagement, and that most individuals are so called ‘lurkers’ that only engage passively (Sun, Rau and Ma, 2014). It could be further discussed how these so called

‘lurkers’ participate in the online environment and how this low level of engagement still affects the overall experience in online communities. A possible direction for future research would be to seek greater understanding of ‘lurking’ behaviour. In relation to this, Lee, Chen and Jiang (2006) suggest that ‘lurkers’ can in some sense be considered more active participants by acting as an audience for the individuals engaging on higher levels. Sun, Rau and Ma (2014) further suggest different ‘de-lurking’ strategies that have the possibility to motivate lurkers to engage on higher levels. Future research could thus adopt an experimental design to investigate if influencers who adopt the suggested motivations from Sun, Rau and Ma (2014), could encourage ‘lurkers’ to reach higher levels of engagement and thus exhibit stronger brand loyalty.

In addition to the different levels of engagement, across all three cases the respondents express differences in usage intensity, e.g. how often they engage by consuming or contributing. Therefore, it could be discussed if considering usage intensity could add further understanding to engagement.

Such a relationship is shown in the research conducted by Jahn and Kunz (2012). They in part study the relation between usage intensity and engagement and conclude that intensity has an effect on both engagement and brand loyalty. In this thesis, these insights could explain the rather weak loyalty as due to low engagement and possibly also low intensity. Since the usage intensity is not included in the definition of engagement in the theoretical framework, the findings do not reflect any possible connections between usage intensity and the other variables of the theoretical framework. Thus, for future research purposes the definition of engagement could be changed in order to encompass usage intensity and the effect it has on brand loyalty as well.

Another notable finding of the interviews is a link between information and inspiration. The respondents express that in order to be able to get inspiration from the content of the influencer they also have to receive some information from the influencer’s content. Further, during the data processing it was difficult to make a clear distinction between information and inspiration as a motivation for engagement. An interrelationship between information and inspiration is supported by Muntinga, Moorman and Smit (2011) who define inspiration as a sub-motivation of information.

This could explain the possible interdependency of the two variables found during the interviews.

There are several other notable directions future research could take based on the findings from this study. In relation to engagement, it would be interesting to explore the meaning behind different forms of contributing and if all contribution has the same motivations. The findings in this study indicate that the respondents are more inclined to ‘like’ than to post comments which indicates that there could be different motivations for the same engagement level, in this case contributing. To better understand this, research could study the different mechanics on different platforms and what they mean to users of these platforms as well as how they relate to the different levels of engagement, consuming, contributing and creating.

Future studies on the relationship between influencers and followers could also benefit from the development of a typology of influencers and followers, similar to the research conducted by Wallace et al. (2014b). A possible research topic could thus be to explore different types of influencers and followers in order to make it possible to further specify and relate specific engagement behaviours and motivations to different types of influencers and followers.

7.4 Reflections on brand loyalty

Further, little evidence of attitudinal loyalty was uncovered during the interviews across all three cases. This finding is unanticipated since the theoretical framework suggests that engagement leads to both behavioural and attitudinal loyalty. Van Doorn et al. (2010) suggest that consumer engagement could lead to emotional connections with brands. Yet, few respondents exhibited strong attitudinal loyalty in any of the three cases. This could be explained by Van Doorn et al.

(2010) who cite that the reason the emotional connections take shape is that individuals engage to

reinforce or reflect their social identity. This is similar to self-expression as a motivation for

engagement described by Wallace et al. (2014a). Thus, one explanation for the weak attitudinal

loyalty could be that the respondents rarely exhibit self-expression as a motivation for their

engagement. A rival explanation for the weak attitudinal loyalty could be the lack of PSI that is also

found to have an effect on attitudinal loyalty as illustrated in the theoretical framework. However,

due to the scarcity of PSI, little can be concluded of the possibilities of the relationship between

these variables.

7.5 Managerial implications

The previous section discussed the findings of this study and offered theoretical implications. This following section will offer managerial recommendations and implications based on the findings. It should be noted that the findings by themselves should not be held as absolute facts with remarks to validity and reliability as described in Chapter 5. In order to reach greater validity, the findings in this study would need to be tested through surveys or further case studies in order to achieve generalizability. However, in the following section managerial implications will be presented that could provide strategic guidance in relation to creating and sustaining relationships between followers and influencers based on the findings of this study.

Since followers seem to seek inspiration, information and entertainment on the influencers’ social media platforms it is suggested that influencers create content that is inspirational, informative or entertaining and that they create a balanced mix of these types of content that reflects the interests and lifestyle of the influencer.

In relation to sponsorships it is suggested that the influencers post sponsored content that is in line with their lifestyle and interests since the findings show that followers do not accept sponsored content that is perceived as not fitting the influencer, which further has a negative effect on engagement. A natural fit between the sponsorship and the influencer would further increase trust in the influencer. Equally, for managers it is important to choose a fitting influencer to collaborate with. The negative outcome of partnering with an unfitting influencer is that followers do not seem to engage with posts that feature sponsorships they do not accept. The findings in this thesis also give some indication as to how managers can evaluate the content an influencer posts and thus assist in finding an influencer that has a natural fit with the sponsored product or brand.

Openness in communication is frequently mentioned in this study and serves as an antecedent of PSI, but it is also shown to have a connection to trust and thus acceptance of sponsorship.

Therefore, influencers striving to gain acceptance of sponsorship could benefit from being open in

their communication. It is mentioned in the study that followers perceive openness as sharing both

positive and negative sides of life as well as being transparent about sponsorships. As a suggestion,

influencers could show the process behind sponsorships and offer a more transparent disclosure

than simply stating that a post is sponsored. For example, allowing the follower a ‘behind the

scenes’ view of the sponsorship could benefit both acceptance of sponsorship and trust towards the

influencer. This also includes that influencers should always disclose sponsorships openly, since

hidden disclosures are found to have a negative effect on trust toward the influencer and acceptance

of sponsorship.

In document BEYOND PICTURE-PERFECT BRUNCH PLATES (Sider 105-112)