• No results found

Influencer versus brand marketing on instagram : which strategy is the most effective in a transparent environment?

N/A
N/A
Protected

Academic year: 2021

Share "Influencer versus brand marketing on instagram : which strategy is the most effective in a transparent environment?"

Copied!
36
0
0

Bezig met laden.... (Bekijk nu de volledige tekst)

Hele tekst

(1)

Influencer versus brand marketing on Instagram

Which strategy is the most effective in a transparent environment?

Phylicia Baars 10092226 Master thesis Graduate School of Communication

MA Persuasive Communication Supervisor: M. L. Fransen Date: 2-02-‘18

(2)

Abstract

The non-obligatory disclosing of sponsored content that is embedded in influencer posts on Instagram raises ethical questions on advertising transparency. Meanwhile advertising codes lobby for a transparent advertising future. This study compares and evaluates which strategy, either brand or influencer marketing, would work best when both strategies are disclosed. This is investigated through the process of

trustworthiness and perceived intrusiveness. Mediation analysis results show that there is no difference in source in a transparent Instagram environment, and neither

trustworthiness nor intrusiveness play a mediating role. This can partially be explained by the discrepancy found in the manipulation check but also through the

correspondence bias and the persuasion knowledge model. On the basis of these findings, implications and recommendations for future research are given.

(3)

Brands and companies are increasingly using the social media platform Instagram for advertising purposes. Business features generated by Instagram make it attractive for marketers to effectively find their target groups. However, targeting has become more difficult since Instagram altered its algorithms to ensure that their users are exposed to less advertising and more to what they prefer to see (Instagram, 2016). Marketers therefore search for alternative marketing strategies to still expose consumers to their brand messages and products. Influencer marketing is one of those alternative

strategies. An influencer, a micro-internet celebrity with high source credibility and established popularity, promotes or endorses a brand or product through their social media channel in exchange for money (De Veirman, Cauberghe, & Hudders, 2017). Influencer marketing appears to be more effective than traditional advertising due to high authenticity and credibility of the influencer, which subsequently leads to lower resistance to the message (De Vries, Gensler, & Leeflang, 2012). This form of advertising however, is not transparent towards the consumer. Most consumers are unaware of the persuasive intent of the influencer posts (Boerman, Willemsen, & Van der Aa, 2017) because influencers are not obligated to display sponsored activities. This raises concerns about the ethics of this strategy. As an effort to increase

advertising transparency on their platform, Instagram gave influencers the opportunity to officially add a sponsorship disclosure to their posts starting from October of 2017. A sponsorship disclosure explicitly informs an audience when commercial content is integrated into editorial content to guarantee fair communication and avoid persuasion without audience awareness (Friestad & Wright 1994; Cain, 2011; Woods, 2008). Brands and company accounts were already obligated to disclose their posts if it contained advertising content, but now influencers are also given this option. With a

(4)

codes committees in the Netherlands who are lobbying for a transparent advertising future (Stichting Reclame Code, 2018), the accompanying disclosure that is now still optional is not far from becoming compulsory. The challenge for advertisers thus becomes to distinguish which type of strategy, either brand marketing or influencer marketing, is more effective when trying to achieve the highest purchase intentions and most positive brand attitudes.

An important construct that attributes to a positive brand attitude and a high purchase intention is trustworthiness. Consumers perceive influencers as highly trustworthy sources of information due to celebrity endorsement and electronic word-of-mouth effects (Djafarova & Rushwort, 2017; De Vries et al., 2012). Brands on the other hand have always been perceived as distrustful by consumers. However, the implementation of the sponsorship disclosure to influencer posts might alter this perception. Once consumers notice that the influencers are motivated by money, consumer will feel violated in their trust. This is why it is expected that once all advertising attempts on Instagram become transparent this will have a backfiring effect for influencers.

Another important construct that attributes to brand attitude and purchase intention is intrusiveness. Intrusiveness, in contrast to trustworthiness, has a negative effect on brand attitude and purchase intention. Since social media platforms were initially developed for people to interact, brands are perceived as intrusive when they enter these platforms (Fournier & Avery, 2011). Yet, the sponsorship disclosure might also cause consumers to consider the influencer post as even more intrusive compared to the brand post.

The expected effects of advertising transparency might alter the perception and future of influencer marketing. It is therefore of additional importance to study

(5)

complete transparency in the Instagram environment, especially since Instagram is the second most used platform to facilitate influencer marketing (Wheeler, 2017) and advertising transparency is not far from becoming reality. Hence, it is aimed to provide more insights in the differences between influencer and brand posts on Instagram when all commercial intents are transparent, and how this affects consumer behaviour and attitudes through trustworthiness and intrusiveness. The research question for this study is:

RQ1: To what extent can we distinguish a difference between influencer and brand posts on brand attitude and purchase intention taking into account perceived

intrusiveness and trustworthiness?

This research contributes to the understanding of consumer behaviour and attitudes on Instagram through influencer and brand marketing. The outcomes of this study will not only provide academic insights but it will also yield valuable information for professionals working in online advertising. Professionals can then, hopefully, make more constructive decisions into which marketing strategy will be more effective in a transparent environment.

Theoretical Background

Instagram

The social media platform Instagram is a mobile, desktop and internet-based photo-sharing application and service that focuses on visual storytelling (Instagram, About us, 2018). It was founded in 2010 and is now the seventh most used social network site worldwide after Facebook, Whatsapp, Facebook Messenger, QQ, We Chat and Qzone (Chaffey, 2017). In September 2017 it had over 800 million active users

(6)

million make use of the platform on a daily basis (Chaffey, 2017; Balakrishnan & Boorstin, 2017). In the eight years of its existence it has grown into one of the big players amongst social network sites. The tremendous growth and popularity of the platform has increasingly gained interest of marketers to use it for advertising purposes. Marketing professionals are especially enthused about the impact and revenue Instagram-advertising generates (Forer, 2017; Morrison, 2017). There are two popular strategies for marketers to engage in Instagram advertising: brand marketing and influencer marketing.

Brand Marketing on Instagram

By the end of 2016, Instagram designed special profiles for brands and businesses. Before, brand profiles were identical to consumer profiles. Additional features were implemented that distinguish a company profile from a normal profile. One of these features is the direct link button. This button allows consumers to directly connect with the brand. Another, more important, feature is ‘Instagram statistics’. This

software implementation allows businesses to obtain a detailed insight in the location, interests, demographics and behaviours of the accounts’ followers and visitors. It not only offers brands detailed statistics of their target group but it also allows tracking of the Instagram advertisements’ marketing achievements. By offering these business features Instagram has proven itself to be an accessible and valuable tool and platform for brands to engage in Instagram marketing. This explains why there is such a large increase of marketing and advertising practices on Instagram (Chaffey, 2017). Today, there are over 25 million Instagram business accounts worldwide (Instagram

(7)

When brands engage in brand marketing on Instagram they usually use a promotional push strategy. This strategy entails “pushing” a product or brand through the channels of distribution, thus aggressively selling and promoting the item (Belch & Belch, 2015). On Instagram this means that brands create advertisements, that take the form of Instagram posts, and advertises these posts on the platform. Instagram saw this immense increase of advertising activity and as a measure altered their algorithm to ensure that their users were exposed to less advertisements (Instagram, 2016). This made it harder for brands to proliferate on the platform. Moreover, Simmons (2008) indicates that firms start to realize that push advertising online will not work as effectively as an online pulling strategy. A promotional pull strategy entails focussing on the ultimate consumer and creating a demand at the end-user level (Belch & Belch, 2015). Simmons (2008) explains that brands should “pull” consumers into brand narratives that appeal them with their online social networks or within developed brand communities. Thus, brands on Instagram started looking for a creative solution in alternative pulling strategies. One of those solutions is influencer marketing.

Influencer Marketing on Instagram

Influencer marketing is a marketing strategy in which an influencer promotes or endorses a brand or product through their social media channel in exchange for money (De Veirman et al., 2017). Social media influencers are referred to as people who have built a sizeable social network of people following them on a, or several, social media platform(s). They can also be perceived as micro-celebrities on the Internet and are seen as a trusted tastemaker in one or several niches (De Veirman et al., 2017). Influencer marketing offers the opportunity to reach target audiences in a

(8)

explained through celebrity endorser- and electronic word-of mouth (eWOM) effects. When a celebrity or in this case an influencer is employed by a brand to spread their message or to endorse their products they become brand ambassadors. Erdogan & Baker (2000) state that when celebrities endorse a product, the meaning developed around the particular celebrity will transfer to the endorsed product, brand or company. This mechanism is also referred to as the Meaning Transfer Model (McCracken, 1989). Celebrity endorsement effects have proven to positively influence brand attitude and purchase intentions (Bergkvist, Hjalmarson, & Magi, 2016). Influencers also generate positive word-of-mouth for the brands when they endorse. Influencers are known for maintaining good and personal contact with their followers (De Veirman et al., 2017). This form of interaction can be perceived as eWOM and it refers to any product information communicated by potential

consumers via the Internet (Daugherty & Hoffman, 2013). EWOM is considered an influential and powerful factor impacting consumer decision making (Daugherty & Hoffman, 2013; Erkan & Evans, 2016). EWOM and celebrity endorsement effects make influencer marketing a very effective and lucrative pulling strategy. In a study by Tomoson (Morrison, 2017) 59% of the marketers interviewed indicated to increase their influencer marketing budget in the next 12 months because influencer marketing offers big revenue: of the average $1 spent, influencer marketing generates a $6 return.

Advertising transparency on Instagram

Wherever there is advertising, there are advertising codes. Advertising codes exist to guarantee fair communication and avoid persuasion without audience awareness. This is also called advertising transparency. Sponsorship disclosures contribute to

(9)

advertising transparency. These disclosures inform an audience when commercial content is integrated into editorial content (Friestad & Wright 1994; Cain, 2011; Woods, 2008). The increase of advertising practices on Instagram is paralleled by the increase of unfair communication concerns by advertising code committees.

Instagram responded to this increased pressure in October 2017. Where brands were already obligated to disclose their advertisements, for influencers it has become optional. The current sponsorship disclosure of brand advertisements states: ‘Sponsored’. For influencer posts this is: ‘Paid partnership with [brand name]’. However, this non-obligatory disclosing of influencer posts is not transparent to the consumer because not all influencers consistently use it or even use it at all. The concern for unfair communication is not ceased. It is therefore expected that all practices of influencer marketing will have to be disclosed in the near-future.

To effectively study the differences between influencer and brand marketing it is important to look at consumers’ attitudes and behaviours. The interesting, and most used, variables to take into consideration are: brand attitude and purchase intention. Brand attitude refers to consumers’ attitudes as the sum of the products of beliefs multiplied by the evaluations (Wang & Scheinbaum, 2017) or the overall evaluations of the brand from the consumers’ perspective (Spears & Singh 2004). Purchase intention is the overall intention to buy or the willingness to buy a product (Spears & Singh, 2004). Brand attitude and purchase intention are two concepts that have proven to be important for marketers because they are useful in predicting consumer

behaviour (Mitchell & Olson,1982).

An important indicator of why people (intend to) make purchases and what their attitude is towards a brand is the trust people have in the source that transmits

(10)

the message (Ohanian, 1990). The following section will elaborate on the trustworthiness of both brands and influencers on social media.

Trustworthiness

Trustworthiness represents the degree of confidence consumers place in a

communicator’s intent to convey the assertions he or she considers most valid (Amos, Holmes, & Strutton, 2008) and is one of the three constructs of the Source Credibility theory by Ohanian (1990). Trustworthiness of the communicator or source is an important construct in persuasion and attitude-change (Ohanian, 1990). Brands and influencers however are not equally perceived as trustworthy.

Brands generate several forms of brand communication of which advertising is the most common form. Advertisements are designed to persuade consumers (Fransen, Verlegh, Kirmani, & Smit, 2015). It is usually made to create brand or product awareness or to communicate social, emotional or functional product benefits (Fransen et al., 2015). Consumers assume brands operate with a persuasive intention to either sell their product or generate brand awareness in the brands’ favour. This pre-assumption makes consumers distrustful towards brands on social media (Hunt & McLaren, 2015). Research by Boerman et al. (2017) shows that the presence of brands online activates persuasion knowledge in consumers. Persuasion knowledge refers to consumers’ knowledge and beliefs about various advertising related issues (Friestad & Wright, 1994). This means that once consumers notice brands online they immediately assume the persuasive motives of the brand. This causes consumers to perceive brands as low in trustworthiness.

Trustworthiness of influencers is, as previously discussed, one of the main reasons why influencer marketing in a non-transparent environment works. Generally,

(11)

influencers are perceived as more trustworthy (Erkan & Evans, 2016) opposed to brands. However, research by Bergkvist et al. (2016) states that if a celebrity endorser is seen as being motivated by money, by for example a sponsorship disclosure, this will affect consumers’ brand evaluations. This phenomenon is also known as the correspondence bias and comes from the attribution theory (Gilbert & Malone, 1995). This bias means that people make causal inferences to explain other people’s

behaviours; they develop a judgement based on assumptions of a person’s feelings or beliefs and perceive this as true (Boerman et al., 2017; Cronley, Kardes, & Houghton, 1999). In other words, Instagram users make assumptions of the motivations of influencers and attribute the endorsement to the influencers’ genuine liking of a product or brand. A recent study by Boerman et al. (2017) indicates that a sponsorship disclosure to a celebrity endorsed posts on Facebook increased critical and distrusting beliefs. The sponsorship disclosure activates critical thinking in consumers and indicates that the influencer has an ulterior motive. The disclosure affects the trust that consumers have in influencers and even causes a backfiring effect. Since

consumers will feel violated in their trust due to the sponsorship disclosure they will show a stronger resistance towards the message. It is therefore expected that

influencer posts generate a lower level of trustworthiness compared to brand posts. The hypothesis taking into account the literature discussed is as follows:

H1: Influencer posts generate a lower level of trustworthiness compared to brand posts.

(12)

Research by Wang and Scheinbaum (2017) has indicated that when consumers perceive the source as highly trustworthy this will result in positive brand evaluations and high purchase intentions. The literature discussed indicates that influencer posts generally have more positive outcomes on brand attitude and purchase intention through the conjunction of trustworthiness compared to brand posts. However, as previously mentioned, advertising transparency will negatively affect the degree of trustworthiness, especially for influencers, and thus it will also subsequently

negatively affect brand attitude and purchase intention. Research by Dekker and Van Reijmersdal (2013) on disclosures in in-program endorsement, states that viewers who did not find the celebrity credible were significantly less persuaded by the endorsement when confronted with the disclosure of persuasive intent and deception. This means that, since trustworthiness is a construct of source credibility, the lower perceived trustworthiness in influencers subsequently causes the consumers to be less favourable towards the brand. It is therefore expected that influencer posts generate more negative brand attitudes and lower purchase intentions through the conjunction of trustworthiness compared to brand posts.

H2: Influencer posts generate more negative brand attitudes compared to brand posts

H3: Influencer posts generate lower purchase intentions compared to brand posts.

H4: Influencer posts generate more negative brand attitudes because of trustworthiness compared to brand posts.

H5: Influencer posts generate lower purchase intentions because of trustworthiness compared to brand posts.

(13)

Once consumers are confronted with sponsorship disclosures, their persuasion knowledge is activated (Boerman, van Reijmersdal, Neijens, 2012) and the post is perceived as advertising content (Boerman et al., 2017: Friestad & Wright, 1994) and consequently will generate a sense of intrusiveness. The next section will elaborate on the extent to which the sources are perceived as intrusive.

Perceived Intrusiveness

Several academics have suggested that consumer perceptions of digital advertising have become increasingly negative (Truong & Simmons, 2010). One of the concepts that attributes to this increased negative perception is perceived intrusiveness. Ha (1996) defines intrusiveness as “the degree to which advertisements in a media vehicle interrupt the flow of an editorial unit”. Or as Edwards, Li and Lee (2002) define it: “the perception or psychological consequence that occurs when an audience’s cognitive processes are interrupted”.

Social network sites were developed for online social interaction amongst people (Fournier & Avery, 2011). However, companies and brands also use these platforms. Fournier and Avery (2011) write that brands are “uninvited crashers to the web 2.0 party”, by which they mean that social network sites are not created to sell branded products but to link people together in collective conversational webs.

Instagram users thoughtfully choose the accounts that they want to follow and want to be shown in their personal news feeds. When suddenly a sponsored brand post

appears, users perceive this content as intrusiveness because it invades their

personally and carefully assembled feed and therefore interrupts their flow. Sponsored posts thus become content that users did not request to see but are unwillingly

(14)

Research by Boerman et al., (2017) shows that a celebrity endorsed Facebook post without a sponsorship disclosure did not activate the persuasion knowledge of consumers and therefore maintained positive brand attitudes. However, the findings also indicate that a celebrity endorsed post with a sponsorship disclosure increased critical and distrusting beliefs (Boerman et al., 2017). Where once influencer marketing is perceived as an approachable and genuine source of information, the disclosure alters this perception. According to the Persuasion Knowledge Model (Friestad & Wright, 1994), consumers’ attitude towards influencers might change when they realize that the content was not only created for entertaining purposes, but also serves a commercial purpose. This is also in line with the correspondence bias. The sponsorship disclosure will, almost painfully, indicate that the predisposition consumers had on the influencer is incorrect. This subsequently causes resistance towards the message. A sponsorship disclosure to a brand post however, does not alter consumers’ responses, compared to a brand post without a disclosure, because the brand itself already triggers their persuasion knowledge (Boerman et al., 2017). Once consumers see the sponsorship disclosure with the influencer post, this will generate an even larger sense of intrusiveness compared to a brand post. The hypothesis formulated taking is as follows:

H6: Influencer posts are perceived as more intrusive compared to brand posts.

Perceived intrusiveness as a mediator

Previous research indicates that perceived intrusiveness is an important construct that attributes to negative brand evaluations and consumer behaviour (Gazley, Hunt, & McLaren, 2015; van Doorn & Hoekstra, 2013). The higher a source is perceived as

(15)

Trustworthiness

intrusive, the lower and more negative consumers’ purchase intention and brand attitude will be. As hypothesized in the previous section, influencers are expected to generate a higher level of intrusiveness compared to brands. Thus it is expected that, subsequently, influencer posts will also generate more negative brand attitudes and lower purchase intentions compared to brand posts. The final hypotheses are as follows:

H7: Influencer posts generate more negative brand attitudes because of intrusiveness compared to brand posts.

H8: Influencer posts generate lower purchase intentions because of trustworthiness compared to brand posts.

The conceptual model visualising the different concepts of this study is shown in figure 1.

perceived

Figure 1: Conceptual model of this study.

Brand attitude Source: influencer/ brand Purchase intention Perceived intrusiveness

(16)

To test the hypotheses, an online experiment was conducted. The between-subjects experimental design included two conditions: exposure to an Instagram post by a brand (brand post) and exposure to an Instagram post by an influencer (influencer post). A total of 135 respondents (65% female) participated in this study. The average age of the sample was 30 (M=30.25, SD= 10,02). The youngest participant is 21 and the oldest participant is 63. 13 respondents started the experiment but were excluded from further participation because they did not meet the criteria of being familiar with the Instagram environment. Another 16 respondents were excluded from the analysis because they did not fully complete the experiment, leaving a working sample of 106 respondents. 25,9% of the sample completed a bachelor education on university level. The majority of the sample is of Dutch origin (63%). This sample is representative to the most common users of the Instagram platform namely young, highly educated adults aged 18-30 (Djafarova & Rushwort, 2017).1

Procedure

Participants were recruited through an online open survey link, which was distributed through Facebook, Instagram and Whatsapp. The participants were first asked to sign an informed consent. To be able to participate in this study a criteria was that

participants should be familiar with the Instagram environment. If they indicated not to be familiar with Instagram they would be excluded from further participation, debriefed and thanked. Participants were asked to pay attention to the post that they were about to be exposed to. They were randomly and evenly assigned to one of the two conditions (1=influencer, 0=brand). To ensure exposure to the stimuli, the

1 The manipulation check for this study showed discrepancies. Thus, it was chosen to proceed the

analyses with a smaller sample (N=68) that ensured correct manipulation. All the following results mentioned are based on this smaller sample. A more elaborate explanation of the manipulation check and motivation to why it was chosen to proceed with the smaller sample will follow in the results

(17)

stimulus was shown to participants for a minimum of 10 seconds. Respondents were only able to proceed with the questionnaire after the 10-second exposure had passed. Previous research shows a longer exposure to be more effective than a short exposure of three seconds (Boerman et al., 2012). After the exposure, participants continued with the questionnaire. This included the following measures: brand attitude, purchase intention, manipulation check, trustworthiness, perceived intrusiveness, persuasion knowledge post, persuasion knowledge on Instagram and descriptive statistics:

gender, age, country of origin and level of education. Finally, they were debriefed and thanked for their participation. Participants were recruited over a 17-day time frame in December 2017 and January 2018. The study took three months to finish.

Independent variable

Two Instagram posts were developed for this study: an influencer post, and a brand post. The posts were developed through an existing photo of a watch brand: Daniel Wellington. This photo was chosen because it could easily be perceived as an influencer as well as a brand photo. The logo of the Daniel Wellington brand was removed from the photo. The brand name ‘Horawatches’ and the influencer ‘Milou Maesen’ are fictional. The follow button on the brand post stresses the push-strategy. The influencer post did not have a follow button to indicate that the viewer follows the influencer. Since the viewer “follows” the influencer this is an indication of a pull-strategy. Both posts are disclosed, the brand post with: ‘Sponsored’, and the

influencer post with: ‘Paid partnership with Horawatches’. These disclosures are similar to the current disclosing practices on Instagram (Instagram, 2018). To maximize the chance of participants noticing the disclosure, it was slightly enlarged.

(18)

possible. The stimuli were not pilot tested due to time constraints and can be found in the appendices (Appendix A, Appendix B).

Measures

Brand attitude. The dependant variable ‘Brand attitude’ is measured using a

semantic differential scale developed by Spear and Sing (2004), seven-point bipolar Likert scale with five items (‘I think Hora watches is’: Bad/good,

unfavourable/favourable, dislikeable/likeable, unappealing/appealing,

unpleasant/pleasant). Reliability analysis showed the scale to be reliable (M= 4.84,

SD= 1,02, α= .93).

Purchase intention. The dependant variable ‘Purchase intentions’ is

measured using a five-item, seven-point Likert semantic differential scale developed by Spear and Singh (2004). Reliability analysis showed the scale to be reliable (M= 3.30, SD= 1.11, α= .90). The scale can be found in the appendix.2

Trustworthiness. The variable ‘Trustworthiness’ is measured using a scale of

the Source Credibility Theory developed by Ohanian (1990). This is a five-item, seven-point Likert scale (‘I think the source who posted this post is: dependable, honest, reliable, sincere, trustworthy, 1= strongly disagree to 7= strongly agree). The scale showed reliable (M=4.01, SD= 1.16, α= .93).

Perceived intrusiveness. The variable ‘Perceived intrusiveness’ is measured

using a scale developed by Li, Edwards, and Lee (2002). It is a item,

2 Initially, this scale was implemented in the questionnaire as a back-up option to measure purchase intention. The first scale was a 3-item 7-point Likert scale developed by Grewal, Monroe, and Krishnan (1998). The wording of the items could deem questionable for non-native English speakers and was expected to result in incorrect interpretation. Since it was expected that the majority of the sample would be non-native English speakers the second scale was implemented to ensure at least one correct measurement of purchase intention. Although the scale showed reliable (M=3.09, SD= 1.29, α= .82) it was not chosen to be used for analysis because the scale by Spear and Singh showed a higher

(19)

point Likert scale (‘When the Instagram post was shown, I thought it was’:

distracting, disturbing, forced, interfering, intrusive, invasive, obtrusive, 1- strongly disagree to 7= strongly agree). The items were collapsed and showed a reliable scale (M= 3.51, SD 1.58, α= .88).

Control variables. Two control variables are taken into account: ‘Persuasion

knowledge post’ and ‘Persuasion knowledge Instagram’. The first measures the extent to which respondents are aware of the persuasive intent of the Instagram post that they are being exposed to. In other words, it measures if and to what extent

respondents are able to identify the post as advertising. The latter measures the extent to which respondents are aware of the extent of which the Instagram platform is used for advertising purposes. This is a more general indication of overall persuasion knowledge on Instagram. These control variables were taken into account to ensure that effects of source were not caused by other differences between the experimental groups. The variable ‘Persuasion knowledge post’ is measured using a single item: ‘The Instagram post that was shown to me, is created to’: Advertise, 1= strongly disagree to 7= strongly agree, (M= 6.51, SD= .61). According to Rossiter (2011) a single-item is sufficient to measure a concrete, singular object. ‘Persuasion

knowledge Instagram’ is measured using a single item with five response categories: ‘Please indicate to what extent you think the following practices happen on

Instagram: Advertising, 1= not at all, 5= a great deal (M= 4.29, SD= .81).

Results

Manipulation check

(20)

The brand Horawatches, Don’t know’. The manipulation was tested using a crosstab analysis. Out of the 106 respondents, 28% answered incorrectly. Because this study focuses on the effects of sources, it is important that respondents recalled their source correctly. Thus, it was chosen to proceed the analyses with only the respondents with whom the manipulation was effective. 38 respondents were excluded from the analysis because of this reason, leaving a working sample of 68 respondents.3 An overview of the manipulation check can be found in table 1.

Table 1 Manipulation check Said influencer Said brand Said don’t know Total Had influencer 33 13 8 54 Had brand 1 35 16 52 Total 34 48 24 106 Randomization check

A randomization check was conducted to detect between-group differences. An independent samples t-tests was used to test: level of education, age, persuasion knowledge post and persuasion knowledge Instagram. A Chi-square test was used to test gender. The experimental groups showed no significant between-group

differences with respect to level of education (t (66)= .82, p= .415, 95% CI= [-.29, .70]), age (t (66)=-.72, p= .473, 95% CI [-6.08, 2.85]), persuasion knowledge post (t

3The results of the total and smaller sample were identical apart from one relationship. The effect of

(21)

-(66))=1.29, p= .762, 95% CI= [-.13, .644]), persuasion knowledge Instagram (t (66)= .40, p= .690, 96% CI= [-.23, .36]) and gender (χ2=1.99, p= .158). This means that the

conditions do not differ on level of education, age, persuasion knowledge post, persuasion knowledge Instagram and gender.

Effects of source on trustworthiness

The first hypothesis tested the effect of source on trustworthiness (H1). To test this effect a one-way ANOVA was conducted. The results showed no statistically significant group difference in sources on trustworthiness (F (1, 67)= .14, p= .709). This means that respondents exposed to an influencer post did not differ in the way that they perceived the post as trustworthy compared to respondents who were exposed to the brand post. These findings reject our first hypothesis.

Direct effects of source on brand attitude and purchase intention

For the second hypothesis we tested the direct-effect of source on brand attitude. One-way ANOVA results showed no statistically significant difference (F, (1,67)= 2.60,

p= .111). This means that there is no direct effect of source on brand attitude and

therefore there is no statistically significant difference between influencer and brand posts on their level of brand attitude. Subjects exposed to the influencer post did not differ in the way they perceived the brand compared to subjects exposed to the brand post. These findings reject our second hypothesis.

In our third hypothesis we state that influencer posts generate lower purchase intentions compared to brand posts. We tested this direct effect using a one-way ANOVA. Results showed no statistically significant effect of source on purchase

(22)

and the influencer post did not differ in their intention to make a purchase. These findings are not in line with our hypothesis and we must therefore reject it.

Mediation effects of trustworthiness

To test the influence of source on brand attitude in conjunction with the effect of trustworthiness (H4) PROCESS model 4 (Hayes, 2018) was used with 5000 bootstrap samples to estimate the bias corrected bootstrap confidence intervals (BCBCI).

This method was used for all simple mediation analyses in this study. Results showed no statistically significant mediation effect of trustworthiness on the effect of source on brand attitude (b= .02, boot SE= .08, 95% BCBCI [-.11, .21]). This means that, as previously established, there is no direct effect of source on brand attitude, there is also no significant relationship between source and brand attitude through

trustworthiness. Trustworthiness therefore does not explain a relationship between source and brand attitude. These findings are not in line with our H4. The results did indicate that trustworthiness has a significant effect on brand attitude (b=.25, t=2.46,

boot SE= .10, 95% BCBCI [.04, .46], p=.016). However, this did not influence the

total effect of the model. These results do however confirm previous studies, stating that trustworthiness is a predictor of brand attitude.

Next, we tested if there is a mediation effect of trustworthiness between source and purchase intention (H5). Results showed no significant mediation effect (b= .03,

boot SE= .09, 95% BCBCI [-.13, .24]). Results did indicate a significant effect of

trustworthiness on purchase intentions (p= .010, t=2.65, boot SE= .11, 95% BCBCI [.07, .52]). This means that although trustworthiness does not seem to explain a relationship between source and purchase intentions it does have an influence on purchase intentions. Even though these findings are not in line with our hypothesis,

(23)

they are in line with previous research indicating that trustworthiness is an important predictor of purchase intentions.

Effect of source on intrusiveness

For the sixth hypothesis we tested the effect of source on intrusiveness (H6). A one-way ANOVA analysis showed no effect (F (1, 67)= 2.13, p= .150). These findings state that subjects that had been exposed to an influencer post did not differ in the way that they perceived the post as intrusive compared to the subjects who had been exposed to the brand post. These findings are not in line with our hypothesis. We therefore reject our sixth hypothesis.

Mediation effects of intrusiveness

We examined if there is a mediation effect of intrusiveness on the relationship between source and brand attitude (H7). Results showed no significant mediation effect (b=-.05, boot SE= .06, BCBCI [-.18, .05]). These findings are not in line with our hypothesis and we must therefore reject it. Interestingly, results did not show an expected significant negative effect of intrusiveness on brand attitude (b= .14, t= -1.31, boot SE= .10, 95% BCBCI [-.35, .07], p= .193). This means that although the results cannot explain a relationship between source and brand attitude through

intrusiveness, they are also not line with previous studies stating that intrusiveness has a negative effect on brand attitude.

Finally, for our last hypothesis (H8), we tested if there is a mediation effect of intrusiveness on the relationship between source and purchase intention. Results showed no significant mediation effect (b= -.03, boot SE= .05, 95% BCBCI [-.17,

(24)

and purchase intention. These findings do not support our eighth hypothesis and we must therefore reject it. What is interesting is that the results also did not show an expected significant relationship between intrusiveness and purchase intention (b= -.09, boot SE= .05, 95% BCBCI [-.17, .07], p= .408). Previous studies have indicated intrusiveness to be an important predictor or purchase intentions (Gazley, et al., 2015). The findings of this research do not support these studies.

Figure 3: Tested mediation model in P values. * BCBCI lies between -1 and 1. Total effects of model is indicated with b.

Conclusion and Discussion

This study examined the difference in influencer marketing opposed to brand marketing in the context of a transparent Instagram environment on consumer behaviour and attitude. The objective of the study was quintuple. First, it aimed to gain insight into the difference of Instagram posting sources, either brand or influencer, on trustworthiness. Second, it studied the difference of source on

(25)

clarify the process through which a difference in source affects consumers’ attitude and behaviour, through a mediation of trustworthiness. The fourth aim was to indicate if there was a difference of source on intrusiveness. And the fifth and final aim was intended to clarify the process of source on brand attitude and purchase intention through a mediation effect of intrusiveness.

With respect to the first aim, no difference in source on trustworthiness was found. These findings indicate that the current advantages of influencer marketing over brand marketing in the context of the level of trustworthiness no longer hold in a transparent Instagram environment. Although the correspondence bias seems to affect the influencer post to an extent to which it is no longer perceived as more trustworthy compared to brands it does not seem to affect it to such an extent that it is perceived as less trustworthy opposed to brands. This could also be an indication that perhaps the bad perception of brands online, as indicated by Fournier and Avery (2011) might be altering. This might mean that consumer’s are getting used to the presence of brands online. Future research should monitor or re-evaluate the perception of brands online. This is also pertinent for the stated advantages of brand posts over influencer posts in the context of intrusiveness with respect to our fourth aim. The persuasion knowledge of consumers is activated by the sponsorship disclosure, but the influencer post does not generate such a perceived intrusiveness that it surpasses the perceived level of intrusiveness of subjects that were exposed to the brand post.

The second aim of this study was to indicate if there was a difference in source on consumers’ attitude towards the brand (H2) and towards their intention to purchase (H3). The results showed no difference. Simmons (2008) states that pushing strategies online will not work as effectively as pulling strategies. As demonstrated in this study,

(26)

transparent advertising future it does not make a difference for marketers if they either choose a pulling strategy like an influencer or a pushing strategy with a brand post to promote their brand with the goal of generating more positive brand evaluations. This also counts for the goal of generating a higher purchase intention.

The third and fifth aims of this study were to explain the process of which source on consumers’ attitude towards the brand and intention to purchase could be explained through trustworthiness (H4, H5) and intrusiveness (H7, H8). These findings also did not show between-source differences. Even though the results did not show mediation effects of trustworthiness or intrusiveness on the relationship between source and consumers’ behaviour and attitude, findings do confirm previous studies and theories that trustworthiness is an important indicator for brand attitude and purchase intentions (Wang & Scheinbaum, 2017). This confirmation can be of importance for future research and future marketing implementations, since

trustworthiness plays a role in consumers’ behaviour and attitude even in a

transparent online social media platform. Previous studies have also indicated that intrusiveness has a negative effect on brand attitude and purchase intentions (Gazley et al.,2015). However, the findings of this study seem to counter these studies. The findings show no effect of intrusiveness on neither brand attitude nor on purchase intention. However, there is discrepancy in these findings. When testing the same effect of intrusiveness on brand attitude with the bigger sample (N=106), findings did indicate intrusiveness to have an effect. These findings are fascinating, because it implies that the people who indicated not to know who posted the post do contribute to a heightened sense of intrusiveness.

Surprisingly, persuasion knowledge post did not seem to correspond with intrusiveness. Previous research has indicated that once consumers are confronted

(27)

with advertising, their persuasion knowledge is activated and this will induce a sense of intrusiveness (Van Doorn & Hoekstra, 2013; Boerman, et al., 2012). The findings counter this assumption. The average persuasion knowledge on the post, and thus recognition of the advertisement, did not seem to affect the extent to which respondent felt the post was intrusive even though the average indication of

persuasion knowledge post was high (M= 6.51, SD= .61). This could imply that the correspondence bias that contributes to a higher perceived intrusiveness of influencer is not a determining factor. Future research should concentrate on this implied change in advertising attitudes of consumers and the effect of the correspondence bias in a transparent environment.

Furthermore, the manipulation check of this study also showed a discrepancy. As previously indicated, the sample that was selected for the analysis was reduced from 135 respondents to 68 respondents. This relatively large exclusion rate is due to the fact that, apart from subjects not completely finishing the survey and not meeting the participation criteria, subjects were unable to identify the source to which they had been exposed to. Interestingly, the respondents from the influencer condition had more trouble indicating their source (N=13) compared to subjects who had been exposed to the brand post (N=1). This is an interesting sponsorship disclosure effect because even though the influencer post was disclosed, these numbers state that it raises a lot of confusion and maybe even suspicion amongst participants. This is an indication that the current disclosing practice on Instagram might not be efficient enough for people to understand how the advertising construction behind the

influencer post works. This is also in line with the research by Boerman et al. (2017) that indicates that the sponsorship disclosure to celebrity endorses posts increases

(28)

experimental context. Which means that generally participants have an increased sense of awareness and therefore do not reflect real-life conditions. The fact that respondents were unable to identify the source in this experimental condition is a prospect that consumers in real-life will almost definitely not notice or pay notion to the source of the post.

This study has some limitations. Firstly, the study involved only one product type (watches). Consumer motives might differ on different types of products and to an extent might impact their attitude and behaviour. The second limitation postulates the fact that the influencer was fictional and that respondents’ attitudes may be affected by the perceived fit of the product to the celebrity (Bergkvist et al., 2016) or in this case: influencer. Since, in this study, the respondent has no framework or background information of the influencer to refer to, they also cannot attribute the fit of the product to the influencer. Future research should therefore focus on more widespread product categories and or provide more information to which respondents can attribute fit of product to the influencers. A third limitation is that this study only compared the effects in a transparent environment. In order to better comprehend how consumers respond to advertising transparency and to completely generalize these findings, future research should study the effects of advertising transparency in comparison to a non-transparent environment. Research should also focus on the content or composition of the sponsorship disclosure to minimize confusion amongst the public and maximize advertising transparency.

This study is one of the first to address complete advertising transparency on the Instagram platform and its findings prove valuable insights into the future of influencer and brand marketing. It shows that the advantages of using influencer marketing over brand marketing no longer withhold when influencer posts are

(29)

disclosed. The findings also show that although the intent of the sponsorship

disclosure is to make advertising practices on Instagram more transparent, it deemed to cause the opposite. Where advertising transparency revolves about providing clarity to the consumers, in this study it appeared to cause only more cloudiness.

(30)

References

Amos, C., Holmes, G., & Strutton, D. (2008). Exploring the relationship between celebrity endorser effects and advertising effectiveness: A quantitative synthesis of effect size. International Journal of Advertising, 27 (2), 209–34 Balakrishnan, A., & Boorstin, J. (2017, September 25). Instagram says it now has 800

million users, up 100 million since April. Retrieved from

https://www.cnbc.com/2017/09/25/how-many-users-does-instagram-have-now-800-million.html

Belch, G.E., & Belch, M.A. (2015). Advertising and promotion: An integrated

marketing communications perspective. International edition: McGraw-Hill

Education.

Bergkvist, L., Hjalmarson, H., & Magi, A. (2016). A new model of how celebrity endorsements work: Attitude toward the endorsement as a mediator of celebrity source and endorsement effects. International Journal of

Advertising, 35 (2), 171–84.

Boerman, S. C., van Reijmersdal, E. A., & Neijens, P. C. (2012). Sponsorship

disclosure: Effects of duration on persuasion knowledge and brand responses.

Journal of Communication, 62(6), 1047-1064.

Boerman, S.C., Willemsen, L.M., & Van der Aa, E.P. (2017). “This post is

sponsored” Effects of sponsorship disclosure on persuasion knowledge and electronic word of mouth in the context of Facebook. Journal of Interactive

Marketing, 38, 82-92.

Cain, R.M. (2011). Embedded advertising on television: Disclosure deception and free speech rights. Journal of Public Policy & Marketing, 30 (2), 226-238. Chaffey, D. (2017, April 27). Global social media research summary. Retrieved from

https://www.smartinsights.com/social-media-marketing/social-media-strategy/new-global-social-media-research/.

(31)

Consumer Research, 26 (1), 627-631.

Daugherty, T., & Hoffman, E. (2013). eWOM and the importance of capturing consumer attention within social media. Journal of Marketing

Communications, 20 (1-2), 82-102.

Dekker, K., & van Reijmersdal, E. A. (2013). Disclosing celebrity endorsement in a television program to mitigate persuasion: How disclosure type and celebrity credibility interact. Journal of Promotion Management, 19 (2), 224-240. De Veirman, M., Cauberghe, V., & Hudders, L. (2017) Marketing through Instagram

influencers: The impact of number of follower and divergence on brand attitude. International Journal of Advertising, 36 (5), 789-828.

De Vries, L., Gensler, S., & Leeflang, P.S.H. (2012). Popularity of brand posts on brand fan pages: An investigation of the effects of social media

marketing. Journal of Interactive Marketing, 26 (2), 83–91.

Djafarova, E., & Rushwort, C. (2017) Exploring the credibility of online celebrities Instagram profiles in influencing the purchase decisions of young female users. Computers in Human Behaviour, 68, 1-7.

Edwards, S., Li, H., & Lee, J.H. (2002). Forced exposure and psychological reactance: Antecedents and consequences of the perceived intrusiveness of pop-up ads. Journal of Advertising, 31 (3), 83-95.

Erkan, I., & Evans, C. (2016). The influence of eWOM in social media on consumers’ purchase intentions: An extended approach to information adoption.

Computers in Human Behaviour, 61, 47-55.

Forer, L. (2017, July 31). The lowdown on the rise of influencer marketing [Infographic]. Retrieved from

https://www.marketingprofs.com/chirp/2017/32472/the-lowdown-on-the-rise-of-influencer-marketing-infographic

(32)

193-Fransen M.L., Verlegh, P.W.J., Kirmani, A., & Smit, E.G. (2015). A typology of consumer strategies for resisting advertising, and a review of mechanisms for countering them. International Journal of Advertising, 34 (1), 6–16.

Friestad, M., & Wright, P. (1994). The persuasion knowledge model: How people cope with persuasion attempts. Journal of Consumer Research, 21(1), 1-31. Gazley, A., Hunt, A., & McLaren, L. (2015). The effects of location-based services on consumer purchase intention at point of purchase. European Journal of

Marketing, 49, 1686-1708.

Grewal, D., Monroe, K. B., & Hrishnan, R. (1998). The effects of price-comparison advertising on buyers’ perceptions of acquisition value, transaction value, and behavioural intentions. Journal of Marketing, 62 (2), 46-59.

Ha, L. (1996). Advertising clutter in consumer magazines: Dimensions and Effects.

Journal of Advertising Research, 36, 76-83.

Instagram, About us. (2018) Retrieved from https://www.instagram.com/about/us/. Instagram, Business. (2018). Retrieved from https://business.instagram.com.

Li, H., Edwards, S., & Lee, J. (2002). Measuring the intrusiveness of advertisements: Scale development and validation. Journal of Advertising, 31(2), 37-47. McCracken, G. (1989). Who is the celebrity endorser? Cultural foundations of the

endorsement process. Journal of Consumer Research, 16, 310- 321. Mitchell, A.A., & Olson, J.C. (1981). Are product beliefs the only mediator of

advertising effects on brand attitude? Journal of Marketing Research, 18, 318-323.

Morrisson, K. (2015, March 27). Survey: Nearly 60% of marketers plan to boost ainfluencer marketing budget. Retrieved from

http://www.adweek.com/digital/survey-tomoson-influence-marketing-budgets/#/.

(33)

and why it must replace psychometrics. New York, NY: Springer Science + Business Media.

Simmons, G. (2008). Marketing to postmodern consumers: Introducing the internet chameleon. European Journal of Marketing, 42, 299–310.

Spears, N., & Singh, S.N. (2004). Measuring attitude toward the brand and purchase intentions. Journal of Current Issues & Research in Advertising, 26 (2), 53-66. Stichting Reclame Code. (2018). Retrieved from

https://www.reclamecode.nl/consument/default.asp?paginaID=141&hID=102 Truong, Y., & Simmons, G. (2010). Perceived intrusiveness in digital advertising:

Strategic marketing implications. Journal of Strategic Marketing, 18 (3), 239-256.

Van Doorn, J., & Hoekstra, J.C. (2013) Customization of online advertising: the role of intrusiveness. Marketing Letters, 24 (4), 339-351 [Peer Reviewed Journal]. Wang, S.W., & Scheinbaum, A. C. (2017). Trustworthiness trumps attractiveness and

expertise: Enhancing brand credibility through celebrity endorsement. Journal

of Advertising Research, 31 (2), 37-47.

Wheeler, S. (2017, August 22). Which social media platforms are relevant for your influencer marketing campaign? [Blogpost]. Retrieved from

https://influence.bloglovin.com/which-social-media-platforms-are-relevant-for-your-influencer-marketing-campaign-15569b79d72d. .

Woods, L. (2008). The consumer and advertising regulation in the television without frontier and audiovisual media services directives. Journal of Consumer

(34)

Appendix A Stimulus 1:

(35)

Appendix B Stimulus 2:

(36)

Appendix C Figure 2:

Referenties

GERELATEERDE DOCUMENTEN

Dit heb ik gedaan door ideaaltypische taken (op papier) te creëren waarin leerlingen twee voor het Religieproject relevante onderzoeksvaardigheden (vergelijken en interviewen)

(Aukema q.q./ING Commercial Finance) r.o.. • De bank wist, althans behoorde te voorzien, dat de vennootschappen ten gevolge van de financieringsconstructie niet langer

For that reason, the current study compares the impact of textual and symbolic disclosure as well as the combination of textual and symbolic disclosure on

This research focuses on the social media platform Instagram as influencers are most active on this platform (Influencermarketinghub, 2021). 1.3 Research question.. The

The second factor analysis with the remaining 24 items was conducted and 6 factors, namely purchase intention, source credibility expertise, source credibility trustworthiness,

• People on Instagram, next to celebrity influencers, who have a large following on social media (Bijen, 2017; Kalavrezos, 2016).. • Comparable to celebrity influencers

All studied alcohols show similar vibrational lifetimes of the OH stretching mode and similar HB dynamics, which is described by the fast (~200 fs) and slow components (~4 ps).

For all converged coupled-cavity bands, we find that light hops predominantly in a few high-symmetry directions including the Cartesian (x , y, z) directions, therefore we propose