Primed Authenticity: How Priming Impacts Authenticity Perception of Social Media Influencers

ABSTRACT

Though organizations increasingly collaborate with social media influencers, such as bloggers and videobloggers, little is known as to how the contextual cues related to sponsored content affect the authenticity perception of the social media influencers among audience members. This study explores how positive and negative priming of sponsored content shapes the authenticity perception of the vlogger among its audience members. Four different manipulation conditions were constructed to study a U.S. based travel vlog on Qualtrics, with data collected via Amazon’s Mechanical Turk. A theory-guided content analysis of 211 open viewer responses was conducted to compare perceptions of authenticity between the manipulation groups. The results verified the significance of priming: the same sponsored content can result in opposite reactions among the audience members depending on the positive/negative valence of the introductory text attached, highlighting the central importance of strategic communication related to the perception of sponsored content. The results also point out the importance of audience member engagement for experienced authenticity: The manipulation of audience participation with the vlog had a stronger effect on the perception of authenticity of the vlogger than the positive/negative valence of the introductory text.

Introduction

As traditional forms of advertising are faced with more challenges from free online content, new forms of sponsored content are gaining ground (Liljander, Gummerus, & Söderlund, 2015). One strategy for organizations is sponsored content in collaboration with social media influencers (SMIs), who are a new type of endorser that shape the attitudes of their audiences through social media (Freberg, Graham, McGaughey, & Freberg, 2011). Blogs and video blogs (vlogs) produced by social media influencers are becoming increasingly popular as brand communication channels as a result of their ability to reach mass audiences with similar interests (Uzunoğlu & Misci Kip, 2014). Online content generated by amateurs or non-journalists has been suggested to be or at least is perceived to be, more authentic than professional content due to its “freshness” and “spontaneity” (Tolson, 2010). This is due to the perceived authenticity of ordinary people in comparison to big brands and political elites (Audrezet, Kerviler, & Moulard, in press; Coleman & Moss, 2008; Linqia, 2017; Montgomery, 2001; Scott, 2015).

In the field of strategic communication, collaboration with “audiences” has always been a key focus (Paul, 2011). Recent studies have also highlighted the importance of engaging audience members in strategic decision making (van Ruler, 2018). Audience members use social media for reasons other than receiving brand-related messages, which is a challenge for the influence process of sponsored content (Liljander et al., 2015). Perceived authenticity has been linked to the higher credibility of sponsored content (Kapitan & Silvera, 2016). However, little is known how audience members’ perceptions of authenticity are formed, and how they can be strategically shaped via communication.

To understand audience members’ perceptions of influencer-produced sponsored content from a variety of perspectives, this research combined theoretical insights from several fields, including corporate communication, psychology, media studies, marketing research, and advertising. The topic is important, as previous research in strategic communication has highlighted how a lack of understanding as to how sponsored content works may lead to poor or even unethical conduct by brands and organizations (Ikonen, Luoma-Aho, & Bowen, 2016). In addition, if new forms of sponsored content confuse audience members because of a lack of transparency, it may lead to a lack of trust (Hallahan, 2014; Howe & Teufel, 2014).

This research sought to better understand the audience members who view sponsored content of social media influencers. Specifically, which factors shaped the audience members’ perceptions of sponsored content in the context of vlogs produced by SMIs? It was proposed that priming and strategic communication about the content’s commercial nature could alter the interpretation of the content and its authenticity. Building on the persuasion knowledge model (Friestad & Wright, 1994) and the theory of priming (Carroll & Einwiller, 2014; Celse & Chang, 2017), mere awareness of an intent to persuade may provoke in individuals various coping mechanisms and resistance to communication (Petty & Cacioppo, 1977; van Reijmersdal et al., 2016).

Audience members’ perceptions of sponsored content were studied via an experimental setting of priming before the influencer’s content was presented. It was hypothesized that priming would make a difference in the perception of the endorser’s authenticity based on the manipulation group. The conceptual framework is presented in Figure 1. Self-reported perceptions of vlog viewers were collected for the four manipulation conditions, which primed the viewing with (positive/negative) introductory texts and (encouraged/discouraged) audience participation with the vlog. A survey was conducted online with the Amazon Mechanical Turk (MTurk) crowdsourcing tool during the summer of 2017. A theory-driven content analysis of the four manipulation conditions was conducted from the open-ended responses received from 211 vlog viewers.

Figure 1. Conceptual framework.

The rest of this study is organized in the following manner. First, sponsored content is introduced and previous research on the that topic is presented and discussed. Then, the concepts of priming and perceived authenticity are described, along with how priming shapes perceived authenticity. Next, audience participation and its role on perceived authenticity is discussed. The impact of priming on attitude formation towards the specific elements of sponsored content (i.e., the video itself and the product) is also shown. The methods for the methods for the current study are then reported, followed by the results of the priming experiment. In conclusion, the findings are discussed. Limitations and implications of the research results for strategic communication theory and practice are presented.

Literature review

Sponsored content

Sponsored content is a hybrid form of strategic communication. It is commercial or paid hybrid content appearing in a media context outside the organization’s own platforms (Taiminen, Luoma-Aho, & Tolvanen, 2015). This context can be either journalistic content (such as an online magazine) or user-generated content (such as blog posts, vlog entries, and other posts on social media).

Sponsoring user-generated content has become popular as studies have shown that brands and organizations seldom manage to truly engage audience members on social media by themselves (Kapitan & Silvera, 2016). Also, user-generated content on social media is often considered more credible than company generated advertising (Lee, Lee, & Hansen, 2017). As audiences have come to consider popular social media influencers as trusted opinion leaders (de Veirman, Cauberghe, & Hudders, 2017), many brands and organizations have chosen to engage in sponsored content with influencers in seeking access to audiences online (Kapitan & Silvera, 2016).

At its simplest, sponsored content in the context of influencers means that a brand or organization offers an influencer either money or free products, and in return, the influencer endorses the brand (or product, service, etc.) in content on the influencer’s platform (Liljander et al., 2015). The type of endorsements can range from a single mention (for example, a photo and recommendation on Instagram) to long-term co-operation on several social media channels, depending on the agreement between the influencer and the brand or organization.

The popularity of sponsored content lies in its persuasive power. Audience members can form strong emotional bonds with their favorite influencers. Endorsements from influencers have been found to affect their audience in terms of: purchase behavior (Ho, Chiu, Chen, & Papazafeiropoulou, 2015), brand perception (Lee & Watkins, 2016), brand attitude (Colliander & Dahlén, 2011; Munnukka, Maity, Reinikainen, & Luoma-Aho, 2019), intention to spread “electronic word of mouth” (eWOM) (Evans, Phua, Lim, & Jun, 2017), etc. Brands also report that co-operation with influencers has had a positive effect on the brand’s ranking in search engines and that it also enables feedback from audience members (Uzunoğlu & Misci Kip, 2014).

As sponsored content may also take the form of authentic opinions and experiences of the influencer, it can sometimes be confusing to audience members (van Reijmersdal et al., 2016). An important consideration is, therefore, the transparency of sponsored content (Munnukka et al., 2019). In the U.S., authorities require that the connections between brands and endorsers must always be fully disclosed (U.S. Federal Trade Commission, 2009). The British Competition and Market Authority recently warned several social media influencers for not clearly stating their sponsorships with brands on their social media channels, suggesting that this might be breaking consumer laws (Wakefield, 2019).

However, disclosing sponsorship may lead to less persuasiveness, as the disclosure may produce stronger recognition of the content as advertising (Evans et al., 2017; van Reijmersdal et al., 2016). Yet, studies have produced contradictory results. Liljander et al. (2015) reported young consumers’ responses to suspected covert (concealed) and overt (transparent) blog marketing and found no negative effect on the credibility of the blogger. Colliander and Erlandsson (2015) reported blogger credibility being harmed when the disclosure of the blog being sponsored came from a third party. Despite this, attitudes toward the sponsoring brand were not affected in a similar manner. Hwang and Jeong (2016) reported that the negative effects of disclosure seem to disappear if the influencer emphasizes giving “honest opinions.” It seems that disclosure of sponsorship alone, does not determine the persuasiveness of an endorsement, but that the perceived authenticity of the endorsement also plays an important role in the overall assessment.

Priming of sponsored content

Priming refers to providing contextual cues to a non-conscious information process that activates certain patterns in memory and leaves out others (Carroll & Einwiller, 2014). A tool of strategic communication, priming is like framing, in that it activates audience members’ awareness of an issue. However, framing directs audience members’ judgments by attributing certain meanings to an issue (Wang, 2007). Priming can have an influence on how people think about a topic or it can direct them to use only a part of their knowledge when evaluating a topic (Carroll, 2016).

Priming may be achieved through words (Dolan, Hallsworth, Halpern, King, & Vlaev, 2010). In the context of sponsored content, priming may refer to both intentional and unintentional contextual messages directing audience members on how the content should be received. By selectively directing a person’s attention, the priming guides the interpretation of product information and further influences brand evaluations (Yi, 1990).

Priming requires a target stimulus (i.e., the content) and a prime stimulus (i.e., the primer), which amends the judgment about the target. The prime stimulus makes content, and the cognitive operations used to comprehend or manipulate it, more accessible, which may influence judgments, decisions, and behavior towards the target stimulus as results of information processing (Janiszewski & Wyer, 2014). The manipulation by priming is assumed to cause biased information processing. In the context of the current study, a positive/negative introductory text was used as a prime stimulus, and it was proposed to bias the information processing of the vlog viewers towards the sponsored content resulting in the varied perceived authenticity of the social media influencer (a vlogger in this case).

Perceived authenticity

The rise of the so-called experience economy has increased the desire for authenticity and it’s even been claimed that authenticity is the new business imperative, as people increasingly interpret the world in terms of what is real and what is not (Gilmore & Pine, 2007). The ability to provide an authentic experience is becoming a competitive advantage (Bruhn, Schoenmüller, Schäfer, & Heinrich, 2012). Differentiating oneself from others has been established as a strong contributor to authenticity (Beverland, Lindgreen, & Vink, 2008).

Perceived authenticity has generally been understood as a person (or a brand) being perceived to be true to one’s self (Gilmore & Pine, 2007; Moulard, Garrity, & Hamilton Rice, 2015; Fritz, Schoenmueller, & Bruhn, 2017; Kernis & Goldman, 2006). Perceived authenticity is the perceived uniqueness, originality, and/or genuineness of an object, a person, an organization, or an idea (Molleda, 2010). It is often also defined by negation. Inauthenticity has been understood as falsity (Enli, 2015; Gilmore & Pine, 2007; Kernis & Goldman, 2006), externally motivated behavior and inconsistency (Moulard et al., 2015), and the act of “reciting a script,” rather than expressing one’s own words (Tolson, 2010).

Authenticity is said to promote audience members’ engagement and trust (Enli, 2015), as well as brand attachment and positive word-of-mouth (Morhart, Malär, Guèvremont, Girardin, & Grohmann, 2015). Authenticity has also been linked with commitment and contributing to customer expectations (Tolvanen, Olkkonen, & Luoma-Aho, 2013). It is also considered a key factor when considering the effectiveness of sponsored endorsements (Kapitan & Silvera, 2016).

Studies on marketing and consumer studies often highlight the concept of brand authenticity, defined as “the perceived consistency of a brand’s behavior that reflects its core values and norms, according to which it is perceived as being true to itself” (Fritz et al., 2017, p. 327). Brand authenticity is seen as consisting of continuity, credibility, integrity, and symbolism (Morhart et al., 2015), and reflecting consumer perceptions (Dwivedi & McDonald, 2018), which can also be managed (Moulard, Raggio, & Garretson Folse, 2016). It is notable, particularly in the context of the current study, that people can also be brands (Thomson, 2006). Like traditional brands, the authenticity of human brands (i.e., celebrities) represent consumer perceptions of them being true to themselves in their relationships with consumers (Ilicic & Webster, 2016; Moulard et al., 2015).

Priming of sponsored content – shaping perceived authenticity

Vloggers have been recognized as new “authentic online celebrities” (Morris & Anderson, 2015). Their authenticity is said to be defined by their directness, transparent amateurishness, and the conversational nature of their content (Tolson, 2010). Vloggers often use self-disclosure and talk directly to the camera; both methods have been found to be positively associated with authenticity (Ferchaud, Grzeslo, Orme, & LaGroue, 2018).

Engaging in sponsored content can put the perceived authenticity of a vlogger under threat (Colucci & Cho, 2014), as perceived opportunism can be a signal of “selling out” (i.e., giving up integrity or compromising principles for personal gain) (Thomson, 2006). Therefore, it is believed that negative priming with an introductory text that emphasizes “selling out” and the opportunism of the vlogger will reduce his/her authenticity perception with the audience. Similarly, positive priming with an introductory text emphasizing the credibility and altruistic attributes of the vlogger will enhance his/her authenticity perception with the audience. Thus, it was proposed:

H_1a:

Priming with a positive introductory text about the sponsored content will enhance the perceived authenticity of the social media influencer.

H_1b:

Priming with a negative introductory text about the sponsored content will reduce the perceived authenticity of the social media influencer.

Audience participation – shaping perceived authenticity

Audience participation is defined as various reactions to content, such as likes, dislikes, comments, shares, and subscriptions. From the viewpoint of a brand, audience members’ participation is seen as positive and beneficial in and of itself, because it promotes a positive brand attitude and customer engagement (Kujur & Singh, 2017). However, the current research focuses on the effects of encouraging and discouraging audience members to participate with the vlog, based on the perceived authenticity of the vlogger.

The possibility to comment, share, subscribe, like, and dislike a vlog post enables two-way communication and therefore, supports interaction with the content (Hayes & Carr, 2015). By being allowed to directly contact the vlogger through comments, viewers may freely express their suspicions over the possible commercial claims of the vlogger. However, the negative consequences of suspected deception might even be diminished by the possibility of interaction (Liljander et al., 2015). According to Yang and Lim (2009), the interactivity of a blog increases relational trust and mediates blogger credibility. According to Tolson (2010), the conversational nature of blogs, as they provide the possibility to comment on the content, supports their authenticity.

Utilizing source credibility and warranting theory, Hayes and Carr (2015) studied the connection between enabled comments (in a blog) and brand attitudes. Warranting theory addresses how cues to validate the self-presentation of others are used in online encounters (Walther & Parks, 2002). The cues or “warrants” provided by sources other than the target have a higher warranting value. A blog enabling comments and providing a higher degree of interactivity provides more warrants to its readers, as the information it provides is possible to be confirmed or questioned through interaction (Hayes & Carr, 2015). According to this theory, encouraging/discouraging audience members to participate should increase/decrease the perceptions of authenticity towards the vloggers among the viewers. Thus, it is proposed:

H_2a:

Encouraging the audience members to participate with the vlog will increase their perceived authenticity towards the vlogger.

H_2b:

Discouraging the audience members to participate with the vlog will decrease their perceived authenticity towards the vlogger.

Priming of sponsored content – shaping attitudes toward the video and the product

As discussed previously, priming activates audience awareness of an issue and can have an influence on how people think about a topic (Carroll, 2016). If priming an audience with a positive/negative introductory text regarding the sponsored content can influence the way they process it in a predictable way. It is possible for such priming to help in the formation of positive/negative attitude towards the specific elements of the sponsored content (such as the video itself and the products). It is believed that such an altered view of the sponsored content or the source as a result of priming is rooted in the literature of source credibility.

Source credibility is a widely studied concept. The tradition of the concept lies in Aristotle’s ethos, which referred to the intelligence, character, and goodwill of the communicator (McCroskey & Teven, 1999). The concept of source credibility, or the credibility of the sponsored content, is commonly understood as consisting of several dimensions. These include: attractiveness of the sponsored content, expertise reflected in the sponsored content, trustworthiness (Chu & Kamal, 2008; Goldsmith, Lafferty, & Newell, 2000; Ohanian, 1990), similarity (Munnukka et al., 2019; Munnukka, Uusitalo, & Toivonen, 2016), competence (Liljander et al., 2015), the quality of the message, and when it comes to endorsements, a good endorser-product fit (Kapitan & Silvera, 2015).

When considering endorsements, the perception of credibility of the sponsored content (which can be controlled via priming of the introductory texts) leads to positive attitudes about the endorsement (Goldsmith et al., 2000; Pornpitakpan, 2004) and enhances the likelihood of message acceptance (Ajzen & Fishbein, 1977; Kapitan & Silvera, 2015). Therefore, the credibility perception of the sponsored content can be considered an integral part of endorsements (Munnukka et al., 2019) and can shape the attitude formation towards the specific elements of the sponsored content (i.e., the video and the products). Therefore, based on the literature, the goals of the study were simplified in the following hypotheses:

H_3a:

Priming with a positive introductory text about the sponsored content will lead the audience to perceive the sponsored content more positively (i.e., positive attitude towards the video and product), and

H_3b:

Priming with a negative introductory text about the sponsored content will lead the audience to perceive the sponsored content more negatively (i.e., negative attitude towards the video and product).

Data and analysis

A U.S. based travel vlog was chosen for the study. A 2 × 2 between-subjects true experimental design was used. The independent variables were type of introductory text (positive or negative) and audience participation (encouraged or discouraged); the combination of the two resulted in four experimental conditions which were constructed using Qualtrics. The data was gathered with the Amazon Mechanical Turk (MTurk) crowdsourcing tool.

In total, 211 valid responses were returned. It was assumed that using an online crowdsourcing tool for data collection would draw a sample that was not dissimilar from the typical vlog follower (e.g., technically engaged, open to new uses of technology, etc.). The respondents were in the age range 15–60 years, the median age was in the range 26–30, and 79% of respondents were 35 years old or younger. The respondents were 59% male and 41% female. Four experimental conditions were created to implement the independent variables of priming and audience participation and test their impact on vlogger authenticity. The subjects were randomly assigned to a condition to improve the internal validity of the experiment.

After a careful shortlisting of potential candidates, a vlogger with around 200,000 subscribers was selected for the present study. The casting of the vlog was viewed and informally assessed to ensure that only one vlogger was featured rather than multiple individuals. The focus of the vlog content, the genre, and the type of products it displayed were also examined. Although the effects being tested were not expected to be dependent on any of these variables, care was taken that the vlog appeared “typical,” and would not be considered an outlier in terms of being a candidate for sponsorship, or displayed extreme characteristics in terms of authenticity or inauthenticity. The vlog material was publicly available, and the vlogger was not connected with the study, nor compensated, in any way.

The experimental conditions, instructions, and the dependent measures were carefully designed and pretested with 30 respondents prior to the actual data collection. This helped to confirm the efficacy of the manipulations, and the appropriateness of the experimental conditions, and instructions. A few minor changes were made before the actual data collection. The respondents received compensation in exchange for their participation in the study, according to MTurk’s suggested compensation rates.

The respondents first read a short priming introductory text (see Appendix 1–3) and then watched the roughly three-and-a-half-minute video, in which the vlogger presented three travel-related products: a lock, soap sheets, and travel containers. Next, the respondents were asked to take a short qualifying quiz to ensure that they had watched the video carefully. Following this, the respondents were asked to provide an open response of 150 words, sharing their thoughts and feelings about the video, the vlogger, the products and brands seen on the video, and the viewing experience.

The negative priming text claimed that the respondents were about to see a video that would be: “overly sponsored,” “done only for money,” “pushing the products,” and “a waste of your time.” The positive priming claimed that the video would be: “credible,” “unbiased,” “of good quality,” and “worthy of your time.” As a prime, the negative introductory text was expected to activate knowledge of intent to persuade in the respondents and, therefore, would provoke coping strategies (van Reijmersdal et al., 2016).

To manipulate participation, respondents were either “strongly” encouraged or discouraged to share, comment, like, dislike, and subscribe to the video (Appendix 2). Figure 2 displays the four experimental conditions.

Figure 2. Groups and manipulation conditions.

The data included 211 open-ended responses in English, each was 150 words in length. A content analysis of themes was conducted to answer the research questions (Eriksson & Kovalainen, 2016). The responses were systematically coded using a scheme based on theory in order to extract the respondents’ perceptions of the endorser’s authenticity and/or their perceptions of the inauthenticity of the endorser.

The cues for an authentic endorser included expressions such as genuine, unbiased, true to herself, and sincere. Whereas, expressions such as artificial, fake, not genuine, and putting on an act implied inauthenticity in the endorser. If the endorser was described in some other way that did not fit the categories of either authentic or inauthentic, for the purposes of this study, the endorser was coded as neutral (see Table 1). Also, in this study, the codes endorser authentic, endorser inauthentic, and endorser neutral were defined as mutually exclusive.

Table 1. Examples of coding of the open-ended responses.

Verbatim	Code
“She was genuine and charismatic.”	endorser authentic
“Her upbeat optimism seems fake and unreal.”	endorser inauthentic
“She doesn’t come across as entirely authentic.”
“The endorser is a little excited for my taste but not so much that I would avoid her vlogs.”	endorser neutral

Results

The Pearson Chi-squared test was conducted to examine the perceived authenticity differences among the manipulation groups. The results showed statistically significant differences in endorser authenticity perceptions between the manipulation groups (p = .002). Group 1 (audience-participation-encouraged, introductory-text-positive) was the most positive and least negative about the video and the products. However, regarding perceived authenticity, Group 1 scored only the second highest, although it had the lowest score for inauthenticity. Group 4 (audience-participation-discouraged, introductory-text-negative) was the most distrustful and negative group with regard to the authenticity perception, as well as in their attitudes towards the video and the products. It scored the highest in all the negative aspects and lowest in all the positive aspects. However, this was expected, since Group 4 was primed by the most negative manipulation condition, a negative introductory text, and discouraged participation. Overall, the results confirmed that priming was strongly associated with the authenticity perception of the vlog viewers in both the positive and negative priming circumstances, confirming H1a and H1b. The results are summarized in Table 2.

Table 2. Summary of the results of the influencer content perceptions.

		Positive text introductory text		Negative introductory text
		Participation (encouraged)	Participation (discouraged)	Participation (encouraged)	Participation (discouraged)
		Group 1	Group 2	Group 3	Group 4	Total
Source authenticity*
Authentic	Count	14	12	13	4	43
	% within group	27%	22%	28%	7%	21%
Inauthentic	Count	9	11	18	25	63
	% within group	17%	20%	39%	45%	30%
No mention/Neutral	Count	29	31	15	26	101
	% within group	56%	57%	33%	47%	49%
Total	Count	52	54	46	55	207
Attitudes towards the video
Video positive**	Count	45	42	28	24	139
	% within group	87%	81%	65%	48%	71%
Video negative***	Count	6	9	15	20	50
	% within group	12%	17%	35%	40%	25%
Video neutral	Count	2	3	2	6	13
	% within group	4%	6%	5%	12%	7%
Total	Count	52	52	43	50	197
Attitudes towards the products
Products positive****	Count	40	42	27	26	135
	% within group	89%	81%	69%	57%	74%
Products negative*****	Count	4	8	5	12	29
	% within group	9%	15%	13%	26%	16%
Products neutral	Count	1	3	6	7	17
	% within group	2%	6%	15%	15%	9%
Total	Count	45	52	39	46	182

* X² = 20.865, p-value 0.002

** X² = 21.863, p-value 0.000

*** X² = 14.743, p-value 0.002

**** X² = 14.354. p-value 0.002

***** X² = 5.412, p-value 0.144 (No statistical significance)

Looking at the authenticity perception scores for the four experimental conditions (see Table 1), the endorser was perceived most authentic by Group 3 with a share of 28%. The second highest share of authenticity perception (27%) was found in Group 1, the second lowest in Group 2 (22%), and the lowest in Group 4 (7%). Whereas, inauthenticity perception was highest in the most negatively manipulated Group 4 (45%) and lowest in the most positively manipulated Group 1 (17%). Based on these results, the manipulation of participation had a significant effect on the perception of authenticity only in the negative introduction text conditions. It did not have such an impact in the positive situations. On the contrary, the negative valence of the introductory text seems to have had a profound impact on inauthenticity perception as shown in Group 4 (45%) and Group 3 (39%). The results partially support H2a, suggesting that participation is significantly associated with the perceived authenticity of the vlogger only in the negative priming circumstance. Hypothesis 2b was not supported, although the group differences appeared to follow a similar pattern as in the case of H2a, but the group differences were not statistically significant.

In the case of endorser neutrality, the requirements of the Pearson Chi-squared test were not met, since four cells (50%) had an expected count of less than 5, while the test requires each expected cell frequency to have a sample size of at least 5 (Hair, Wolfinbarger, Money, Samouel, & Page, 2015). The endorser was mostly perceived neutral in Group 1 with a share of 56%.

The Pearson Chi-squared test showed statistically significant differences for both positive (p < .001) and negative (p < .01) attitudes towards the video among the manipulation groups. The Chi-squared test requirements were not met in the case of neutral attitudes towards the video due to a low cell frequency (count of less than five) as explained previous. As presented in Table 1, the attitudes were the most positive in Group 1 with a share of 87%, followed by Group 2 with 81%. The lowest share of positive attitudes towards the video was in Group 4 (48%), and the second lowest in Group 3 (65%). The share of negative attitudes towards the video was highest in Group 4 (40%), followed by Group 3 (35%), then Group 2 (17%), and the lowest for Group 1 (12%), indicating a strong priming effect of the positive and negative introductory texts.

The video was mostly found neutral by Group 4 with a share of 12%. In total, there was a neutral attitude towards the video in 7% of all the responses mentioning the video.

Based on the results regarding the video, the manipulation effect again worked as predicted. The combination of a positive introductory text and encouraging participation produced the most positive results the combination of a negative introductory text and discouraging participation produced the most negative results. The mixed conditions (positive introduction + discouragement and negative introduction + encouragement) produced results that were in between the extremes, the valence of the introductory text being the dominating factor.

The Chi-squared test produced statistically significant differences in positive attitudes (p < .01) towards the products among the manipulation groups. However, in the case of negative attitudes toward the products, the Chi-squared test results yielded no statistical significance (p > .05). Hence, no differences in negative attitudes towards the products between the manipulation groups were detected. The Chi-squared test requirements were not met in the case of neutral attitudes towards the products due to low cell frequencies (counts less than 5) as discussed previously (Hair et al., 2015).

As presented in Table 1, the share of positive attitudes towards the products was highest in Group 1 (89%). The second highest share of positive attitudes was found in Group 2 (81%). The groups primed with a negative introductory text had the lowest shares of positive attitudes towards the products, Group 3 (69%) and Group 4 (57%). These results suggest that the priming manipulation operated as hypothesized, since the positive introductory text produced the most positive results and the negative introductory text the most negative results, while encouragement for participation increased positivity towards the products. As the results were statistically significant, it can be stated that the negative manipulation resulted in less positive attitudes towards the products. Hence, H3a was confirmed, while H3b was only partly confirmed.

The share of negative attitudes towards the products was highest in Group 4 (26%) and second highest in Group 2 (15%). The share of negative attitudes towards the products was lowest in Group 1 (9%) and second lowest in Group 3 (13%). Though no differences were found in the negative manipulations, it can be concluded that the manipulation did not influence the perception of the endorsed products. In other words, negativity was not projected on the products as expected.

The tested results suggest that priming shapes the authenticity perception formed by the vlog viewers, and the hypotheses of a clear difference between groups were confirmed. The Chi-squared test indicated statistically significant differences of endorser authenticity perceptions among the manipulation groups (p < .01).

Discussion

The present study demonstrates that priming shapes the authenticity and source credibility perceptions of the vlog audience cf. (Kapitan & Silvera, 2016). It also supports highlighting the role of strategic communication, and especially, the calls to action in the context of sponsored content. The analyses revealed significant differences among the manipulation groups concerning the variables of authenticity and source credibility perceptions.

The endorser was most often perceived as authentic in groups that encouraged to participate, regardless of the nature of the priming (positive or negative). The negative priming group that was encouraged to participate also had the second highest inauthenticity perception and the lowest share of non-responses. Audience participation was found to have an important role in priming effectiveness. Participation increased the positive authenticity perceptions in the negative priming case but had no such effect in the opposite case. This suggests that the priming may have activated a stepwise process of persuasion. It is suggested that the disclosure of a persuasion attempt first triggers cognitive persuasion knowledge that leads to activation of attitudinal knowledge. However, as vlogs have been traditionally perceived as a non-traditional advertising outlet (e.g., Wojdynski & Evans, 2016), the audience may be uncertain about the persuasion intentions of vlog product presentations (as in the video of the current study). Therefore, participation with the video channel or outlet allows more careful information processing, leading to either confirmation or rejection of the priming (i.e., the persuasion attempt). In the current study, if the video was perceived as positive and credible, higher participation will result in a rejection of the negative priming attempt, and thus, to improved authenticity perceptions. This supports the important role of audience participation in the effectiveness of priming in social media.

The priming effectiveness was further verified by the finding that the negative introductory text was a dominant factor. The vlogger was perceived as least authentic among those whose perceptions were negatively primed and discouraged from participating. Thus, the cognitive persuasion knowledge of this group was activated. Positive priming and encouragement to participate led to an opposite outcome. The results also support the subjectivity of authenticity perceptions (Gilpin, Palazzolo, & Brody, 2010; Henderson & Bowley, 2010) and verify that they are prone to priming. Vlog participation was found to act as a type of coping strategy in response to the priming message by allowing the audience to process and assess it, and thus, the vlogger’s authenticity. Therefore, participation may act as a means of mitigating the negative effects of persuasion knowledge if other cues support the vlogger’s authenticity.

This study has several implications. As the same sponsored content can result in opposite reactions among the audience depending on the positive/negative valence of an introductory text, who recommends the content and what they say becomes of central strategic importance when understanding the effectiveness of sponsored content. The managerial implications for strategic communication call for strategic planning of how sponsored content is primed, failure to introduce it properly may lead to less success. Sponsored content effectiveness depends on the context in which it is introduced; positive, encouraging messages can support success (e.g., Munnukka et al., 2019). On the other hand, overly positive or overly negative messages may harm the perceived authenticity of the content. This calls for organizations to be more responsible in their communication, and emphasizes the strategic value of understanding the audiences and stakeholders well. As for the theoretical implications, the findings seem to confirm the increased importance communication plays in engagement (Johnston & Taylor, 2018). Specifically, the findings highlight that the behavioral aspects of engagement help the individual processing of sponsored content. Similar results have recently been demonstrated in work place communication and social media use, where higher engagement was a result of the actions of individuals (van Zoonen & Banghart, 2018).

Conclusions

The current study highlighted the role of priming in strategic communication. Priming was found to be capable of influencing the authenticity perceptions an audience has of a vlogger, and thus, endorsement effectiveness. Positive priming increased the authenticity perception, and negative priming decreased it. Encouraging participation was found to hinder the priming effect in that the priming message was in conflict with the perceived reality. The findings also showed that positive priming influences both the vlogger’s authenticity as well as attitudes towards the endorsed video and product, while the effect of negative priming was restricted to only the vlogger. The results suggest that sponsored content is less risky for organizations than influencers: the vloggers will receive a much greater share of negativity when content is perceived as inauthentic. Moreover, encouraging people to interact may improve their ability to process the content, highlighting the benefits of viewer and stakeholder engagement.

The present study does have some limitations (namely, the setting). These limitations also suggest opportunities for future research. First, the research design involved only one vlog and one video. Although the literature suggests the studied effects will be widespread, the findings may be specific to this vlog, vlogger, and the presented product category. Thus, future studies should explore authenticity formation and priming effects in a wider variety of contexts. Second, the present study examined authenticity and priming in the case of a positive brand endorsement only. Priming effectiveness may not be symmetrical in the case of negative or neutral brand media coverage. Therefore, future studies should concentrate on examining the results in various endorsement situations. Finally, this research was conducted in the U.S. The findings of the present study could be validated internationally or in other settings, and there may be other important factors that contribute to perceptions of authenticity.

Supplemental Material

Supplemental data for this article can be accessed on the publisher’s website.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

This work was supported by industry partners and a research grant from Business Finland for the research project "Opening the Black Box of Content Marketing".