The effects of micro vs macro influencers on brand awareness, brand attitude, and purchase intention, and the moderating role of advertising appeals
Almperta Christodoulaki (11103752) Master’s Thesis
University of Amsterdam Graduate School of Communication Master’s Programme Communication Science
Supervisor: Dr. Stephan Winter Date of Completion: June 29, 2018
Abstract
The emergence and growth of social media platforms have created valuable opportunities for influencer marketing. Social media influencers (SMIs) may be perceived as trustworthy and accessible, shaping, thus, their audience’s beliefs and behaviours. The aim of the study was to determine whether different types of influencers, namely micro-influencers (100.000
followers or below) and macro-influencers (1 million followers or above) can have an effect on their audience’s brand awareness, brand attitude, and purchase intention. Moreover, the study sought to uncover whether this effect could be moderated by the advertising appeal type used in the influencer’s Instagram endorsement, as well as mediated by perceived source credibility. In an online experiment (N = 232) comprised of a 2 (Type of influencer: micro vs macro-influencer) x 2 (Advertising appeal: soft vs hard-sell) between-subjects factorial design, participants were evaluated in terms of their brand recall, brand attitude, and purchase intention towards a fictitious brand of women’s watches. The results displayed no effect of influencer type on brand recall, brand attitude, and purchase intention. No moderation effect of advertising appeal type on the relationship between influencer type and brand recall, brand attitude, and purchase intention was found. Lastly, no indirect effect was uncovered between the three dimensions of perceived source credibility; trustworthiness, attractiveness, and expertise.
The effects of micro vs macro influencers on brand awareness, brand attitude, and purchase intention, and the moderating role of advertising appeals
Following the rapid growth of social media, companies are now leveraging these platforms to attract new customers and establish relationships (Michaelidou, Siamagka, &
Christodoulides, 2011). For instance, Instagram is hosting effective marketing tactics through sponsored social messages (eMarketer, 2015). This emerging type of marketing is called influencer marketing (Ranga & Sharma, 2014). Freberg and colleagues (2011) defined “social media influencers” (SMIs) as «a new type of third party endorsers who shape audience attitudes».
This concept is related to Katz and Lazasfeld’s study (1955), which argued that opinion leaders filter mass media messages and transmit them throughout their social interactions. Similarly, SMIs are able to influence their audience’s behaviours and drive action to brands’ benefits (Gardner, 2005; Uzunoglu & Kip, 2014). Establishing partnerships with SMIs requires effective management. However, 75% of brands claim that identifying the right influencer is their biggest challenge (Roy, 2015).
Most marketers identify appropriate SMIs for their brand based on the most daily hits or the highest number of followers (Freberg, Graham, McGaughey, & Freberg, 2011). Nowadays there has been an increasing trend for marketers to distinguish between micro and macro-influencers (Chen, 2016). Micro-influencers have between 10.000 and 100.000 followers and are believed to produce higher levels of engagement comparing to macro-influencers (Dinesh, 2017). Macro-macro-influencers have millions of followers, achieving, thus, great amounts of reach and increasing their audience’s brand awareness.
In academic literature, there is no general agreement on the effects of the number of followers on one’s social media account. Several scholars argue that a large number of subscribers implies a more attractive and trustworthy influencer (Djafarova & Rushworth,
2017; Jin & Phua, 2014). However, there is evidence that “too many” followers can elicit decreased perceived credibility (Tong, Van Der Heide, Langwell, & Walther 2008; Utz, 2010; Westerman, Spence, & Van Der Heide, 2012). Further academic research needs to be conducted, in order to determine the impact a high versus low number of followers on SMIs’ accounts can have on consumers.
Other than the number of followers, SMIs may also differ in terms of the layout of their brand-endorsing picture (Hu, Manikonda, & Kambhampati, 2014). Their practices may include a close-up shot of the product or a more creative shot that subtly features the product. The present study will investigate if the effects of SMIs on their audience depend on the type of advertising appeal; soft-sell or hard-sell advertising. Soft-selling refers to imaginative and transformational advertising, while hard-selling concerns rational and informational
advertising (Okazaki, Mueller, & Taylor, 2010). Previous studies have explored the different effects of soft vs hard-sell advertising appeals (Chu, Gerstner, & Hess, 1995; Okazaki, Mueller, & Taylor, 2010); however, no previous literature has connected this with influencer marketing.
For influencer marketing campaigns, the present study can provide insights on the type of influencer that is recommended to partner with based on the campaign’s needs (regarding brand awareness, brand attitude, or purchase intention). For the scientific community, the present research can provide a deeper understanding on the formation of people’s perceptions and attitudes according to the recommending source, as well as the effect of advertising appeals in influencer marketing. Taking Katz and Lazarsfeld’s two-step flow theory (1955) to the digital era, the present study aims to empirically recognize SMIs as opinion leaders that influence their audience. Additionally, the goal of this exploratory study includes developing academic background that supports marketers’ distinction of micro and
macro-influencers, by providing evidence that different types of influencers can produce different effects on their audience.
Overall, the overarching goal of this study is to assess firstly, whether micro versus macro-SMIs elicit different effects on the receivers’ brand awareness, brand attitude, and purchase intention, secondly, whether those effects are mediated by perceived source credibility, and lastly, whether those effects are moderated by the advertising appeal type.
Theoretical framework
As a basis for this research, previous literature on social media, influencer marketing, and micro/macro-influencers will be introduced. Brand awareness, brand attitude, and purchase intentions will be presented as dependent variables. This order reflects the classical hierarchy of consuming advertised information; people first build awareness about a brand, develop positive or negative feelings towards it, and lastly buy or reject the product/brand (Kotler & Bliemel, 2001). Additionally, perceived source credibility will be discussed as a potential explanation of the effects that the type of influencer produces on the receivers. Lastly, the role of advertising appeals will be summarized through the current body of research, and connected to influencer marketing.
Social media
The emergence and rapid growth of social media (e.g. Facebook, Instagram, Snapchat, YouTube) has altered the way businesses communicate their marketing efforts to the consumers. It is estimated that in 2019 there will be 2.77 billion social media users
worldwide (“Statista”, 2018). Social media platforms have been found to be effective tools for increasing brand awareness, encouraging WOM activities, and guiding purchasing intentions (Hutter, Hautz, Dennhardt, & Fuller, 2013). People use social media in order to
create, express, and share content about numerous topics, including brands (Muntinga, Moorman, & Smit, 2011). This type of electronic word-of-mouth (eWOM) is especially stimulated by influential online users, known as social media influencers (SMIs).
Influencer marketing
Nowadays, influencer marketing involves SMIs uploading brand-related content, by either sharing products at their own initiative, or partnering with brands in exchange for (in)direct monetary benefits (Lu, Chang, & Chang, 2014). These practices resemble celebrity endorsements, aiming to create eWOM. The Word of Mouth Marketing Association
(WOMMA, 2016) recognized influencer marketing’s ability to influence consumers’
opinions through influential leaders in key communities. For instance, Instagram influencers may affect their audience’s purchase behaviour (Djafarova & Rushworth, 2017).
Katz and Lazarsfeld (1955) were among the first to recognize opinion leaders who mediate the transmission of information. In their two-step flow theory, mass media messages are disseminated by opinion leaders, who transmit the interpreted information to their large number of personal connections. The importance of opinion leaders lies in the fact that they are able to familiarize their peers with social issues or even consumer choices (Nisbet & Kotcher, 2009). Rogers and Cartano (1962) have referred to opinion leaders as people who can impose “an unequal amount of influence on the decisions of others”.
SMIs may be perceived as digital opinion leaders, as they transmit brand-related messages and influence their audience. Katz and Lazarsfeld (1955) referred to people who know their opinion leaders in person. However, due to technological advancements, face-to-face contact is no longer a prerequisite for opinion leadership (Uzunoglu & Kip, 2014). SMIs manage to influence audiences through an online environment (Abidin, 2016).
Lastly, Rogers (1962) viewed opinion leaders’ influence as a continuous variable which depends on the number of near-peers; meaning that sometimes opinion leaders are asked for advice by a large number of peers, whereas in other cases only by a small number. The same concept may apply to the modern-day opinion leaders, recognizing a distinction between SMIs’ influence based on the volume of their following.
Micro vs macro influencers
There has been an increasing interest in researching SMIs and their characteristics (Eirinaki, Monga, & Sundaram 2012). However, limited academic research exists on the grounds of distinguishing micro and macro-influencers. Past literature has mainly focused on “celebrity” influencers, or macro-influencers, who have a large number of subscribers and are able to transform a product’s popularity by expanding its visibility to a wide audience (Jin & Phua, 2014; Scott, 2011).
However, there is a recognisable trend for brands to partner with micro-influencers (Chen, 2016). According to marketers, micro-influencers are Instagrammers, YouTubers, or Snapchatters with a small fan base (less than 100.000). Following micro-influencers are “middle-power” influencers, with 100.000 to a million followers, and macro-influencers with more than one million followers (Barker, 2017; Pavlika, 2017; Pierucci, 2018).
The number of followers is the main distinction between these influencer types. Additionally, the content of their posts may differ depending on their categorization. It is believed that the micro-influencer trend arose because they were able to address smaller audiences in niche market segments (Barker, 2017). For example, Allison Graham is an Instagram micro-influencer who is considered an expert to those who steer off of big fashion labels. Macro-influencers, on the other hand, may share a broader variety of topics. For example, Chriselle Lim’s posts, an Instagrammer with one million followers, extend from
sharing her favourite fashion and travels, to endorsing a car. Chriselle has received negative feedback due to endorsing incongruent content and has been accused of poor influencer marketing (“Curalate”, 2016). This suggests that macro-influencers may run the risk of appearing insincere due to their larger spectrum of content1. In sum, micro and macro-influencers differ in terms of their number of followers, their content, and the effects they produce on their audience.
Kapitan and Silveira (2016) argue that the number of followers affects the types of attributions made in consumers’ minds about the SMIs, which are crucial for the
effectiveness of the endorsement. The Keller Fay Group (2016) suggested that micro-influencers are targeting more niche categories and appear knowledgeable and authentic when it comes to their recommendations, especially comparing to celebrities. Tong and colleagues (2008) indicated that the number of one’s Facebook friends has an effect on his/her social attractiveness. More specifically, people who had moderate numbers of friends were perceived as more likeable than those with greater numbers of friends. The same findings were revealed for Twitter, where “too many” followers resulted in lower judgments of expertise and trustworthiness (Westerman, Spence, & Van Der Heide, 2012). This effect may also apply to SMIs; macro-influencers may be considered less credible due to their lack of expertise.
Brand awareness
One of the ways in which SMIs may influence their audience is by enhancing the brand awareness for certain products/services. According to Michaelidou, Siamagka, and
1
Although micro and macro-influencers may portray different content, the present paper will show the same content for both, in order to isolate potential effects solely based on the number of followers.
Christodoulides (2011), 82% of companies use social media platforms to increase brand awareness. Scholars have defined brand awareness through two dimensions: brand recognition and recall (Aaker, 1991; Keller, 1993; Radder & Huang, 2008). Brand
recognition assumes prior exposure to the brand and involves identifying a brand when given its name as a cue. Brand recall refers to the ability to remember a brand when given the brand category (Keller, 1993).
Marketers have been advised to leverage social media platforms to strengthen the consumers’ brand awareness at a low cost (Wang, Hsiao, Yang, & Hajli, 2016). Advertising tools, such as influencer marketing, can increase brand awareness and place the brand name in the memory of the consumers (Brouwer, 2017; Keller, 2003).
Marketing practitioners have suggested that one of the most important advantages of macro-influencers is their ability to generate visibility (Hatton, 2018; Pavlika, 2017). Similar to celebrities, macro-influencers are expected to generate positive eWOM, due to the large number of following that is exposed to their opinions (Djafarova & Rushworth, 2017; Spry, Pappu, & Cornwell, 2011). The effectiveness of macro-influencers enhancing their
audience’s brand awareness is expected to be a result of their posts being broadcasted to a wide population, who may re-post the message and spread the brand’s exposure (Scott, 2011). Upon seeing a high number of followers on one’s social media profile, people may assume that the influencer is an important celebrity, which brings value to the brand name (Djafarova & Rushworth, 2017). Eventually, consumers may associate brands with the SMI endorsers, which adds to the depth of brand awareness and brand credibility (Djafarova & Rushworth, 2017; Spry, Pappu, & Cornwell, 2011).
H1. Macro-influencers will be more effective in forming their audience’s brand awareness compared to micro-influencers.
Brand attitude
Following a consumer’s familiarization with the brand is the formation of a positive or negative brand attitude (Rossiter, 2014). Mitchell and Olson (1981) defined brand attitude as one’s internal evaluation concerning a brand stimulus. Abzari, Ghassemi, and Vosta (2014) measured people’s attitude towards the same brand through traditional and social media. Their findings suggested that social media induce a more positive brand attitude than traditional media. Furthermore, Jin (2018) observed that a peer’s Facebook post created more positive attitudes toward the endorsement, comparing to a sponsored Facebook post on a celebrity’s profile. These findings may connect to the present study, with micro-influencers being perceived more similar to a peer, and macro-influencers being closer to celebrities.
As far as influencer marketing is concerned, the evaluation that consumers make involves the SMI brand-endorsing post. Scholars have argued that source credibility is an important factor when forming brand attitude (Bhatt, Jayswal, & Patel, 2013; Boerman, Willemsen, & Van Der Aa, 2017; Wang, Kao, & Ngamsiriudom, 2017). The more credible the endorser is perceived, the more likely consumers will form a favourable brand attitude. Djafarova and Rushworth (2017) argued that the persuasive strength of the endorser to alter one’s brand attitude may depend on the quality of their argument. This suggests that when SMIs endorse brands in a valid and persuasive way, their audience will likely develop a positive brand attitude. If the SMI endorsement seems invalid, consumers may form an unfavourable attitude (Cheung, Luo, Sia, & Chen, 2009). The fact that micro-influencers address to niche audiences in a consistent theme may constitute them more credible than macro-influencers, who may use their popularity to endorse brands that they do not truly like. Therefore, it is assumed that micro-influencers’ authenticity is more likely to induce positive brand attitudes to the consumers.
H2. Micro-influencers will be more effective in shaping their audience’s brand attitude than macro-influencers.
Purchase intention
Purchase intention is a consumer’s willingness to undertake effort into buying a product (Spears & Singh, 2004). The Theory of Reasoned Action (Fishbein & Ajzen, 1975) and the Theory of Planned Behavior (Ajzen, 1985) suggest that people’s attitudes may influence their intentions, which can consecutively affect their purchase behavior. The present study will focus on purchase intention rather than behavior, since the latter may be influenced by external factors (e.g. unanticipated financial changes; Sun & Morwitz, 2010).
Djafarova and Rushworth’s study (2017) revealed that celebrities on Instagram were able to influence the behavior of young female consumers. Their findings also revealed that non-traditional celebrities, such as “Instagrammers”, were more powerful in influencing the audience’s purchase intention comparing to traditional celebrities, due to being perceived more credible. Source credibility plays an important role when assessing a communicator, and it is evaluated based on the following characteristics: expertise (i.e. whether the communicator is knowledgeable), trustworthiness (i.e. the source’s honesty), and
attractiveness (i.e. whether the audience likes the communicator; Hwang, & Jeong, 2016; Ohanian, 1990).
Gladwell (2005) studied the characteristics of those who are able to influence consumers, and distinguished “connectors”, “mavens”, and “salesmen”. As Mourdoukoutas and Siomkos (2009) suggest, connectors communicate with people only if they receive something in return, mavens brag about being masters of information, and salesmen seek to sell products. Macro-influencers could be parallelized with these categorizations, since they are seen as online celebrities who promote products through brand partnerships. This could
induce low perceived credibility and declined influencing power in regards to people’s purchase intention. On the other hand, micro-influencers can be compared to “innovators” and “early adopters”. Their communality lies in their ability to act as opinion leaders in niche and more specific fields, where they are believed to be experts (“Mediakix”, 2016;
Mourdoukoutas & Siomkos, 2009).
Since previous studies have suggested that “too many” social media followers may be associated with declined expertise and trustworthiness (Westerman, Spence, & Van Der Heide, 2012; Utz, 2010), and that SMIs who are perceived as credible may be able to
influence their audience’s purchase intention (Djafarova & Rushworth, 2017), the following will be examined:
H3. Micro-influencers will be more effective in shaping their audience’s purchase intentions than macro-influencers.
H4. The effect that the type of influencer (micro vs macro) can have on their audience’s brand awareness, brand attitude, and purchase intentions is mediated by the source’s perceived credibility.
Advertising appeal (soft vs hard sell)
When SMIs upload brand-endorsing posts, they may use different advertising appeals; for example, the endorsed product could be portrayed on a close-up shot, or inserted in the background. Mueller (1987) argued that soft-sell appeals are subtle and convey sentiments through a beautiful scene or storytelling means. Hard-sell appeals place the emphasis on the sales of the advertised product, highlighting the brand name (Okazaki, Mueller, & Taylor, 2010). Alden, Steenkamp, and Batra (1999) interpreted soft-sell appeals as image-oriented content that communicates associations with the brand, and hard-sell appeals as message arguments that emphasize the features of the product/brand.
Previous studies have suggested that when someone endorses a product on social media, the audience develops a positive brand attitude if they believe that the product’s review is valid (Spry, Pappu, & Cornwell, 2011). However, if the review is perceived false and invalid, the audience will likely develop a negative attitude towards the brand and the endorser (Cheung, Luo, Sia, & Chen, 2009). This suggests that the quality of the social media post is evaluated in regards to perception of the endorser, as well as relevance (Djafarova & Rushworth, 2017).
In the present study, it is assumed that macro-influencers are perceived as less credible than micro-influencers, due to people’s skepticism when seeing their large number of followers. This means that their endorsements are more likely to activate the consumers’ persuasion knowledge, who may recognize the macro-influencer post as a persuasion tactic (Hwang & Jeong, 2016). Therefore, it is assumed that macro-influencers will be more effective in influencing the audience’s brand awareness when using a soft-sell advertising appeal, since this subtly presents their endorsed product, and potentially avoids skepticism and resistance (Lu, Chang, & Chang, 2014). On the other hand, since micro-influencers are assumed to be more trustworthy and credible, a hard-sell appeal may be considered as more direct and appropriate for giving recommendations to consumers, whose purchase intentions and brand attitudes will be positively affected. In sum:
H5a. The effect of micro-influencers shaping their audience’s purchase intentions more effectively than macro-influencers is expected to be stronger for hard-sell advertising appeals, comparing to soft-sell appeals.
H5b. The effect of micro-influencers shaping their audience’s brand attitude more effectively than macro-influencers is expected to be stronger for hard-sell advertising appeals, comparing to soft-sell appeals.
H5c. The effect of macro-influencers shaping their audience’s brand awareness more effectively than micro-influencers is expected to be stronger for soft-sell advertising appeals, comparing to hard-sell appeals.
Study overview
The information above is summarized in the conceptual model shown in Figure 1. This model presents an overview of the hypotheses discussed along the theoretical
framework of the study.
Figure 1. Conceptual model.
Research methodology Design
The study used an experimental 2 (influencer type: micro vs macro) x 2 (advertising appeal: soft-sell vs hard-sell) between-subjects design, with three dependent variables: brand awareness, brand attitude, and purchase intention, and one mediator variable: perceived
source credibility. The type of influencer was manipulated based on the number of the influencer’s followers (micro-SMI: 85 thousand, macro-SMI: 2.1 million). The advertising appeal presented the endorsed brand either subtly in the Instagram picture (soft-sell), or had the endorsed brand as the main focus of the Instagram picture (hard-sell). The design resulted in four conditions.
Procedure
In the beginning of the survey, the theme was described vaguely as marketing on social media, as the less aware participants were about the goal of the experiment, the closer their reactions would be to real life. Anyone who did not select ‘female’ as their gender was excluded from the study. Since the study demonstrated a women’s fashion product (watch), and in order for the results of the study to be generalizable to a wider population, a sample consisting only of women was deemed appropriate. Previous studies that have explored fashion Instagram effects have also limited their research on females (Djafarova & Rushworth, 2017; Duffy & Hund, 2015). Additionally, women can be found in 68% of Instagram users (Aslam, 2018).
Firstly, participants were asked questions about their social media use2, their most frequently visited platforms, their Instagram usage, and whether they follow any SMIs. The participants were then randomly exposed to the stimuli based on their condition and the moderator (influencer type and advertising appeal). Following the manipulation check (recalling the number of followers), the dependent measures of the study were assessed;
2
The question about having an account on any social media platform was meant as a control that would exclude participants who were not familiar with social media. However, this was not the case for any of the subjects of the study.
brand awareness, brand attitude, and purchase intention. The manipulation check of the advertising appeal was added towards the end of the survey, so that the results would be as least bias as possible. Following the measurement of the participants’ perceived source credibility were socio-demographic questions (age, nationality, level of education).
Participants
The online-based survey experiment reached 262 respondents. After excluding the
participants who did not fully complete the questionnaire and those who were not female, the number dropped to 232. The participants were recruited through social media platforms (Facebook and Instagram). This constituted it as a non-probabilistic self-selection convenience sampling method.
The participants’ average age was 26years old (M = 25.53, SD = 3.34). In total, 42 nationalities took part in the study (30.6% Greek, 17.2% Dutch, 8.6% German). The majority completed a Bachelor’s (51.3%) or a Master’s degree (37.1%), suggesting that the sample had a relatively high education level.
Facebook was participants’ most used platform (48.7%), followed by Instagram (40.1%). Almost everyone claimed that they use social media several times a day (92.2%). Numerous participants (81%) claimed that they follow SMIs, some examples being
beauty/lifestyle influencers (e.g. Chiara Ferragni), celebrities (e.g. Kim Kardashian), and artists (e.g. Martin Garrix).
Pilot test
The aim of the pilot test was to select the appropriate stimuli for the main study concerning the advertising appeal. The pilot test followed a between-subjects design with influencer type (micro vs macro) as conditions. Within the two conditions, stimulus type x
advertising appeal type was assessed (within-subjects), exposing subjects to both a soft-sell and a hard-sell stimulus.
A convenience sample (N = 63) participated in the pilot test, with a mean age of 26 (SD = 6.50). In total, 33 women were randomly exposed to the micro-influencer, and 30 to the macro-influencer stimuli. After the exposure to the stimuli, the participants were asked to recall the number of followers through a continuous variable with 6 options3. A
cross-tabulation revealed a significant difference between the two conditions, χ2
(5) = 40.46, p < .001. These findings show that participants perceived a clear difference between the number of followers for the micro (M = 2.24, SD = .87) and macro (M = 4.37, SD = 1.27) influencer.
Moreover, an analysis of variance (ANOVA) was conducted, in order to examine the stimuli presented in the pilot test (two soft-sell and two hard-sell stimuli). The stimuli had a significant effect on people’s perceptions in regards to the prominence of the watch, F(2,123)
= 14.775, p < .001. A Bonferroni post-hoc test revealed a significant difference between the soft-sell picture of the first condition (M = 4.33, SD = 1.20) and the hard-sell picture of the second condition (M = 5.68, SD = 1.04, Mdifference= -1.34, p < .001). Thus, the two Instagram pictures were chosen for the main study, representing the hard-sell and soft-sell advertising appeals (see Appendix A).
Stimuli
For the four experimental conditions, stimulus material was designed through Instagram and Adobe Photoshop. A fictitious Instagram profile was presented across all conditions
3
(1) Less than 10.000; (2) From 10.000 to 100.000; (3) From 100.000 to 500.000; (4) From 500.000 to 1 million; (5) From 1 million to 3 million; (6) 3 million or above. For the micro-influencer, the correct answer was the second, and for the macro-influencer it was the fifth.
(“Cloud nine”). The micro-influencer profile had 85 thousand followers, while the macro-influencer’s followers were 2.1 million4
.
Additionally, a fictitious brand was created (“Millie”), in order to eliminate any
(un)favourable effects of a known brand on consumers’ awareness, attitudes, and purchase intentions. A watch was selected as the endorsed product. Watches have been used in previous studies, as they are a product category in which younger adults are intrinsically interested, with medium involvement (Griffith & Chen, 2004; Li & Lo, 2015).
People spend more time on Instagram than other similar sites, which suggests the importance of researching this medium (Sheldon & Bryant, 2016). Since Instagram has 800 million users (“Statista”, 2018), and brand endorsements are becoming increasingly popular, this platform was deemed the most relevant to the current online consumer behaviours.
Across all experimental conditions, a caption was presented below the stimulus/picture, aiming at making the Instagram post close to real-life5. The soft-sell picture portrayed a table with a number of different objects and hand, wearing the “Millie” watch, reaching for a pie. This image was meant to be perceived as an aesthetic scene that subtly features the endorsed product. The hard-sell picture aimed at drawing the focus on the watch, which was shown in a close-up wrist shot of a hand holding flowers, making the name “Millie” effortlessly observable (see Appendix B).
4
This differentiation was made on the basis of the distinction that marketing practitioners have established, according to which micro-influencers have less than 100.000 followers and macro-influencers have one million followers and above (Barker, 2017; Pavlika, 2017; Pierucci, 2018).
5
The caption read: “Good morning everyone! I hope you are all doing well. I’m about to start my day and I’m wearing my new #Millie watch! Off to today’s adventure”.
Measures
The dependent variables (brand awareness, brand attitude, and purchase intention), and the mediator variable (perceived source credibility) were measured during the online survey.
Brand awareness. The study employed two of the more classical measures of brand awareness: brand recall and brand recognition (Laurent, Kapferer, & Roussel, 1995). Firstly, the participants’ spontaneous awareness was measured, by asking them to name the brand of the watch, without any prompting. The item was coded with 0 (not recalled) and 1 (correctly recalled). Slight variations of the word “Millie” (e.g. “millie”, “Milli”, “Milie”, “Mille”) counted as correct. 61.2% of the participants recalled the brand correctly. In order to measure aided awareness, the participants were presented with seven brands of watches (Rolex, Daniel Wellington, Maurice Lacroix, Cluse, Millie, Omega, Swatch), and were asked to indicate which ones they recognized. Participants who selected “Millie” were coded as 1, and the rest were coded as 0. 11.6% of the participants recognized “Millie”. However, it was expected that the score for brand recognition would be equal, if not higher, to the score of brand recall; the participants who typed the word “Millie” (brand recall) were expected to also select this brand when exposed to the list of brands (brand recognition). Additionally, some participants who did not remember the word “Millie” were expected to nonetheless recognize it when prompted with the name. It is assumed that the question used in the survey “Which of these brands of watches do you know?” was perceived as a deeper sense of familiarity with the brand, rather than mere recognition of the brand name. Therefore, the item of brand recognition was not deemed appropriate to be measured in the study, and consequently, cannot contribute into measuring the latent variable of brand awareness. Only the variable of brand recall was examined instead.
Brand attitude. Participants were asked how they felt about the brand they saw on the Instagram post, based on a published scale (Spears & Singh, 2004) that used seven-point semantic differential items (bad/good, unappealing/appealing, unpleasant/pleasant,
attractive/unattractive, boring/interesting, and likeable/unlikeable). A principal component analysis showed that the six items loaded on two factors (eigenvalues 3.60 and 1.16),
explaining in total 79.23% of the variance. The second factor consisted of the two items that measured attractiveness and likeability, which were reversed in the questionnaire, in order to avoid acquiescence. The respondents may have possibly failed to notice the reversed scale and did not correctly perceive these two items. Therefore, these two variables were not taken into account when computing the scale for brand attitude. A principal component analysis showed that the four remaining items loaded on one factor (eigenvalue 3.03, variance explained 75.83%). The reliability of the scale was good (Cronbach’s alpha= .89). The average score of the four items was used to measure this concept (M = 4.32, SD = 1.17).
Purchase intention. Participants were asked to measure the likelihood that they would buy the evaluated product through a seven-point semantic differential scale
(unlikely/likely, definitely not/definitely, and improbable/probable) that is based on previous literature (Bearden, Lichtenstein, & Teel, 1984). A principal component analysis reported that the three items loaded on one factor (eigenvalue 2.68, explained variance = 89.44%). The scale appeared to be reliable (Cronbach’s alpha= .94). Therefore, the average score of the three items formed the measurement of purchase intention (M = 2.67, SD = 1.45).
Perceived source credibility. The concept of perceived source credibility of the influencer “Cloud nine” was measured based on Ohanian’s study (1990). The scale consisted of 15 items with 7-point semantic differential scales (attractive/unattractive, classy/not classy, beautiful/ugly, elegant/plain, sexy/not sexy, dependable/undependable, honest/dishonest, reliable/unreliable, sincere/insincere, trustworthy/untrustworthy, expert/not an expert,
experienced/inexperienced, knowledgeable/unknowledgeable, qualified/unqualified, skilled/unskilled). A principal component analysis showed that the 15 items loaded on 3 factors. This supports Ohanian’s (1990) study, which claims that source credibility consists of three dimensions (attractiveness, expertise, and trustworthiness). A principal component analysis was conducted for each of the three dimensions; attractiveness (eigenvalue 2.75, explained variance = 55.09%), trustworthiness (eigenvalue 3.31, explained variance =
66.11%), and expertise (eigenvalue 3.57, explained variance = 71.41%). Reliability was good for all dimensions (attractiveness; Cronbach’s alpha= .78, trustworthiness; Cronbach’s
alpha= .86; expertise; Cronbach’s alpha= .90). New scales were constructed for attractiveness (M = 3.81, SD = .92), trustworthiness (M = 4.35, SD = .97), and expertise (M = 4.10, SD = 1.01).
Results Manipulation checks
The manipulation check for the influencer type revealed that participants perceived a clear difference in the number of followers between the two conditions (micro vs macro-influencer), χ2
(5) = 141.13, p < .001. 193 out of 232 participants (90.1%) correctly recalled the number of followers6. The distribution of the participants within the four groups was considered even7 (see Table 1).
6
In the multiple-choice question that they were asked, the correct option for the micro influencer was “from 10.000 to 100.000 followers”, and for the macro influencer the correct option was “from 1 million to 3 million followers”.
7
All 232 participants will be included in the analyses. Separate analysis that exclude the 39 respondents (N = 193) who failed to recognize the manipulation between micro and macro-influencers will be included in footnotes.
An independent samples t-test was conducted to examine whether participants perceived a difference in terms of the advertising appeal of the two stimuli. Results revealed a significant difference between soft-sell (M = 4.36, SD = 1.65) and hard-sell appeal (M = 5.30, SD = 1.49), t(223.21) = 4.55, p <.001, d = 0.60, 95% CI [.53, 1.35]. As expected, the assessment of the prominence of the watch in the stimuli was lower for the soft-sell (M = 4.36, SD = 1.65) than for the hard-sell appeal (M = 5.30, SD = 1.49). Both experimental manipulations were successful.
Randomization checks
The experimental conditions did not differ with respect to age; a one-way ANOVA examined whether the age of the participants was evenly distributed across all four
conditions; F (3, 228) = .32, p = .810. The average age across all conditions was around 26 years old (M = 25.53, SD = 3.34). Additionally, the distribution of the participants’
educational level was measured through a cross-tabulation, and it was shown that it did not significantly differ across conditions, χ2
(18) = 11.28, p = .882. This indicates successful randomization within the sample.
Main analysis
Hypothesis 1. As discussed in the methodology section, brand recall was examined in
the place of brand awareness. The first hypothesis claimed that macro-influencers would elicit greater brand recall compared to micro-influencers. A one-way ANOVA indicated that the macro-influencer (M= .56, SD = .50) did not have a significantly different effect on the audience’s brand recall comparing to the micro-influencer (M= .66, SD = .48), F(1, 230) =
2.21, p = .139. The first hypothesis was rejected8.
Hypothesis 2, that the micro-influencer would be more effective in shaping the
audience’s brand attitude than the macro-influencer was not supported9
. A one-way ANOVA did not find a significantly different effect between micro (M = 4.36, SD = 1.21) and macro-influencers (M = 4.27, SD = 1.22) on the audience’s brand attitude, F(1, 230) = .30, p = .582.
With respect to Hypothesis 3, no significant effect on the audience’s purchase
intention was found when comparing micro (M = 2.79, SD = 1.58) and macro-influencers (M
= 2.54, SD = 1.27). An analysis of variance (one-way ANOVA) did not find a significant
8
No support was found for the first hypothesis for N = 193 either. The macro-influencer post (M= .61, SD = .49) was not found to be more effective than the micro-influencer post (M= .66, SD = .48) in regards to people’s brand recall, F(1, 191) = .66, p = .417.
9
The second hypothesis was also rejected for N = 193. Micro (M = 4.27, SD = 1.18) and macro-influencers (M = 4.27, SD = 1.02) did not significantly differ in regards to the audience’s brand attitude; F(1, 191) = .002, p = .964.
effect of influencer type on purchase intention, F(1, 230) = 1.74, p = .189. The third hypothesis was not supported10.
In order to test the Hypothesis 4, a mediation analysis was conducted using a
PROCESS macro command with the Model 4 (Hayes, 2013) for each of the three dependent variables. A bootstrapping analysis with 5000 samples was used.
Firstly, no significant mediation was found for influencer type on purchase intention through perceived source credibility (trustworthiness; indirect effect = .0001, SE = .001, 95% confidence interval [-.002, .003], attractiveness; indirect effect = -.002, SE = .006, 95% confidence interval [-.019, .001], expertise; indirect effect = .0001, SE = .001, 95% confidence interval [-.002, .003]). Perceived source credibility did not mediate the relationship between type of influencer and purchase intention.
Secondly, the same bootstrapping analysis was conducted to examine the mediation of perceived source credibility with brand recall as the outcome variable. Based on the PROCESS analysis, a non-significant effect of brand recall was extracted through the mediation of trustworthiness; indirect effect = -.001, SE = .002, 95% confidence interval [-.003, .002]. Non-significant effects were also found for attractiveness; indirect effect = .0002,
SE = .004, 95% confidence interval [-.004, .015], and expertise; indirect effect = -.0001, SE =
.002, 95% confidence interval [-.004, .003]. Therefore, perceived source credibility did not mediate the relationship between influencer type and brand recall.
When tested for the relationship between influencer type and brand attitude, perceived source credibility did not serve as a mediator. No significant mediation effects were found for either or the three mediator variables with brand attitude as the outcome variable;
10
For N = 193, the effect on the audience’s purchase intention was not significantly different for micro (M = 2.59, SD = 1.43) and macro-influencers (M = 2.46, SD = 1.27), F(1, 191) = .43, p = .513.
trustworthiness; indirect effect = -.0002, SE = .002, 95% confidence interval [-.004, .001], attractiveness; indirect effect = .002, SE = .006, 95% confidence interval [-.021, .002], expertise; indirect effect = -.0001, SE = .001, 95% confidence interval [-.002, .002].
In sum, no support was found for the fourth hypothesis11. Participants did not perceive a difference in the credibility of the two profiles for none of the three outcome variables; trustworthiness (Mmicro= 4.31, SD =.91; Mmacro = 4.39, SD = 1.03), attractiveness (Mmicro= 3.72, SD =.89; Mmacro = 3.91, SD = .94), and expertise (Mmicro= 4.16, SD =1.04; Mmacro = 4.04,
SD = 1.10). Altogether, perceived source credibility was not proven a successful mediator
(see Appendix C for coefficients).
Hypothesis 5a, examining whether the type of advertising appeal (soft vs hard-sell)
influences the relationship between the type of influencer and the audience’s purchase
intentions, was rejected. The analysis (two-way ANOVA) found no significant main effect of advertising appeal type on purchase intention, F(1, 228) = 2.61, p = .107. No significant interaction effect was found between type of influencer and advertising appeal, on the audience’s purchase intention, F(1, 228) = .18, p = .672.
With respect to Hypothesis 5b, the moderating effect of the type of advertising appeal (soft vs hard sell) on the relationship between type of influencer (micro vs macro) and the audience’s brand attitude was examined. A two-way ANOVA found no significant effect of the advertising appeal type on brand attitude, F(1, 228) = 3.52, p = .062. There was no significant interaction between influencer type and advertising appeal, on the audience’s brand attitude, F(1, 228) = .03, p = .877. Hypothesis 5b was rejected.
11
No support was found for the fourth hypothesis when testing for N = 193. Perceived source credibility (trustworthiness, attractiveness, and expertise) did not mediate the relationship between influencer type and purchase intention, brand recall, or brand attitude.
Since brand recognition was deemed as an inappropriate measure, Hypothesis 5c examined advertising appeal type as a moderator of the relationship between influencer type and brand recall, instead of brand awareness. A two-way ANOVA revealed a significant effect of advertising appeal type on brand recall, F(1, 228) = 5.18, p = .024. The analysis showed that the hard-sell appeal (Mtotal = .68, SD = .47) induced greater brand recall than the soft-sell appeal condition (Mtotal = .54, SD = .50).The interaction between advertising appeal type and influencer type on brand recall was not significant, F(1, 228) = .01, p = .925. Therefore, hypothesis 5c was rejected12. The overall results of the study’s hypotheses are summarized in Figure 2.
12
When testing for N = 193, no moderation effect was found for the three outcome variables; purchase intention, F(1, 189) = .51, p = .478, brand attitude, F(1, 189) = .23, p = .635, or brand recall, , F(1, 189) = .34, p = .562. No main effects of the moderator variables were found either (purchase intention, F(1, 189) = 1.86, p = .175; brand attitude, F(1, 189) = 1.90,
Conclusion and discussion
The aim of this study was to determine whether the influencer type (micro or macro-influencer) has an effect on people’s brand recall, brand attitude, and purchase intention in online endorsements, and whether perceived source credibility could explain such effect. Additionally, the study sought to investigate whether the type of advertising appeal (soft or hard-sell) moderates this effect. In order to answer those questions, five hypotheses were tested through an online experiment.
The first hypothesized path of the model was not supported by the results of the study. This regards to Hypotheses 1 to 3. Firstly, the results revealed that the type of influencer that is endorsing a product/brand, namely a micro or macro influencer, did not have a significant effect on people’s brand recall. Contrary to marketers’ claims (Hatton, 2018; Pavlika, 2017),
which recognized macro-influencers to be more effective in generating brand awareness compared to micro-influencers, no such difference was found in the present study. More specifically, due to their larger number of followers which suggests higher exposure, celebrity influencers have been previously found to induce higher brand awareness than traditional influencers (Djafarova & Rushworth, 2017; Scott, 2011). However, the study’s macro-influencer did not have such differentiated effects from the micro-influencer, which could be due to the fictitious nature of the profile. Other than the number of followers, celebrity influencers have several cues that demonstrate their popularity, such as the number of likes and comments, and an extensive collection of high quality pictures. The lack of such cues on the “Cloud nine” profile may have prevented the audience from perceiving the macro-influencer as a successful celebrity account.
Furthermore, the study sought to find whether the type of influencer has an effect on the audience’s attitude towards the endorsed brand (Hypothesis 2). Micro-influencers were expected to impact brand attitude more positively, due to the fact that they address niche audiences and seem more credible than macro-influencers, who may endorse a large variety of brands, thus lacking honesty (Djafarova & Rushworth, 2017; Cheung, Luo, Sia, & Chen, 2009; Wang, Kao, & Ngamsiriudom, 2017). However, no significance was found between the two types of influencer posts. People being exposed to both the micro and the macro-influencers developed similar brand attitudes.
The last main effect examined was the impact of the influencer type on the audience’s purchase intention (Hypothesis 3), which proved not to be significant. According to
Djafarova and Rushworth’s study (2017), non-traditional celebrities, such as
“Instagrammers” were more powerful in impacting women’s purchase intentions comparing to traditional celebrities. The present study attempted to parallelize macro-influencers with traditional celebrities, due to their millions of followers, while micro-influencers were
described as more similar to peers who share recommendations with a closer circle (Jin, 2018; Tong, Van Der Heide, Langwell, & Walther, 2008). However, this third hypothesis was rejected by the study’s findings.
The lack of support for the first three hypotheses is inconsistent with the claims of marketing professionals, who recognized macro-influencers more appropriate for brand awareness campaigns and micro-influencers more influential on people’s brand attitude and purchase intention (Barker, 2017; Chen, 2016; Morrell, 2017; Pavlika, 2017; Pierucci, 2018; Scott, 2011). Previous studies have found a distinction in terms of one’s number of following in social media (Tong, Van Der Heide, Langwell, & Walther 2008; Utz, 2010; Westerman, Spence, & Van Der Heide, 2012). Accordingly, it was hypothesized that the more followers a SMI has the less credible they are perceived. The fact that this was not found in the study could be due to several reasons.
Firstly, Tong, Van Der Heide, Langwell, and Walther’s study (2008) was conducted on Facebook. According to Kietzmann, Hermkens, McCarthy, and Silvestre (2011),
Facebook’s main functionality is building relationships. Therefore, the number of friends one has on this medium might be a more crucial criterion of determining their credibility in comparison to Instagram followers. Moreover, Westerman, Spence, & Van Der Heide (2012) argued that “too many” Twitter followers resulted in lower judgments of trustworthiness. This did not apply to Instagram in this study. A possible explanation for this might be that Twitter focuses mainly on creating conversation, while Instagram is meant to showcase one’s identity (Kietzmann, Hermkens, McCarthy, & Silvestre, 2011). Therefore, it could be
assumed that certain platforms, such as Facebook and Twitter, centre around creating small communities, where the number of followers can act as a cue that determines the credibility of the profile. Instagram, however, did not appear to produce different effects for different numbers of followers, possibly due to its more inclusive and expressive identity.
Secondly, the present study exposed participants only to one picture of the SMI. The lack of significant results could also be by reason of the one-time exposure. Previous
literature has recognized that the SMIs’ success to influence their audiences can be attributed to the fact that they are perceived approachable and relatable (Abidin 2016; Schau & Gilly 2003). However, in order to manifest their approachability, repeated exposure to their profiles might be essential. A single exposure may give a superficial insight on the SMI, whereas repeated exposure on multiple posts may give more substantial conclusions on the SMI’s personality and honesty. This is in line with Katz and Lazarsfeld (1955), who argued that opinion leaders can exert influence on a larger audience. Their two-step flow theory stipulated that the audience knew the opinion leaders in person and had established a relationship with them. This is in contrast with the present study, where the audience only experienced a single exposure to the influencer and had not established a rapport with them. This differentiation could be a potential explanation for the lack of support in regards to the different effects between micro and macro-influencers.
Furthermore, no significant indirect effect was found when examining whether trustworthiness, attractiveness, and expertise mediated the relationship between influencer type and purchase intention, brand recall, and brand attitude. Perceived source credibility could not explain the effects between the type of influencer and people’s brand recall, brand attitude, and purchase intention, since no such main effects were found.
The lower path of the assumed model expected to uncover a moderation effect such that a soft-sell appeal would be more effective for macro-influencers raising people’s brand awareness, while a hard-sell appeal would be more effective for micro-influencers affecting people’s purchase intention and brand attitude. The fact that no support was found for this hypothesis demonstrates that both the soft and hard-sell appeal post may have caused the same amount of skepticism in regards to the SMI promoting the watch (Hwang & Jeong,
2016; Lu, Chang, & Chang, 2014). The only significant result in regards to the moderator was that the hard-sell appeal induced greater brand recall, which is due to the fact that in this condition the watch was much more prominent in the picture and aided the brand name to stand out.
Limitations and future research
When building off of this study in the future, there are some important limitations and recommendations to consider. Firstly, the convenience sampling of the research might have influenced the objectiveness of the data collection. Additionally, this collection adopted a cross-sectional approach, which limited the measurement of the effects at one point in time. Since the effectiveness of SMIs lies in the fact that they build trusting relationships with their audience, future research could choose longitudinal approaches that investigate such effects in a longer period of time, increasing, thus, the internal validity of the study. Another approach for this limitation would be using real influencers and testing the effects that they have on their followers. This would already ensure a trusting relationship between the SMI and his/her audience, allowing, thus, a more in-depth examination of the SMI’s perceived credibility and influencing power.
Furthermore, the sample only focused on young women, and therefore, the findings can only be generalized to this part of the population. Future studies could examine similar hypotheses on men or older participants, assessing if these populations would produce different effects. Moreover, the present research used a fictitious brand, in order to avoid predetermined attitudes towards an existing brand. This decreased the external validity of the study, which could be improved by examining real brands.
The present study exposed participants to Instagram, limiting the findings to this platform. Since Kapitan and Silveira (2016) have argued that different social media platforms
may induce different effects, future research could focus on different platforms. For example, YouTube is an important medium for promoting brands through product reviews. YouTubers upload “monthly favourites” videos, where they talk about their favourite products/brands. Studies that aim to examine the influence of different types of SMIs and the effects they have on their audience could investigate such YouTube videos.
Despite the lack of significant findings between SMIs’ posts in the present study, the differentiation between micro and macro-influencers may still be valuable for marketers. More specifically, while the distinction between the number of followers was successfully manipulated in the present study, it was lacking additional context. As marketers have suggested, micro and macro-SMIs differ not only in their number of followers, but also in their content; macro-influencers adopt a broader range of topics in their profiles, while micro-influencers focus on more specific and niche material (Barker, 2017; Mourdoukoutas & Siomkos, 2009). Hence, the variety of macro-influencers may increase brand awareness, and the relatability of micro-influencers may induce greater brand attitude and purchase intention. Therefore, when academics or marketers wish to examine differentiated effects between types of influencers, they should take into account not only the number of followers, but also the content of the profile. In conclusion, longitudinal studies with a more detailed
operationalization of micro and macro-influencers could eliminate errors and examine the SMIs’ effects on consumers within the trending topic of influencer marketing.
References
Aaker, D. A. (1991). Managing brand equity. New York, NY: The Free Press.
Abidin, C. (2016). Visibility labour: Engaging with influencers’ fashion brands and #OOTD advertorial campaigns on Instagram. Media International Australia 161(1), 86–100. Abzari, M., Ghassemi, R. A., & Vosta, L. N. (2014). Analysing the effect of social media on
brand attitude and purchase intention: The case of Iran Khodro company. Procedia-
Social and Behavioral Sciences, 143, 822-826. doi:10.1016/j.sbspro.2014.07.483
Ajzen, I. (1985). From intentions to action: A theory of planned behaviour. Action-control:
From cognitions to behaviours. Heidelberg: Springer.
Alden, D. L., Steenkamp, J.-B. E. M., & Batra R. (1999). Brand positioning through
advertising in Asia, North America, and Europe: The role of global consumer culture.
Journal of Marketing, 63(1), 75–87.
Aslam, S. (2018, January 1). Instagram by the numbers: stats, demographics & fun facts. Omnicore Agency. Available at: https://www.omnicoreagency.com/instagram-statistics/ Barker, S. (2017, September 29). Using micro-influencers to successfully promote your
brand. Forbes. Retrieved from:
https://www.forbes.com/sites/forbescoachescouncil/2017/09/29/using-micro-influencers-to-successfully-promote-your-brand/#2444eb8b1763
Bearden, W. O., Lichtenstein, D. R., & Teel, J. E. (1984). Comparison price, coupon, and brand effects on consumer reactions to retail newspaper advertisements. Journal of
Retailing, 60(2), 11-34.
Berger, J., & Keller Fay Group (2016). Research shows micro-influencers have more impact
than average consumers. Experticity. Retrieved from:
http://go2.experticity.com/rs/288-azs-731/images/experticity-kellerfaysurveysummary_.pdf
Bhatt, N., Jayswal, R. and D Patel, J. (2013). Impact of celebrity endorser's source credibility on attitude towards advertisements and brands. South Asian Journal of Management,
20(4), 74-95.
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.
http://dx.doi.org/10.1016/j.intmar.2016.12.002
Brouwer, B. (2017). Why brands are investing more into influencer marketing in 2017.
Econtent, 40(3), 32-33.
Chen, Y. (2016, April 8). The rise of ‘micro-influencers’ on Instagram. Digiday. Retrieved from: http://digiday.com/marketing/micro-influencers/
Cheung, M., Luo, C., Sia, C., & Chen, H. (2009). Credibility of electronic word-of- mouth: Informational and normative determinants of on-line consumer
recommendations. International Journal of Electronic Commerce, 13(4), 9-38. Chu, W., Gerstner, E., & Hess, J. D. (1995). Costs and benefits of hard-sell. Journal of
Marketing Research, 32(1), 97-102.
Curalate. (2016, July 27). Influencer marketing: The goof, the bad and the downright ugly. Retrieved from: https://www.curalate.com/blog/influencer-marketing-fails/
Djafarova, E., & Rushworth, C. (2016). Exploring the credibility of online celebrities’ Instagram profiles in influencing the purchase decisions of young female users.
Computers in Human Behavior, 68, 1-7. doi: 10.1016/j.chb.2016.11.009
Duffy, B. E., & Hund, E. (2015). “Having it all” on social media: Entrepreneurial femininity and self-branding among fashion bloggers. Social Media + Society, 1(2). doi:
2056305115604337.
Eirinaki, M., Monga, S.P.S., Sundaram, S. (2012). Identification of influential social networkers. International Journal of Web Based Communities, 8(2), 136–158.
eMarketer (2015, November 18). Are sponsored social posts the most effective marketing channel? Retrieved from: https://www.emarketer.com/Article/Sponsored-Social-Posts-Most-Effective-Marketing-Channel/1013242
Fishbein, M., & Ajzen, I. (1975). Belief, attitude, intention and behavior: An
introduction to theory and research. Reading, MA: Addison-Wesley Publishing Co.
Freberg, K., Graham, K., McGaughey, K. & Freberg, L. (2011). Who are the social media influencers? A study of public perceptions of personality. Public Relations Review,
37(1), 90-92.
Gardner, S. (2005). Buzz marketing with blogs for dummies. New York, NY: Wiley Publishing Inc.
Gladwell, M. (2005). The tipping point. New York, NY: Back Bay Books.
Griffith, D. A., & Chen, Q. (2004). The influence of virtual direct experience (Vde) on online ad message effectiveness. Journal of Advertising, 33(1), 55-68. doi:
10.1080/00913367.2004.10639153
Hatton, G. (2018, February 13). Micro influencers vs macro influencers. Social Media Today. Retrieved from: https://www.socialmediatoday.com/news/micro-influencers-vs-macro-influencers/516896/
Hayes, A. F. (2013). Introduction to mediation, moderation, and conditional process
analysis: a regression-based approach. New York, NY: The Guilford Press.
Hu, Y., Manikonda, L., & Kambhampati, S. (2014). What we Instagram: a first analysis of
Instagram photo content and user types. Proceedings of AAAI International
Conference on Web and Social Media. Retrieved from:
https://www.aaai.org/ocs/index.php/ICWSM/ICWSM14/paper/viewFile/8118/8087 Hutter, K., Hautz, J., Dennhardt, S., & Fuller, J. (2013). The impact of user interactions in
Facebook. Journal of Product & Brand Management, 22(5-6), 342-351. https://doi.org/10.1108/JPBM-05-2013-0299
Hwang, Y., & Jeong, S.-H. (2016). “This is a sponsored blog post, but all opinions are my own”: The effects of sponsorship disclosure on responses to sponsored blog posts.
Computers in Human Behavior, 62, 528-535.
http://dx.doi.org/10.1016/j.chb.2016.04.026
Jin, S. V. (2018). “Celebrity 2.0 and beyond!”. Effects of Facebook profile sources on social networking advertising. Computers in Human Behaviour, 79, 154-168.
https://doi.org/10.1016/j.chb.2017.10.033
Jin, S., & Phua, J. (2014). Following celebrities' Tweets about brands: The impact of Twitter-based electronic word-of-mouth on consumers' source credibility perception, buying intention, and social identification with celebrities. Journal of Advertising, 43(2), 181-195.
Kapitan, S., & Silvera, D. (2016). From digital media influencers to celebrity endorsers: attributions drive endorser effectiveness. Marketing Letters, 27(3), pp.553-567. Katz, E., & Lazarsfeld, P. F. (1955). Personal Influence: The part played by people in the
flow of mass communications. New York, NY: Free Press.
Keller, K. L. (1993). Conceptualizing, measuring, managing customer-based brand equity.
Journal of Marketing, 57(1), 1–22.
Keller, K. L. (2003). Strategic brand management: Building, measuring, and managing
brand equity. Upper Saddle River, NJ: Pearson.
Kietzmann, J., Hermkens, K., McCarthy, I. P., & Silvestre, B. (2011). Social media? Get serious! Understanding the functional building blocks of social media. Business
Horizons, 54(3), 241-251. doi: 10.1016/j.bushor.2011.01.005
Laurent, G., Kapferer, J. N., & Roussel, F. (1995). The underlying structure of brand awareness scores. Marketing Science, 14(3), 170-179.
Li, H., & Lo, H. (2015). Do You Recognize Its Brand? The Effectiveness of Online In-Stream Video Advertisements. Journal of Advertising, 44(3), 208-218, doi: 10.1080/00913367.2014.956376.
Lu, L. C., Chang, W. P., & Chang, H. H. (2014). Consumer attitudes toward blogger's sponsored recommendations and purchase intention: the effect of sponsorship type, product type, and brand awareness. Computers in Human Behavior, 34, 258-266. ‘Mediakix’. (2016, June 29). What are Micro-Influencers: Definitions, trends & advantages.
Retrieved from: http://mediakix.com/2016/06/micro-influencers-definition-marketing/ Michaelidou, N., Siamagka, N. T., & Christodoulides, G. (2011). Usage, barriers and
measurement of social media marketing: An exploratory investigation of small and medium B2B brands. Industrial Marketing Management, 40(7), 1153-59.
Mitchell, A. & Olson, J. (1981). Are product attribute beliefs the only mediator of advertising effects on brand attitude? Journal of Marketing Research, 18(3), 318-332.
Mourdoukoutas, P. and Siomkos, G.J. (2009). The Seven principles of WOM and buzz
marketing: Crossing the tipping point. Germany: Springer-Verlag Berlin and
Heidelberg GmbH & Co. K
Mueller, B. (1987). Reflections of culture: An analysis of Japanese and American advertising appeals. Journal of Advertising Research, 27(3), 51–59.
Muntinga, D. G., Moorman, M., & Smit, E. G. (2011). Introducing COBRAs: Exploring motivations for brand-related social media use. International Journal of
Nisbet, M. C., & Kotcher, J. E. (2009). A two-step flow of influence? : Opinion-leader campaigns on climate change. Science Communication, 30(3), 328-354.
Ohanian, R. (1990). Construction and validation of a scale to measure celebrity endorsers’ perceived expertise, trustworthiness, and attractiveness. Journal of Advertising, 19(3), 39-52.
Okazaki, S., Mueller, B., & Taylor, C. R. (2010). Measuring soft-sell versus hard-sell advertising appeals. Journal of Advertising, 39(2), 5-20. doi: 10.2753/JOA0091-3367290201
Pavlika, H. (2017, April 11). Which is best: Micro, power middle or celebrity influencers? Collective bias. Retrieved from: https://collectivebias.com/blog/2017/04/which-is-best-micro-power-middle-or-celebrity-influencers/
Pierucci, S. (2017, January 18). Why micro-influencer marketing is ‘The Game’ in 2018. The Startup. Retrieved from: https://medium.com/swlh/why-micro-influencer-marketing-is-the-game-in-2018-fdeda0993c36
Radder, L., & Huang, W. (2008). High-involvement and low-involvement products: A
comparison of brand awareness among students at a South African university. Journal
of Fashion Marketing and Management: An International Journal, 12(2), 232-243.
https://doi.org/10.1108/13612020810874908
Ranga, M. & Sharma, D. (2014). Influencer marketing- A marketing tool in the age of social media. Abhinav Publications, 3(8), 16-21.
Rogers, E. (1962). Diffusion of Innovations. New York, NY: Free Press.
Rogers, E. M., & Cartano, D. G. (1962). Methods of measuring opinion leadership. Public
Opinion Quarterly, 26, 435-441.
Rossiter, J. R. (2014). ‘Branding’ explained: Defining and measuring brand awareness and brand attitude. Journal of Brand Management, 21, 7-8.
https://doi-org.proxy.uba.uva.nl:2443/10.1057/bm.2014.33
Roy, A. (2015). Status and practice of influencer engagement. Augure Report. Available at: http://www.augure.com/blog/state-influencer-engagement-20150618
Schau, J. H., & Gilly, M. C. (2003). We are what we post? Self-presentation in personal web space. Journal of consumer research, 30(3), 385-404.
Scott, D. (2011). The new rules of marketing and PR: How to use social media, online
video, mobile applications, blogs, news releases and viral marketing to reach buyers direct. New York, NY: Wiley.
Scott, K. (2015, November 4). 'Instafamous' teen reveals social media's ugly truth. ABC News. Retrieved from: http://www.abc.net.au/news/2015-11-03/instagram- personality-essena-o'neill-reveals-social-media-truth/6908270.
Sheldon, P., & Bryant, K. (2016). Instagram: Motives for its use and relationship to narcissism and contextual age. Computers in Human Behavior, 58, 89-97. 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. Spry, A., Pappu, R., & Cornwell, B. T. (2011). Celebrity endorsement, brand
credibility and brand equity. European Journal of Marketing, 45(6), 882-909. Statista. (2018). Number of social media users worldwide from 2010 to 2021 (in billions).
Statista. Retrieved from: https://www.statista.com/statistics/278414/number-of-worldwide-social-network-users/
Sun, B., & Morwitz, V. G. (2010). Stated intentions and purchase behaviour: A unified model. International Journal of Research in Marketing, 27(4), 356-366.
http://doi.org.proxy.uba.uva.nl:2048/10.1016/j.ijresmar.2010.06.001
Tong, S. T., Van Der Heide, B., Langwell, L., & Walther, J. B. (2008). Too much of a good thing? The relationship between number of friends and interpersonal impressions on
Facebook. Journal of Computer-Mediated Communication, 13, 531–549.
Utz, S. (2010). Show me your friends and I will tell you what type of person you are: How one’s profile, number of friends, and type of friends influence impression formation on social network sites. Journal of Computer-Mediated Communication, 15(2), 314-335. Uzunoglu, E., & Kip, S. M. (2014). Brand communication through digital influencers:
leveraging blogger engagement. International Journal of Information Management,
34(5), 592-602. https://doi.org/10.1016/j.ijinfomgt.2014.04.007
Wang, Y., Hsiao, S.-H., Yang, Z., & Hajli, N. (2016). The impact of sellers’ social influence on the co-creation of innovation with customers and brand awareness in online communities. Industrial Marketing Management, 54, 56-70.
https://doi.org/10.1016/j.indmarman.2015.12.008
Wang, S., Kao, G., & Ngamsiriudom, W. (2017). Consumers' attitude of endorser credibility, brand and intention with respect to celebrity endorsement of the airline sector.
Journal of Air Transport Management, 60, 10-17.
Westerman, D., Spence, P. R., & Van Der Heide, B. (2012). A social network as information: The effect of system generated reports of connectedness on credibility on Twitter.
Computers in Human Behavior, 28, 199-206. doi:10.1016/j.chb.2011.09.001
WOMMA. (2016, April 28). Research shows micro-influencers are marketing’s biggest
opportunity. WOMMA. Retrieved from:
Appendix A
Figure 1. Micro-influencer’s hard-sell stimuli (pilot study).
Figure 3. Macro-influencer’s hard-sell stimuli (pilot study).
Appendix B
Figure 1. Condition 1; micro-influencer with hard-sell advertising appeal.
Figure 3. Condition 3; micro-influencer with soft-sell advertising appeal.
