ArticlePDF Available

Disclosing sponsored Instagram posts: the role of material connection with the brand and message-sidedness when disclosing covert advertising

Authors:

Abstract and Figures

Regulation prescribes that sponsored social media posts should clearly disclose their material connection with brands. However, research on the impact of such disclosures is limited. This study used an experimental 4 × 2 between-subjects design (N = 414), reflecting different types of material connections between the influencer and the brand and two kinds of message sidedness, to investigate how the sponsorship disclosure affects consumers' responses to sponsored Instagram posts. Findings show that including a sponsorship disclosure (compared to no disclosure) negatively affects brand attitude through enhanced ad recognition, which activates ad skepticism, which, in turn, negatively affects the influencer's credibility. Further, results show a significant moderated mediation effect in that source's credibility and, consequently, brand attitude was only negatively affected when the influencer used a one-sided message and not when the message was two-sided. Also, influencers who post genuine product recommendations and thus have no commercial relationship with the brand, do well by explicitly mentioning this. Iincluding a statement that a post is not sponsored, rather than leaving thisit unclear whether their post is sponsored, may generate more positive brand responses through lowered ad recognition and skepticism.
Content may be subject to copyright.
Full Terms & Conditions of access and use can be found at
https://www.tandfonline.com/action/journalInformation?journalCode=rina20
International Journal of Advertising
The Review of Marketing Communications
ISSN: 0265-0487 (Print) 1759-3948 (Online) Journal homepage: https://www.tandfonline.com/loi/rina20
Disclosing sponsored Instagram posts: the role of
material connection with the brand and message-
sidedness when disclosing covert advertising
Marijke De Veirman & Liselot Hudders
To cite this article: Marijke De Veirman & Liselot Hudders (2019): Disclosing sponsored Instagram
posts: the role of material connection with the brand and message-sidedness when disclosing
covert advertising, International Journal of Advertising, DOI: 10.1080/02650487.2019.1575108
To link to this article: https://doi.org/10.1080/02650487.2019.1575108
Published online: 11 Feb 2019.
Submit your article to this journal
View Crossmark data
Disclosing sponsored Instagram posts: the role of
material connection with the brand and message-
sidedness when disclosing covert advertising
Marijke De Veirman
a
and Liselot Hudders
b
a
Department of Communication Sciences, Ghent University, Ghent, Belgium;
b
Department of
Communication Sciences and Department of Marketing, Ghent University, Ghent, Belgium
ABSTRACT
Regulation prescribes that sponsored social media posts should
clearly disclose their material connection with brands. However,
research on the impact of such disclosures is limited. This study
used an experimental 4 2 between-subjects design (N¼414),
reflecting different types of material connections between the
influencer and the brand and two kinds of message sidedness, to
investigate how the sponsorship disclosure affects consumers
responses to sponsored Instagram posts. Findings show that
including a sponsorship disclosure (compared to no disclosure)
negatively affects brand attitude through enhanced ad recogni-
tion, which activates ad skepticism, which, in turn, negatively
affects the influencers credibility. Further, results show a signifi-
cant moderated mediation effect in that sources credibility and,
consequently, brand attitude was only negatively affected when
the influencer used a one-sided message and not when the mes-
sage was two-sided. Also, influencers who post genuine product
recommendations and thus have no commercial relationship with
the brand, do well by explicitly mentioning this. Iincluding a
statement that a post is not sponsored, rather than leaving thisit
unclear whether their post is sponsored, may generate more posi-
tive brand responses through lowered ad recognition
and skepticism.
ARTICLE HISTORY
Received 2 April 2018
Revised 17 December 2018
Accepted 8 January 2019
KEYWORDS
Influencer marketing;
Instagram; sponsorship
disclosure; mes-
sage sidedness
Introduction
The proliferation of digital advertisements has encouraged consumers to ignore ads or
avoid them by using technology such as ad blockers. This practice has significantly
eroded the value of traditional brand-driven advertisements (Petty and Andrews 2008;
Kelley, Kerr, and Drennan 2010). Consumers are empowered to avoid unwanted adver-
tising, and they aim to decide for themselves which information, commercial or not, is
needed and used in their decision-making process. At the same time, consumers are
well-informed that traditional advertisements only provide arguments that put the
promoted products in a favorable light, reducing their credibility (De Veirman,
CONTACT Marijke De Veirman marijke.deveirman@ugent.be Department of Communication Sciences, Ghent
University, Korte Meer 11, 9000 Ghent, Belgium
ß2019 Advertising Association
INTERNATIONAL JOURNAL OF ADVERTISING
https://doi.org/10.1080/02650487.2019.1575108
Cauberghe, and Hudders 2017; Fransen et al. 2015). Therefore, they value and trust
the opinions of peers more, which is generally defined as (electronic) word-of-mouth
(eWOM), and use these opinions to guide their purchase decisions. This eWOM is per-
ceived to be more genuine and unbiased and is widely available online (e.g. online
review platforms as TripAdvisor or Yelp) and on social media (e.g. Facebook,
Instagram, or Twitter), in particular (Hennig-Thurau et al. 2004; Schindler and Bickart
2005). As awareness has been raised about the value and persuasiveness of eWOM
among advertisers, eliciting eWOM has become an important marketing strategy
(Goldsmith and Clark 2008).
One way to do so is to utilize influencers to endorse brands via eWOM on social
media (De Veirman, Cauberghe, and Hudders 2017). Influential social media users with
numerous followers typically receive products for free (i.e. material compensation) or
are being paid (i.e. financial compensation) to include and recommend brands on their
social media profiles and, in this way, shape their followersopinions (De Veirman,
Cauberghe, and Hudders 2017; Phua, Jin, and Kim 2017). These sponsored posts can
be considered a form of native advertising, defined as any paid advertising that takes
the specific form and appearance of editorial content from the publisher itself
(Wojdynski and Evans 2016). Indeed, sponsored Instagram posts mimic and blend with
organic, non-sponsored posts, appearing in usersnews feeds without interrupting
their social media experience, unlike non-integrated social media ads (Abidin 2016;
Wojdynski and Evans 2016; De Veirman, Cauberghe, and Hudders 2017). Moreover,
influencers may remain vague about their commercial relationship with the sponsoring
brand. Therefore, along with the emergence and popularity of sponsored social media
posts, there is a growing concern that consumers might not recognize these spon-
sored posts as paid ads but rather as highly credible word-of-mouth endorsements,
which could lead to deceptive persuasion (Kaikati and Kaikati 2004; Fransen et al.
2015; Evans et al. 2017; De Veirman, Cauberghe, and Hudders 2017; Evans et al. 2017;
Boerman, Van Reijmersdal, and Neijens 2015). As a result, a clear and prominent dis-
closure of their commercial nature is recommended to comply with The Federal Trade
Commissions (FTC) guides concerning native advertising (2015) and EU rules applic-
able to online native advertising [Interactive Advertising Bureau (IAB) 2016].
These disclosures may trigger consumersability to recognize the commercial con-
tent and subsequently activate a critical reflection of the post, skills that have been
defined as advertising literacy (Hudders et al. 2017). Indeed, prior studies found that a
clear disclosure is necessary to trigger advertising recognition, which, in turn, may
reduce resistance to ad persuasion. Evans et al. (2017) recently investigated the impact
of disclosure language in sponsored Instagram posts and found that a clear disclosure
(PaidAdor Sponsored), produced more advertising recognition compared to a vague
abbreviation (SP) or no disclosure. A clear disclosure had a positive effect on advertis-
ing recognition and disclosure memory, which could negatively impact brand attitudes
and userswillingness to share a post or advertisement. Similarly, Wojdynski and Evans
(2016) found that using sponsoredor advertisingin sponsored news stories online
led to greater advertising recognition than nebulous disclosure language, which nega-
tively affected brand attitudes. The same tendency was discovered in blogs, where
van Reijmersdal et al. (2016) found that consumersattitudes regarding the product
2 M. DE VEIRMAN AND L. HUDDERS
being reviewed and likelihood to purchase were lowered as a result of increased
advertising literacy in response to clear disclosures. Our study builds on these findings
and investigates the role of disclosing the type of commercial relationship between
the influencer and the brand (i.e. financial versus material compensation) in sponsored
posts and compares it with a non-sponsorship disclosure and a non-disclosed post. In
particular, we inquire into how differently disclosed sponsored Instagram posts affect
brand attitude. Moreover, we investigate whether and how the content of a (spon-
sored) post, more specifically message sidedness, impacts consumersresponses to
sponsorship disclosures.
Although disclosures are indispensable in a transparent and ethical advertising pol-
icy, advertisers (and influencers) may be reluctant to use them, as most previous
research shows that disclosures negatively affect brand attitudes and even influencer
evaluations. Accordingly, the current study will examine whether message sidedness
may counter the negative impact of advertising disclosures on brand evaluations.
Although consumers assign greater weight to negative information in their decision
making, two-sided messages may turn out positive for the sponsoring brand (Shoham,
Moldovan, and Steinhart 2017). This is because message sidedness may reassure con-
sumers that the influencer is credible because a more nuanced view of the product
has been given (Uribe, Buzeta, and Vel
asquez 2016; Eisend 2006,2007). This is of par-
ticular interest as advertising disclosures may challenge credibility perceptions and,
consequently, attitudes toward the sponsoring brand (Friestad and Wright 1994;
Campbell and Kirmani 2008; Obermiller and Spangenberg 1998; van Reijmersdal,
Neijens, and Smit 2010).
The obtained novel insights advance our theoretical understanding of how the
disclosure of covert advertising tactics affects consumer responses and the role of
advertising literacy in this process. Furthermore, they provide insights into the con-
sequences of the FTC guidelines, which state that influencers should clearly and
conspicuously disclose any material connectionwith advertisers, which is mostly
monetary payment or free product (FTC 2015). Moreover, findings on message
sidedness are particularly important, as there has been little research to date about
its impact in covert advertising formats, while the impact of message sidedness in
traditional advertising has been widely investigated. Furthermore, while the link
between two-sidedness and source credibility has been well established, it is more
tenuous whether this link may also restore perceptions of credibility and, conse-
quently, brand attitudes and evaluations when they have been challenged by a dis-
closure indicating that the message is, in fact, an ad (Eisend 2006). To conclude, our
study is the first to investigate the impact of explicitly disclosing non-sponsorship in
social media posts, a hashtag that yields almost 200,000 search results on
Instagram, alone.
We conducted an experimental study based on a 4 2 between-subjects design
(sponsorship disclosure: material compensation versus financial compensation versus
not sponsored versus no disclosure message sidedness: one-sided versus two-sided)
among Instagram users to test our hypotheses.
We focused on Instagram, as this is currently the most used platform for influencer
marketing [World Federation of Advertisers (WFA) 2018]. In 2017, there were
INTERNATIONAL JOURNAL OF ADVERTISING 3
14.5 million brand-sponsored influencer posts on Instagram, and this number is
expected to double by 2019 (Statista 2018a). Moreover, now that Instagram officially
has more than 1 billion monthly active users, the need for proper disclosure of
Instagram influencer marketing is more relevant than ever.
Theoretical background and hypothesis development
Figure 1 presents our conceptual model. We focus on how different sponsorship dis-
closure types affect attitudes toward a brand, which is consumersoverall evaluation
of a brand (Wu and Wang 2011) and is a primary element of customer-based equity
(Lane and Jacobson 1995; Morgan and Hunt 1994). Because it is an important pre-
dictor of consumer behavior and brand preference, in particular, reinforcing brand atti-
tude has been recognized as one of the most essential marketing objectives (Wu and
Wang 2011). Moreover, we take into account the possible moderating effect of mes-
sage sidedness.
The aim of this study is fourfold: first, we aim to unravel the process through
which sponsorship disclosures (#sponsored) are likely to affect consumersevalua-
tions of the influencer and brand attitudes. Second, we will take a closer look at
influencer posts about genuine product recommendations and investigate whether
consumer responses differ when the influencer explicitly mentions he or she is not
sponsored compared to when a post does not include any disclosure. Third,
we investigate the role of disclosing the type of material connection (i.e. material
versus financial compensation disclosure) and compare it to a non-sponsorship dis-
closure. To conclude, we will take into account the moderating impact of mes-
sage sidedness.
The effects of advertising disclosures in sponsored social media posts
Digital advertising is often fully integrated into media content, disguising its commer-
cial nature (Hudders et al. 2017). For instance, sponsored social media posts resemble
genuine posts and are difficult to identify as commercial content. Accordingly, product
Figure 1. Conceptual model.
4 M. DE VEIRMAN AND L. HUDDERS
recommendations posted by influencers are perceived to be authentic, honest, and
sincere recommendations from influential people who are perceived to be experts in
the field (De Veirman, Cauberghe, and Hudders 2017). These recommendations are
therefore much more persuasive than brand-generated messages, which are perceived
as being biased (Bickart and Schindler 2001).
According to FTC guidelines (FTC 2015) and EU legislation (IAB 2006), consumers
have the right to know when they are exposed to commercial content. This know-
ledge will help them to cope with persuasion critically. Following the persuasion
knowledge model (Friestad and Wright 1994), when people recognize a persuasive
attempt, they can use their knowledge about persuasion to process the attempt critic-
ally and decide either to be persuaded or to resist the persuasion. More specifically, all
individualsknowledge and skills, specifically regarding advertising, can be defined as
advertising literacy (Hudders et al. 2017). The phrase refers to individualscapacity to
recognize advertising, understand its persuasive intent, and have insight into persua-
sive tactics. In addition, it refers to individualsknowledge about detecting and con-
trolling emotions that are aroused by advertisements and their ability to evaluate ads
on perceived fairness and appropriateness (Hudders et al. 2017). This knowledge,
which should be activated when confronted with an advertising message, helps an
individual to process the advertising critically. An important precondition for the
proper activation of this knowledge is that people are aware that they have been con-
fronted with advertising. This awareness is, however, difficult when exposed to covert
or native advertising that resembles editorial content. Clear advertising disclosure
should, therefore, help consumers recognize the commercial nature of a sponsored
post and cope with persuasion (e.g. Boerman, van Reijmersdal, and Neijens 2012; van
Reijmersdal et al. 2015,2016).
Indeed, recent research suggests that sponsorship disclosures may help individuals
recognize the commercial intent of a sponsored social media post (Boerman,
Willemsen, and Van der Aa 2017; Evans et al. 2017). When consumers are made aware
of the persuasive nature of a message through disclosure, resistance to the persuasion
attempt is likely to occur (Main, Dahl, and Darke 2007). As a result, critical feelings
such as ad skepticism may be instigated (Boerman, Van Reijmersdal, and Neijens 2012;
Rozendaal, Buijzen, and Valkenburg 2011). The latter is supported by reactance theory
(Brehm and Brehm 1981), which suggests that consumers do not want to be manipu-
lated and attach great importance to their autonomy and independence. When con-
sumers become aware of a persuasion attempt, for instance by disclosing sponsorship,
they develop reactance, as they feel threatened in their freedom to make their own
decisions. Therefore, disclosing sponsorship may negatively affect brand responses
through a critical reflection on the ad (e.g. Campbell, Mohr, and Verlegh 2013;
Wojdynski and Evans 2016; van Reijmersdal et al. 2016).
Two studies have investigated whether sponsorship disclosure can make sponsored
social media posts more transparent. In particular, Boerman, Willemsen, and Van Der
Aa (2017) found that a sponsoreddisclosure added to a sponsored Facebook post
from a celebrity increased individualsability to identify this post as advertising. This
label, in turn, negatively affected eWOM intentions through decreased trust in the
message. Similarly, Evans et al. (2017) found that including disclosure in the form of a
INTERNATIONAL JOURNAL OF ADVERTISING 5
hashtag #sponsoredin the caption of a sponsored post on Instagram is effective in
increasing consumersrecognition of the post as advertising. This label also lowered
brand attitudes and userswillingness to share the post.
When consumers encounter a social media post that displays a brand, they may
draw upon their advertising literacy to evaluate the influencers underlying motives for
incorporating a certain brand in an Instagram post. Inferences of ulterior motives may
result in less favorable evaluations of the influencer (Vonk 1998, Campbell and Kirmani
2000). When a sponsorship disclosure is included, any suspicion of ulterior motives
turns into certainty. A sponsorship disclosure leaves no doubt about the reason why
an influencer depicts a brand in a social media post. An advertising disclosure may
therefore lead consumers to think that the influencer is biased, which may negatively
impact the influencers credibility (Hwang and Jeong 2016; Uribe, Buzeta, and
Vel
asquez 2016) and consequently result in decreasing attitudes towards the brand
(Friestad and Wright 1994; Campbell and Kirmani 2008; Obermiller and Spangenberg
1998; van Reijmersdal, Neijens, and Smit 2010). For instance, Uribe, Buzeta, and
Vel
asquez (2016) found that explicitly disclosing a blogs intent to advertise negatively
impacted the perceived sources credibility and intentions. Similar, Janssen, van
Sprang, and Fransen (2017) recently found that the negative effects of disclosures in
sponsored blogs on brand attitude and purchase intention could be explained by a
decrease in the perceived credibility of the influencer. Therefore, we posit that when a
sponsorship disclosure is included in a sponsored social media post, it will likely insti-
gate consumers to question the influencers credibility and, consequently, negatively
affect their attitudes toward the sponsoring brand.
Drawing on the persuasion knowledge model (Friestad and Wright 1994) and react-
ance theory (Brehm and Brehm 1981), we propose that sponsorship disclosure is likely
to help consumers in detecting the commercial nature of a sponsored Instagram post.
The realization that an Instagram post has a commercial purpose will activate a skep-
tical evaluation of the post, which will lower the influencers credibility and ultimately
result in a negative evaluation of the brand. Accordingly, we suggest a sequential
mediation path explaining the effects of a sponsorship disclosure on brand attitudes
through ad recognition, ad skepticism, and source credibility:
Hypothesis 1: A disclosed sponsored Instagram post will lead to lower brand attitudes
compared to a non-disclosed Instagram post through enhanced ad recognition, activating
ad skepticism, and lowering perceptions of influencer credibility.
The effects of explicitly disclosing non-sponsored influencer posts
Due to the increase of sponsored influencer posts on Instagram, influencers tend to
explicitly mention if a post including a brand is a genuine recommendation by includ-
ing the hashtag #notsponsored. This non-sponsorship disclosure serves as a cue that
confirms the Instagram post is an expression of genuine liking of the brand and the
influencer has altruistic motives in sharing opinions with his or her followers. In other
words, if consumers infer that the influencer has no commercial motive, they are likely
to curtail their use of advertising literacy because it does not apply to the situation
(Friestad and Wright 1994). This difference can be explained by Kelleys(1973)
6 M. DE VEIRMAN AND L. HUDDERS
attribution theory, which holds that consumersidea of why someone is sharing a
message influences how the message is received. A non-sponsorship disclosure leaves
no doubt about the influencers true intentions and will, therefore, most likely be per-
ceived as highly credible eWOM, which may induce more positive reactions toward
the brand compared to a non-disclosed post, which makes possible brand interference
unclear (Campbell and Kirmani 2008). Therefore, we expect that:
Hypothesis 2: A non-sponsorship disclosure in an Instagram post will lead to higher brand
attitudes compared to a non-disclosed Instagram post. We expect this effect to be
sequentially mediated by lowered ad recognition and ad skepticism and higher
perceptions of source credibility.
The role of the material connection between the influencer and
the brand
Influencers should clearly and conspicuously disclose their material connection with
brands (FTC 2017). In practice, material connections may include either monetary pay-
ment or free product samples. As explained previously, advertising literacy can be
used to evaluate an influencers motives behind including a brand in a post (Campbell
and Kirmani 2000). However, we suggest that effects could also vary based on the
type of compensation. In particular, we compare consumersinfluencer and brand
evaluations when the influencer discloses being paid by a brand to include products
in their social media posts versus receiving free products. We compare both to posts
in which the influencer explicitly disclosed that a recommendation was not sponsored.
Both forms of disclosed compensation will lead to lower brand attitudes compared
to the non-sponsorship disclosure due to higher ad recognition, which activates ad
skepticism. However, we expect that the honesty of an influencer might be perceived
differently. When influencers are financially compensated, consumers will likely per-
ceive the post as boughtby the brand, whereas when the influencer was given a
free product, they tend to perceive it as a product trial. An influencer who is being
paid to recommend a brand on social media may be less likely to express an honest
opinion about the product and, therefore, be positively biased. As a result, consumers
might feel misled (Campbell and Kirmani 2000). When an influencer is sent a free
product, the brand has no guaranteed exposure nor control over the content of the
sponsored post. If an influencer chooses to endorse a brand, his or her followers tend
to believe that the influencer genuinely wants to share an opinion about the brand,
because the post would not exist, otherwise. This is known as the correspondence
bias (Gilbert and Malone 1995), which is the tendency to assume that a persons
behavior is a true reflection of their underlying dispositions (beliefs and opinions),
when, in fact, their behavior could be explained entirely by situational factors, e.g.
when the influencer is gifted a free product. Therefore, we expect that material com-
pensation will have a significantly less negative impact on both the influencer and the
brand than financial compensation.
Hypothesis 3: Both a disclosure indicating financial compensation and material
compensation will lead to lower brand attitudes through increased ad recognition, ad
skepticism, and lower source credibility compared to a non-sponsorship disclosure.
INTERNATIONAL JOURNAL OF ADVERTISING 7
However, the negative effects will be less significant in the case of material compensation
compared to financial compensation.
The moderating role of message sidedness
Research on how the content of a sponsored social media post affects its effectiveness
is mainly absent. However, we expect that the content may impact the effects of
sponsorship disclosure. In particular, we expect that a disclosure of sponsorship may
affect consumersattitudes differently in one-sided messages than they do in two-
sided messages. Message sidedness refers to the content of the propositions within a
persuasive appeal. When an influencer only highlights the positive aspects of a prod-
uct without mentioning any disadvantages, this is a one-sided message. When he or
she chooses to highlight both positive and negative aspects of the product, a two-
sided message is used (Winter and Kr
amer 2012). Previous research in the context of
sponsored blogs proposed that two-sided messages enhance the credibility of the
endorser (Huang and Lin 2009; Uribe, Buzeta, and Vel
asquez 2016, Janssen, van
Sprang, and Fransen 2017). The inclusion of negative information in the message may
make people believe the influencer is more genuine and honest as he or she discusses
not only the overt positive facets of the product (Crowley and Hoyer 1994; Eisend
2007). This phenomenon can be linked to the discounting principle of attribution the-
ory, which states that the role of a given cause in producing a given effect is dis-
counted if other plausible causes are also present(Kelley 1973, 113). When an
Instagram post has a sponsorship disclosure, consumers will attribute the sponsored
post to persuasive motives. However, consumersattribution to the given cause (i.e.
the influencer was compensated for including the brand in a post) will be weakened,
and, consequently, persuasion motives will be discounted if other plausible causes are
also present. By using a two-sided message, a second plausible cause is introduced,
which is the influencer has altruistic motives and genuinely wants to share opinions
with other consumers. In this sense, adding both positive and negative arguments
serves as validation that the influencer is an authentic consumer who actually tried
the product and is not merely interested in (financial) compensation in return for the
promotion (Schindler and Bickart 2005). These perceptions of honesty can mitigate
adverse effects on source credibility and message attitudes due to sponsorship disclo-
sures (Carr and Hayes 2014; Hwang and Jeong 2016). For instance, Hwang and Jeong
(2016) found that including a sponsorship disclosure in a blog post had negative
effects on source credibility perceptions and message attitudes. However, these nega-
tive effects were not present when the blogger emphasized that although the blog
was sponsored, he or she was sharing personal and honest opinions. In a similar vein,
Uribe, Buzeta, and Vel
asquez (2016) found that when a blog included an advertorial
label, a two-sided message was perceived as more credible than a one-sided one.
Hence, we expect that using two-sided messages in a sponsored Instagram post may
reduce the negative impact of sponsorship disclosure.
Hypothesis 4: When an influencer uses a one-sided message in a sponsored post, the
indirect negative effect of a sponsorship disclosure on brand attitude, through the
perceived credibility of the influencer, is more likely to occur compared to when an
influencer uses a two-sided message in a sponsored post.
8 M. DE VEIRMAN AND L. HUDDERS
We expect that message sidedness will work in similar ways for financial versus
material compensation sponsorship disclosures; in that, two-sided messages reduce
the negative impact of sponsorship disclosures through increased perceptions of the
credibility of influencers. When an influencer is financially motivated, consumers
assume that the sponsored post typically has more positive information about the
brand, whereas an influencer who received free products will likely be perceived as
less biased (Lu, Chang, and Chang 2014). Therefore, following attribution theory, we
expect that negative effects might be stronger for sponsored posts indicating financial
compensation. (Winter et al. 2015) One-sided messages, on the other hand, confirm
consumersexpectation of influencer bias and a distorted evaluation of the product
and may further strengthen the negative impact of a financial compensation sponsor-
ship disclosure.
Hypothesis 5: The impact of a financial compensation sponsorship disclosure (versus
non-sponsorship disclosure) on brand attitude through influencer credibility will be
stronger when the brand post displays one-sided messages than when it displays two-
sided messages. These effects will be weaker on a material compensation sponsor-
ship disclosure.
Method
Experimental design and procedure
We conducted an experiment based on a 4 2 between-subjects design (sponsorship
disclosure: material compensation versus financial compensation versus not sponsored
versus no disclosure message sidedness: one- versus two-sided). After reading a
short introduction text, participants were asked if they have an Instagram account, as
this was a criterion for inclusion in the sample. If they did not, they were directed to
the end of the survey. Hereafter, participantsage and gender were questioned, after
which the experiment started.
Participants were randomly exposed to one of the conditions. They were asked to
imagine they were scrolling through their Instagram feed and came across a post by
an influencer they follow. This post contained an endorsement for a fictitious brand
and contained different types of disclosures and either one-sided or two-sided argu-
ments. They were asked to have a good look at the Instagram post and afterward, par-
ticipants completed a questionnaire. First, the dependent variable, brand attitude, was
included, after which the mediating variables, namely source credibility, ad recogni-
tion, and ad skepticism were questioned. Hereafter, the manipulation checks were
included, followed by questioning participantsinvolvement with a healthy lifestyle
which is considered as a covariate in the analyses.
Stimulus material
An Instagram post of a fictitious influencer was created, using both a male and female
version (Stephan/Stephanie Jones). The gender of the influencer has been matched to
the gender of respondents to make sure that men were only exposed to the male
influencer and women were only exposed to the female influencer. The two versions
INTERNATIONAL JOURNAL OF ADVERTISING 9
have been merged in the analyses. The post was created using Adobe Photoshop soft-
ware. It was designed to resemble an authentic Instagram post appearing in the news
feed of an Instagram user, consisting of a picture of a product and a picture caption.
The post pictured energy bars from an unknown brand, Innerme(Figure 2).
Sponsorship disclosure was manipulated by adding four different types of sponsor-
ship disclosures in the picture caption indicating either material, financial, or no com-
pensation or not disclosed (see Table 1 for an overview of the manipulations). Next,
message sidedness was manipulated by adding a short review of the product depicted
in the Instagram post below the picture. This description consisted of positive product
features (one-sided message) or both positive and negative product features (two-
sided message). In particular, the one-sided message stated, Snacktime! These
#innerme energybars are my absolute favorite. Sure theyll give you the energy you
need!While the two-sided message stated, Snacktime! These #innerme energybars
are my absolute favorite. Although they contain quite some sugar, sure theyll give
you the energy you need!
A pretest among an Amazon Mechanical Turk sample (N¼167, 107 females,
mean
age
¼31.83 years, SD
age
¼7.09) was conducted to check manipulations of mes-
sage sidedness and the type of material connection between the influencer and the
brand and to measure disclosure recognition. To check manipulations of sidedness, a
multiple choice question, asking what kind of information the influencer included in
his or her Instagram post (Only arguments in favor of Innermeor Both arguments
Figure 2. Stimulus material.
10 M. DE VEIRMAN AND L. HUDDERS
favoring and disfavoring Innerme) was used. To check manipulations of the type of
material connection between the influencer and the brand, three items measured on
a seven-point Likert scale were used. Disclosure recognition was measured by asking
participants Which of the following text entries did you notice in Stephen Joness
Instagram post?Participants could choose one option from a list of possible disclo-
sures. A detailed overview of the used measurement scales can be found in
Methodological detailsAppendix. The detailed results are shown in Table 2. In par-
ticular, results show that 85.4% of the respondents could correctly recognize the dis-
closure type in the no disclosure condition while 14.6% could not (z¼6.4, p<.001). In
the not sponsored disclosure condition, 70% of the respondents correctly recognized
the disclosure while 30% did not (z¼3.6, p<.001). Although respondents seemed to
have more difficulties in recognizing the correct disclosure type in the material and
financial compensation conditions, further analyses show that respondents indicated
more often that the influencer received financial compensation in the financial com-
pensation condition (M ¼4.16) compared to the material compensation condition
(M ¼3.15, p<.001). Regarding perceived message sidedness, results reveal that 97.6%
of respondents correctly indicated that the post contained only arguments in favor of
the brand in the one-sided condition, while 2.4% of the respondents failed the
manipulation check (z¼12.3, p<.001). The manipulation of the two-sided condition
was less successful, as 66.3% of the respondents indicated that the post contained
both arguments against and arguments in favor of the brand in the two-sided condi-
tion, while 33.7% of the respondents did not notice (z¼4.2, p<.001).
Measures
To check manipulations of sidedness, participantsperceptions about the Instagram
post were questioned with three items measured on a seven-point Likert scale (The
Instagram post only gives opinions in favor of Innerme,The Instagram post gives
Table 1. Manipulation stimuli.
One-sided Two-sided
Not disclosed Snacktime! These #innerme energybars
are my absolute favorite. Sure theyll
give you the energy you need!
Snacktime! These #innerme energybars are my absolute
favorite. Although they contain quite some sugar,
sure theyll give you the energy you need!
Not-sponsored Snacktime! These #innerme energybars
are my absolute favorite. Sure theyll
give you the energy you need! This
isnt a sponsored post. I just felt like
trying them out and share my opnin-
ions. #notsponsored
Snacktime! These #innerme energybars are my absolute
favorite. Although they contain quite some sugar,
sure theyll give you the energy you need! This isnt
a sponsored post. I just felt like trying them out and
share my opninions. #notsponsored
Material
compensation
Snacktime! These #innerme energybars
are my absolute favorite. Sure theyll
give you the energy you need! Thanks
to @innerme I could try these out for
free and share my opin-
ions. #sponsored
Snacktime! These #innerme energybars are my absolute
favorite. Although they contain quite some sugar,
sure theyll give you the energy you need! Thanks to
@innerme I could try these out for free and share
my opinions. #sponsored
Financial
compensation
Snacktime! These #innerme energybars
are my absolute favorite. Sure theyll
give you the energy you need!
@innerme paid me to try these out
and share my opinions #sponsored
Snacktime! These #innerme energybars are my absolute
favorite. Although they contain quite some sugar,
sure theyll give you the energy you need! @innerme
paid me to try these out and share my opin-
ions #sponsored
INTERNATIONAL JOURNAL OF ADVERTISING 11
both opinions in favor as arguments against Innerme, and The information in the
Instagram post is balanced). To check manipulations of the type of material connec-
tion between the influencer and the brand and disclosure recognition, the same scales
as in the pretest were used.
Dependent and mediating variables were measured with validated scales, measured
on seven-point Likert scales. Brand attitude (ab) was measured with five items (Spears
and Singh 2004,M¼4.47, SD ¼1.42, a¼0.96). Credibility of the influencer was meas-
ured with Ohanians source credibility scale (1990, M ¼4.23, SD ¼1.13, a¼0.96) con-
sisting of 14 items. Advertising recognition was measured with three items reflecting
participantsability to recognize advertising (The Instagram post is adverting,
‘…contains advertising, and ‘…is commercial), adapted from van Reijmersdal et al.
(2016,M¼5.69, SD ¼1.30, a¼0.86). Ad skepticism was measured with four items [I
think the Instagram post tells the truth(reversed item), I dont believe what the influ-
encer wrote in the Instagram post,I can learn real product information from this
Table 2. Pretest results.
Message sidedness
Correct Wrong z-test
One-sided 97.6% 2.4% 12.3
Two-sided 66.3% 33.7% 4.2
Disclosure type classification
Correct Wrong z-test
Not disclosed 85.4% 14.6% 6.4
#notsponsored 70.0% 30.0% 3.6
Material compensation 43.9% 56.1% 1.1 (n.s.)
Financial compensation 53.3% 46.7% 0.6 (n.s.)
Material connection
Condition
Perceived compensation
Not disclosed Not-sponsored Material compensation Financial compensation
Perceived material compensation 3.59 (1.12) 2.90 (1.37) 4.34 (1.04) 4.22 (.90)
F(163)¼15.02, p<.001 j_____________________________________________j
j______________________________j

j______________________________j

j______________j

Perceived financial compensation 3.07 (1.10) 2.60 (1.32) 3.15 (1.17) 4.16 (.98)
F(163)¼14.12, p<.001 j_____________________________________________j

j______________________________j

j______________j

No compensation
F(163)¼16.38, p<.001
2.95 (1.05) 3.48 (1.43) 2.10 (1.14) 1.84 (1.19)
j_____________________________________________j

j______________________________j

j______________________________j

j______________j

12 M. DE VEIRMAN AND L. HUDDERS
Instagram post(reversed item), and After reading this Instagram post, I have been
accurately informed about the product information(reversed item)] from Lu, Chang,
and Chang (2014, M ¼4.15, SD ¼1.35, a¼0.85). Involvement with a healthy lifestyle
was included as a covariate in the analyses and was measured with the following two
items: A healthy lifestyle is important to meand Information about a healthy lifestyle
interests me(Van Reijmersdal, Neijens, and Smit 2010,M¼5.72, SD ¼1.02, r¼0.721,
p<.001). A detailed overview of the used measurement scales can be found in
Methodological detailsAppendix.
Participants
There were 414 US-based participants recruited from Amazons Mechanical Turk who
completed the study (229 females, M
age
¼32.57 years, SD
age
¼8.29) in return for a
small fee. Precisely, 59 responses were deleted from the data file, as these participants
failed to answer one or more control questions correctly. This exclusion leaves 355
participants (195 females, M
age
¼32.66 years, SD
age
¼8.29).
Statistical analyses
We first report the results of a correlational analysis to examine the correlations
between the mediating and dependent variables. Next, we use the PROCESS macro
(Hayes, 2018, v.3, 5000 bootstrap resamples, 95% confidence interval) to conduct
sequential mediation and moderated sequential analyses to test the effects of add-
ing a sponsorship disclosure to a branded Instagram post on brand attitudes and
the moderating effects of message sidedness. In particular, we compare the effects
of three different types of sponsorship disclosures, supplemented with a control
condition without sponsorship disclosure. We report the results in different steps,
comparing different sets of disclosures, in line with the hypotheses proposed in the
theoretical framework. First, we compare the non-disclosed post (coded as 0) with
the sponsorship disclosure posts (coded as 1) in which we merged the material and
financial compensation conditions (N¼266).Second,wecomparethenon-disclosed
post (coded as 0) with the non-sponsorship disclosure (coded as 1) (N¼178). Third,
we compare the non-sponsorship disclosure (coded as 0) with the material compen-
sation (coded as 1) and the financial compensation (coded as 2) disclosure using a
multi-categorical mediation and moderation analysis (N¼266). In all analyses,
involvement with a healthy lifestyle was included as a covariate as it was found to
be significantly related to brand attitude (r¼0.16, p<.001). The detailed results of
the analyses can be found in Methodological detailsAppendix.
Results
Table 3 provides an overview of the hypotheses and the results, along with informa-
tion on whether each hypothesis is supported or not. In addition, Figure 3 represents
the corrected model illustrating significant and insignificant paths.
INTERNATIONAL JOURNAL OF ADVERTISING 13
Manipulation checks
First, manipulation checks revealed a successful manipulation of the independent vari-
ables. A detailed overview of the manipulation check results can be found in Table 4.
First, results show that most participants correctly classified the disclosure type.
Results reveal that 80.9% of respondents in the no disclosure condition (z¼8.9,
p<.001), 76.4% of respondents in the non-sponsored condition (z¼7.0, p<.001),
63.5% of respondents in the material compensation condition (z¼3.7, p<.001), and
70.7% of respondents in the financial compensation condition (z¼5.6, p<.001) cor-
rectly recognized the sponsorship disclosure. In Table 4, an overview can be found of
results on participantsperceptions of material versus financial compensation the influ-
encer received from the brand.
Concerning message sidedness, results reveal that respondents perceived significantly
more arguments in favor of the brand in the one-sided condition (M
one-sided
¼6.13,
SD
one-sided
¼0.98, M
two-sided
¼4.42, SD
two-sided
¼1.83; t (353) ¼10.92, p<.001), while
respondents perceived significantly more arguments that were balanced (M
one-sided
¼
2.19, SD
one-sided
¼1.55, M
two-sided
¼4.13, SD
two-sided
¼1.85; t (353) ¼10.70, p<.001) and
both against and in favor of the brand (M
one-sided
¼3.01, SD
one-sided
¼1.63, M
two-sided
¼
4.09, SD
two-sided
¼1.65; t (353) ¼6.20, p<.001) in the two-sided condition.
Correlation analysis
First, we checked for correlations between the mediating variables ad recognition, ad
skepticism and source credibility and the dependent variable brand attitude.
Table 3. Hypotheses overview.
Hypothesis Supported? Main result
Hypothesis 1 Supported ab¼0.05, SE ¼0.03, 95% CI ¼[0.12; 0.01]
Hypothesis 2 Supported ab ¼0.27, SE ¼0.07, 95% CI ¼[0.14; 0.42]
Hypothesis 3 Partly supported ab
mat. comp.
¼0.26, SE ¼0.07, 95% CI¼[0.42; 0.14]
ab
fin. comp.
¼0.21, SE ¼0.06, 95% CI ¼[0.34; 0.11]
Hypothesis 4 Supported ab
one-sided
¼0.47, SE ¼0.18, 95% CI¼[0.82; 0.13]
ab
two-sided
¼0.02, SE ¼0.19, 95% CI ¼[0.35; 0.40]
Hypothesis 5 Not supported ab
mat. comp.
¼0.11, SE ¼0.30, 95% CI ¼[0.47; 0.68]
ab
fin. comp.
¼0.23, SE ¼0.28, 95% CI ¼[0.31; 0.79]
Figure 3. Corrected conceptual model.
14 M. DE VEIRMAN AND L. HUDDERS
All variables significantly correlated at the 0.01 level. Ad recognition was positively cor-
related with ad skepticism (r¼0.40) and negatively with source credibility (r¼0.24)
and brand attitude (r¼0.26). Ad skepticism was negatively correlated with source
Table 4. Manipulation check results.
Message sidedness
One-sided M (SD) Two-sided M (SD) t (df)
The Instagram post only gives opinions in favor of Innerme.
6.13 (0.98) 4.42 (1.83) 10.92 (353)
The Instagram post gives both opinions in favor as arguments against Innerme.
2.19 (1.55) 4.13 (1.85) 10.70 (353)
The information in the Instagram post is balanced.
3.01 (1.63) 4.09 (1.65) 6.20 (353)
Disclosure type classification
Correct Wrong z-test
Not disclosed 80.9% 14.6% 8.9
#notsponsored 76.4% 23.6% 7
Material compensation 63.5% 36.5% 3.7
Financial compensation 70.7% 29.3% 5.6
Material connection
Condition
Perceived compensation
Not disclosed Not-sponsored Material compensation Financial compensation
Perceived material compensation 4.64 (1.46) 3.75 (1.65) 5.73 (1.25) 5.42 (1.43)
F(351)¼32.41, p<.001 j_____________________________________________j

j______________________________j

j______________j

j______________________________j

j______________j

Perceived financial compensation 4.51 (1.64) 3.33 (1.64) 4.45 (1.73) 5.34 (1.59)
F(351)¼22.61, p<.001 j_____________________________________________j

j______________j

j______________________________j

j______________j

j______________j

No compensation
F(351)¼39.98, p<.001
3.53 (1.68) 4.70 (1.73) 2.61 (1.70) 2.17 (1.57)
j_____________________________________________j

j______________________________j

j______________j

j______________________________j

j______________j

INTERNATIONAL JOURNAL OF ADVERTISING 15
credibility (r¼0.72) and brand attitude (r¼0.68). To conclude, source credibility
and brand attitude were positively correlated (r¼0.73).
Sequential mediation analyses to test the mediating role of ad recognition, ad
skepticism, and source credibility
Comparing no disclosure with the #sponsored condition
First of all, a sequential mediation analysis (PROCESS model 6) with the disclosure con-
dition as an independent variable, advertising recognition, ad skepticism, and source
credibility as mediators, brand attitude as a dependent variable, and involvement with
a healthy lifestyle as a covariate showed a significant indirect effect from the presence
of a disclosure on brand attitude through advertising recognition, ad skepticism, and
source credibility (ab ¼0.05, SE ¼0.03, 95% CI ¼[0.12; 0.01]). A sponsorship dis-
closure increased ad recognition (a1 ¼0.31, SE ¼0.13, p¼.02), which, in turn, led to
higher ad skepticism (d21 ¼0.48, SE ¼0.07, p<.001) and negatively affected source
credibility (d32 ¼0.63, SE ¼0.04, p<.001), which was positively related with brand
attitude (b3 ¼0.58, SE ¼0.07, p<.001). The analyses additionally revealed a sequential
mediation through ad recognition and ad skepticism only (ab ¼0.05, SE ¼0.03, 95%
CI ¼[0.12; 0.00]). Ad skepticism had a negative effect on brand attitude
(b2 ¼0.36, SE ¼0.07, p<.001). No significant indirect effects appeared through ad
recognition only (ab ¼0.01, SE ¼0.02, 95% CI ¼[0.05; 0.03]), nor through ad skepti-
cism only (ab ¼0.02; SE ¼0.06, 95% CI ¼[0.16; 0.09]), nor through source credibility
only (ab ¼0.07, SE ¼0.06, 95% CI ¼[0.20; 0.04]). In conclusion, there was no signifi-
cant net direct effect of the presence of a disclosure on brand attitude (c0¼0.06,
SE ¼0.12, p¼.61). These results support Hypothesis 1 (Figure 4).
Comparing no disclosure with the #notsponsored condition
Next, we compared the effects of explicit disclosure that a post was not sponsored
with the control condition in a serial mediation analysis (PROCESS model 6) with the
disclosure condition as an independent variable, advertising recognition, ad skepti-
cism, and source credibility as mediators, brand attitude as a dependent variable, and
involvement with a healthy lifestyle as a covariate. Results showed a significant
Figure 4. Sequential mediation analysis: comparing the no disclosure with the #spon-
sored condition.
16 M. DE VEIRMAN AND L. HUDDERS
indirect effect from the presence of a disclosure on brand attitude through advertising
literacy and, sequentially, skepticism and source credibility (ab ¼0.27, SE ¼0.07, 95%
CI ¼[0.14; 0.42]), confirming Hypothesis 2. Ad recognition was lower when participants
were exposed to an Instagram post that explicitly mentioned that it was not spon-
sored (a1 ¼1.09, SE ¼0.20, p<.001), which, in turn, had a positive effect on skepti-
cism (d21 ¼0.50, SE ¼0.07, p<.001). Increased skepticism, consequently, appeared to
negatively affect source credibility (d32 ¼0.64, SE ¼0.04, p<.001), which was posi-
tively related to brand attitude (b3 ¼0.76, SE ¼0.09, p<.001). The analyses also
revealed a significant indirect effect from explicitly mentioning a post was not spon-
sored on brand attitude through ad recognition and ad skepticism (ab ¼0.12,
SE ¼0.05, 95% CI ¼[0.04; 0.23]). Ad skepticism negatively affected brand attitude
(b2 ¼0.23, SE ¼0.08, p<.001). Similar to the first analyses, no significant indirect
effects appeared through ad recognition only (ab ¼0.08, SE ¼0.06, 95% CI ¼[0.04;
0.18]). However, a significant indirect effect through ad skepticism only (ab ¼0.11,
SE ¼0.06, 95% CI ¼[0.25; 0.02]) was found. There was no significant indirect effect
from the presence of a disclosure on brand attitude through source credibility only
(ab ¼0.12, SE ¼0.09, 95% CI ¼[0.30; 0.05]). Further, there was no significant net
direct effect of the presence of a disclosure on brand attitude (c0¼0.03, SE ¼0.14,
p¼.82) (Figure 5).
Comparing the #notsponsored, #sponsored material compensation, and #spon-
sored financial compensation conditions
Next, we compare posts that explicitly mentioned compensation (either material or
financial) to the explicitly not sponsored post. First, a multi-categorical serial mediation
analysis (PROCESS model 6) was used, with the connection to the brand as an inde-
pendent variable, ad recognition, ad skepticism, and source credibility as mediators,
brand attitude as a dependent variable, and involvement with a healthy lifestyle as a
covariate. Results showed a significant negative indirect effect from the connection to
the brand on brand attitude through ad recognition, ad skepticism, and credibility,
sequentially, when comparing the material compensation condition to the not spon-
sored condition and when comparing the financial compensation condition to the not
sponsored condition (ab
mat. comp.
¼0.26, SE ¼0.07, 95% CI ¼[0.42; 0.14];
Figure 5. Sequential mediation analysis: comparing the no disclosure with the #notspon-
sored condition.
INTERNATIONAL JOURNAL OF ADVERTISING 17
ab
fin. comp.
¼0.21, SE ¼0.06, 95% CI ¼[0.34; 0.11]). Compared to an explicit dis-
closure of non-sponsorship, both material compensation (a1 ¼1.57, SE ¼0.18, p<.001)
and financial compensation (a1 ¼1.27, SE ¼0.17, p<.001) conditions led to higher ad
recognition, leading to more ad skepticism (d21 ¼0.46, SE ¼0.06, p<.001) and nega-
tively affecting source credibility (d32 ¼.60, SE ¼.04, p<.001), which was positively
related with brand attitude (b3 ¼0.61, SE ¼0.07, p<.001). Although there appeared to
be no significant difference between the impact of paid and gift conditions on ad rec-
ognition, a comparison of the standardized indirect effects showed that the negative
indirect effects on brand attitude were stronger when comparing the material com-
pensation disclosure with the non-sponsorship disclosure (ab ¼0.19, SE ¼0.05, 95%
CI ¼[0.30; 0.10]) than when comparing the financial compensation with the non-
sponsorship disclosure (ab ¼0.15, SE ¼0.04, 95% CI ¼[0.24; 0.08]). Similar to the
previous analyses reported, we also found a sequential mediation through ad recogni-
tion and ad skepticism only (ab
mat. comp.
¼0.24, SE ¼0.08, 95% CI ¼[0.40; 0.11];
ab
fin. comp.
¼0.20, SE ¼0.07, 95% CI ¼[0.34; 0.08]). Ad skepticism negatively
affected brand attitude (b2 ¼0.34, SE ¼0.06, p<.001). No significant indirect effects
appeared when comparing the effects of the three disclosure conditions on brand atti-
tude through ad recognition only (ab
mat. comp.
¼0.03, SE ¼0.09, 95% CI ¼[0.22;
0.15]; ab
fin. comp.
¼0.02, SE ¼0.07, 95% CI ¼[0.17; 0.13]), nor through ad skepticism
only (ab
mat. comp.
¼0.12, SE ¼0.08, 95% CI ¼[0.03; 0.29]; ab
fin. comp.
¼0.12, SE ¼0.07,
95% CI ¼[0.01; 0.28]), nor through source credibility only (ab
mat. comp.
¼0.06,
SE ¼0.08, 95% CI ¼[0.11; 0.23]; ab
fin. comp.
¼0.02, SE ¼0.07, 95% CI ¼[0.11; 0.17]).
To conclude, there was no significant net direct effect of the presence of a disclosure
on brand attitude (c0
mat. comp.
¼0.12, SE ¼0.16, p¼.44; c0
fin. comp.
¼0.01, SE ¼0.15,
p¼.92). Based on these findings, we can partly accept Hypothesis 3, as both types of
disclosure negatively affected brand attitudes through ad recognition, ad skepticism,
and source credibility. However, disclosing material compensation unexpectedly
appeared to be more detrimental in terms of brand attitude than disclosing financial
compensation (Figure 6).
Moderated mediation analyses
Comparing no disclosure with the #sponsored condition
First, we conducted a moderated mediation analysis (PROCESS model 7) with message
sidedness as a moderator of the effect of the presence of a sponsorship disclosure on
brand attitude through source credibility. Involvement with a healthy lifestyle was
included as a covariate. The analysis revealed the presence of moderated mediation
through source credibility (ab ¼0.49, SE ¼0.26, 95% CI ¼[0.00; 1.02]). Next, we further
examined the conditional indirect effects of the disclosure condition on brand attitude
for one- versus two-sided Instagram posts. Only for one-sided Instagram posts, there
was a significant indirect negative effect of the disclosure condition on brand attitude
through source credibility (ab ¼0.47, SE ¼0.18, 95% CI ¼[0.82; 0.13]. In contrast,
when the Instagram post included a two-sided message, this indirect effect was not
significant (ab ¼0.02, SE ¼0.19, 95% CI ¼[0.35; 0.40]). These results support
Hypothesis 4. However, interestingly to note, a more detailed analysis of the effects on
18 M. DE VEIRMAN AND L. HUDDERS
influencer credibility shows that influencer credibility is not significantly different
between disclosed one-sided and two-sided messages (B ¼0.19, SE ¼0.17, p¼.27),
while non-disclosed one-sided messages seem to lead to higher credibility ratings
compared to two-sided messages (B ¼0.75, SE ¼0.24, p¼.002) (Figure 7).
Comparing the no disclosure with the #notsponsored condition
Next, a moderated mediation analysis (PROCESS model 7) was conducted with mes-
sage sidedness as a moderator of the effect of a not sponsored disclosure (no disclos-
ure, coded as 0 versus #sponsored, coded as 1) on brand attitude, through source
credibility. Involvement with a healthy lifestyle was included as a covariate. The ana-
lysis revealed no significant moderated mediation (ab ¼0.35, SE ¼0.32, 95%
CI ¼[0.26; 0.98]).
Comparing the #notsponsored, #sponsored material compensation, and #spon-
sored financial compensation conditions
A multi-categorical moderated mediation analysis (PROCESS model 7) was conducted
with message sidedness as a moderator of the effect of connection with the brand
(#notsponsored, coded as 0 versus material compensation, coded as 1 versus financial
compensation, coded as 2) on brand attitude, through source credibility. Involvement
with a healthy lifestyle was included as a covariate. The analysis revealed no presence
of moderated mediation when the influencer was given a free product (ab ¼0.11,
SE ¼0.30, 95% CI ¼[0.47; 0.68]), nor when the influencer was paid for promoting the
brand (ab ¼0.23, SE ¼0.28, 95% CI ¼[0.31; 0.79]). Therefore, we need to reject
Hypothesis 5. However, further analyses revealed significantly lower ad recognition
when an influencer was paid for endorsing a brand compared to when he or she was
given free products when the sponsored post used a two-sided message
(M
mat. comp.
¼6.29, SD
mat. comp.
¼0.81; M
fin. comp.
¼5.87, SD
fin. comp.
¼1.11; p¼.05). No
differences between the two forms of compensation were found when the sponsored
post used a one-sided message (M
mat. comp.
¼6.24, SD
mat. comp.
¼0.81; M
fin. comp.
¼6.05,
SD
fin. comp.
¼1.04; p¼.35).
Figure 6. Sequential Mediation Analysis: Comparing the #notsponsored, #sponsored material com-
pensation and #sponsored financial compensation conditions.
INTERNATIONAL JOURNAL OF ADVERTISING 19
General discussion
Theoretical contributions and future research
This study contributes to the further understanding of the function of disclosures in
sponsored Instagram posts as a type of native advertising. First, it was found that dis-
closing sponsored content, regardless of the type of material connection between the
influencer and the brand, activated advertising recognition and resulted in more ad
skepticism, which, in turn, negatively affected brand attitude. These results provide fur-
ther support for previous findings suggesting that adding a sponsorship disclosure
helps consumers recognize sponsored posts as advertising, which reduces the effect-
iveness of the ad (Evans et al. 2017; Boerman, Willemsen, and Van der Aa 2017).
Advertising literacy (Friestad and Wright 1994) is thus an important underlying mech-
anism that explains how sponsorship disclosures may induce advertising recognition,
which results in coping mechanisms such as heightened skepticism that may nega-
tively affect the effectiveness of sponsored influencer posts (Nelson, Wood, and
Paek 2009).
Moreover it was found that in its turn, ad skepticism leads to more negative per-
ceptions of influencer credibility, in line with the recent study by Janssen, van Sprang,
and Fransen (2017). Consumers tend to believe that influencers are trustworthy sour-
ces who express their genuine liking for a brand, which makes them more likely to be
persuaded, according to source credibility theory (Erdogan 1999; Hovland and Weiss
1951; Ohanian 1991). When they are informed about the commercial nature of a post,
and consumers understand the true motives of the influencer, they become skeptical,
which causes negative responses toward the influencer and, consequently, the brand.
Second, results reveal that it may be beneficial for both influencers and brands to
disclose non-sponsored posts clearly. This disclosure not only leads to more positive
brand effects but also to more positive evaluations of the influencer, in terms of cred-
ibility, compared to a non-disclosed post. At first sight, these results are not in line
with the study by Carr and Hayes (2014), which compared a review including a non-
Figure 7. The impact of message sidedness of (non)disclosed influencer posts on influencer
credibility.
20 M. DE VEIRMAN AND L. HUDDERS
sponsorship disclosure with a non-disclosed review and found no differences in per-
ceived source credibility between both conditions. Our results also revealed no direct
impact on source credibility; however, an indirect effect through ad recognition and
ad skepticism was found. Therefore, it is important that future research further unrav-
els the importance of such a non-sponsorship disclosure, especially now that spon-
sored content is a common practice, whereas four years ago Carr and Hayes (2014)
assumed that consumers perceive a non-sponsorship disclosure as an unnecessary
statement that the blogger is credible.
Third, the current study aimed to investigate whether the type of material connec-
tion between the influencer and the brand might impact brand attitude differently.
Results show that sponsorship disclosure indicating financial or material compensation
led to higher ad recognition compared to a non-sponsorship disclosure. This higher ad
recognition led to higher ad skepticism, lower influencer credibility, and lower brand
attitude. However, unexpectedly, the difference between a material connection and a
non-sponsorship disclosure appeared to be greater than the difference between a
financial compensation sponsorship and a non-sponsorship disclosure. Below, we dis-
cuss two possible explanations for this finding.
A first explanation can be found in perceptions about the influencers transpar-
ency. When an influencer honestly discloses that he or she was paid to endorse a
brand on social media, consumers are made fully aware of the influencersbiasand
commercial motives. When the influencer states that he or she was given a free
product in return for exposure, there is more uncertainty about the influencerstrue
motives. That is, the brand may have an agreement with the influencer including
free products in return for a recommendation, or the brand may have sent free
products to the influencer, hoping that it would be included in social media posts.
This result may be in line with the results of Carr and Hayes (2014), which showed
that bloggers were perceived to be less credible when they were more vague about
the sponsorship relation compared to when they explicitly disclosed a financial rela-
tionship. Thus, it might be that the material compensation sponsorship disclosure
was perceived as the influencer being less straightforward than when he or she
admitted to receive financial compensation. Therefore, future research should probe
for perceptions about the influencers honesty and trustworthiness and question to
what extent consumers think the brand determined the content of the sponsored
post. Moreover, feelings of deception could be an important mediator, explaining
the effects of different types of disclosures on attitudes toward both the influencer
and the brand. In support of this, recently, Evans, Wojdynski and Hoy (2018) found
that the negative impact of covertness of advertising format on attitudes and
behavioral intention via advertising recognition could be mitigated by perceptions
ofsponsorshiptransparency.Thus,incaseofmorecovertadvertisingformats,such
as sponsored social media posts, being clear about the posts advertising nature,
may result in less feelings of deception and consequently reactance (Evans, Phua,
and Grubbs Hoy 2018;Brehm,1966). It seems that regardless of the type of com-
mercial relationship, being straightforward about the postscommercialnatureand
specifically, the form of compensation is important to minimize feelings of decep-
tion and consequently negative consumer responses.
INTERNATIONAL JOURNAL OF ADVERTISING 21
A second plausible explanation for the indication that a material compensation dis-
closure may be more detrimental than we expected at first is that free gifts could trig-
ger feelings of envy among consumers, more than paid brand promotions. An
influencer who is paid to promote products may be perceived as being more profes-
sional, which may make it more acceptable to consumers. This ambiguity should be
addressed in future research by measuring feelings of envy and appropriateness of
the compensation method. In general, future research should provide further insights
into how to mitigate the negative effects of disclosing a material connection between
the influencer and the brand without misleading consumers. Related, examining the
difference between disclosures from influencers (e.g. using hashtags, as we examined
in our study) versus disclosures added by media platforms (standardized disclosure)
could be an interesting future research topic. The source of the disclosure may heavily
affect the credibility of an influencer, as he or she is no longer disclosing the spon-
sored content; rather, an external source does it. This difference may lead to higher
feelings of manipulativeness and lead to lower credibility for the advertising message.
The use of two-sided messages may be even more important for standardized
disclosures.
Fourth, we examined the importance of message framing in a sponsored
Instagram post. In particular, we examined how message sidedness affects the
effects of sponsorship disclosures. More specifically, negative evaluations of influen-
cer credibility were found to be more likely when a disclosed (compared to a non-
disclosed post) Instagram post used a one-sided message to promote the product.
When a two-sided message was being used, no significant difference appeared
between disclosed and non-disclosed influencer posts, in line with Uribe, Buzeta,
and Vel
asquez (2016). However, when looking into the results in more detail, they
reveal that credibility perceptions are particularly high when the post was one-
sided and not disclosed. Influencer credibility ratings of the one-sided disclosed
post were at the same level compared to the disclosed and non-disclosed two-
sided messages. Contrary to our expectations, two-sided messages that were dis-
closed did not lead to higher credibility ratings compared to one-sided messages.
Accordingly, further research is necessary to examine how discussing both negative
and positive product features may serve as a validation cue for the influencers
honesty (Schindler and Bickart 2005). In particular, it remains unclear why an influ-
encer posting a two-sided non-disclosed post is perceived as less credible com-
pared to when he or she posts a one-sided post. This finding is inconsistent with
previous research that indicates that sources of two-sided messages are generally
perceived as more honest, which benefits the overall credibility of the message
(Crowley and Hoyer 1994;Eisend2007).
Practical implications
As influencer marketing has evolved to an exponentially growing marketing tactic, pub-
lic policy concerns on its persuasiveness and covert nature have been rising.
Transparency about the commercial nature of the post has become crucial. Recent
guidelines and (self-)regulatory initiatives force influencers to disclose any commercial
relationship (FTC 2015;IAB2016). They emphasize the importance of sponsorship
22 M. DE VEIRMAN AND L. HUDDERS
transparency so that consumers are able to differentiate sponsored content from editor-
ial content or genuine (and non-sponsored) product reviews. The current study shows
that adding a detailed disclosure in which information is given about the nature of the
commercial relationship (either a financial compensation or a free product) helps con-
sumers recognize sponsored influencer posts as advertising. Moreover, the current study
shows that disclosures increase the critical processing of influencer posts as ad recogni-
tion evoked by a sponsorship disclosure induces ad skepticism, which in turn negatively
affects brand responses. Furthermore, it was found that sponsorship disclosures may
also affect perceptions of influencer credibility and consequently brand attitudes due to
increased ad skepticism following ad recognition. These findings indicate that clear dis-
closures do not miss their purpose and indeed help consumers recognize sponsored
influencer posts as advertising, which may result in increased skepticism and conse-
quently negative outcomes for both the influencer and the brand. However, as also rec-
ognized by Reijmersdal et al. (2016) transparency might be appreciated and recognized
by the influencers followers in the long run and in the end might soften their resist-
ance. This is also supported by the finding that consumers appreciate a clear non-spon-
sorship disclosure. As such, the influencer can eliminate any suspicion that consumers
may have towards a non-disclosed post. Consumers dislike the feeling of being misled,
and a clear non-sponsorship disclosure offers them more certainty about the truthful-
ness of the product recommendation, as there is no third-party influence.
Concerning the type of compensation method, both a disclosure indicating financial
as material compensation negatively affected brand attitudes through ad recognition,
ad skepticism, and source credibility. Still and unexpected, disclosing material compen-
sation appeared to be more detrimental in terms of brand attitude than disclosing
financial compensation. This might be important for marketers to take into account
when deciding on the type of remuneration for influencers.
As negative brand outcomes through perceived credibility of the influencer are more
likely to occur when an influencer uses a one-sided message in a sponsored post com-
pared to a two-sided message, influencers using two sided-messages in their sponsored
posts seem to be desirable for brands. However, in terms of influencer credibility, the
results of this study suggest that including two-sided arguments to a disclosed
Instagram post is equally effective compared to one-sided arguments. It seems that,
when including a sponsorship disclosure, it does not matter which argument style the
influencer uses; however, future research is needed to get more certainty about this.
To conclude, when comparing the two types of compensation methods, it was found
that when the influencer used a two-sided message when disclosing financial compen-
sation, this led to significantly lower ad recognition compared to when he or she was
given free products while no differences between the two forms of compensation were
found when the influencer used a one-sided message. These results suggest that two-
sided messages could decrease ad recognition for paid brand endorsements.
Limitations
To conclude, we discuss some limitations of the current study that may provide fur-
ther guidance for future research. First, we selected one particular type of product,
INTERNATIONAL JOURNAL OF ADVERTISING 23
energy bars, which may limit the generalizability of the findings. Results may be differ-
ent for products which are more expensive (e.g. luxury clothes) or more experiential
(e.g. travel). These products may lead to higher feelings of envy and may result in
more negative effects when disclosing financial or material compensation. It would be
interesting to examine further how the product type impacts the effectiveness of
advertising disclosures. Moreover, the used brand stimuli were unknown to respond-
ents to avoid influences of consumersestablished brand knowledge. However, con-
sumersfamiliarity or prior experiences with a product may affect brand evaluations
(Hong and Sternthal 2010). Therefore, future research should investigate whether our
findings are generalizable to known brands and whether results differ depending on
consumersprior brand knowledge. Furthermore, it might be that the energy bars
used in the stimuli were not necessarily perceived as healthy by the respondents.
Thus, rather than measuring and controlling for involvement with a healthy lifestyle,
product involvement might have been a better control measure.
Second, we only examined one particular message strategy, two-sided messages,
and we only manipulated the figure caption and not the picture. Adding an element
of two-sidedness may more strongly affect consumers compared to manipulating the
caption. Pictures may have a more significant impact than message elements. Further
research should also examine the impact of other message strategies, as research on
how a sponsored influencer post should be designed is quite limited. When consum-
ers encounter a sponsored post, they will most likely rely on cues or heuristics to aid
them in their evaluations and decision process (Kardes, Posavac, and Cronley 2004). In
this sense, the language used in the sponsored post may serve as a cue for an assess-
ment of credibility. For instance, it would be interesting to examine the impact of lan-
guage elements such as typos and language errors, using foreign languages, slang, or
dialects. Moreover, language is particularly important, as consumers should decode
and comprehend the post as intended by the influencer (Christiansen and Tax 2000).
A third limitation concerns the manipulation of message sidedness. While the
manipulation of the one-sided message was clearly successful, the manipulation of the
two-sided message was less clear to respondents. Even though the difference between
the respondents who correctly understood the manipulation (66.3%) and respondents
who did not (33.7%; z¼4.2, p<.001), proved to be statistically different, it does not con-
vincingly support the conclusion of a successful manipulation. However, as message
sidedness is an intrinsic feature of the message, it does not require a manipulation
check as argued by OKeefe (2003). Regardless of the respondents evaluation, it is an
objectively verifiable manipulation. Still, future research might search for better manipu-
lations of two-sidedness. For instance, we assume that the fact that the energy bars
contain a lot of sugar might not be perceived as a negative characteristic by some par-
ticipants. Moreover, the manipulation of sidedness is important; for instance, Hwang
and Jeong (2016) blame the lack of a moderating effect of message sidedness on the
effect of sponsorship disclosure on an unsuccessful manipulation of two-sidedness.
Fourth, the study did not control for influencer credibility and the influencer used
in the study was fictitious and thus unknown to participants. It could be that an influ-
encers prior credibility might have an impact on sponsorship disclosure effects such
that effects might differ for people who are followers of the influencer and people
24 M. DE VEIRMAN AND L. HUDDERS
who have no prior experience with the influencer. For instance, when an influencer is
known to post authentic, reliable recommendations and is always clear about his or
her commercial partnerships, a clear disclosure might be less detrimental in terms of
influencer credibility and brand attitude.
In addition, although Instagram is one of the most popular social network sites
worldwide (Statista 2018b) and currently the most used platform for influencer mar-
keting (WFA 2018), only including sponsored Instagram posts in the stimuli limits the
applicability of the study results to other platforms and consumers, as Instagram is
most popular with younger users below the age of 35 and more used by females than
males (Statista 2018b,2018c). Moreover, we focused on adult Instagram users; how-
ever, the effects may be stronger for youths (between 12 and 18), who are heavy
Instagram users and have a lower advertising literacy. Future research should further
unravel the effects among a younger audience.
To conclude, a more in-depth analysis of the design of sponsorship disclosures is
necessary to be able to formulate guidelines on the design of such disclosures. In sum,
the current study provides further evidence for the importance of sponsorship disclosures
to disclose sponsored content that is covert and potentially misleading for consumers.
Acknowledgements
The authors would like to thank Lisa Van Den Abbeele for her help with the creation of the
manipulation stimuli.
Disclosure statement
No potential conflict of interest was reported by the authors.
Funding
This work was financially supported by Fonds Wetenschappelijk Onderzoek [Grant
FWO3E02015003501].
References
Abidin, C. 2016. Visibility labour: Engaging with influencersfashion brands and #OOTD advertor-
ial campaigns on Instagram. Media International Australia 161, no. 1:86100.
Bickart, B., and R.M. Schindler. 2001. Internet forums as influential sources of consumer informa-
tion. Journal of Interactive Marketing 15, no. 3:3140.
Brehm, J.W. 1966. A theory of psychological reactance. New York: Academic Press.
Brehm, S.S., and J.W. Brehm. 1981. Psychological reactance: A theory of freedom and control. San
Diego, CA: Academic Press.
Boerman, S.C., E.A. van Reijmersdal, and P.C. Neijens. 2012. Sponsorship disclosure: Effects of
duration on persuasion knowledge and brand responses. Journal of Communication, 62:
10471064.
Boerman, S.C., E.A. van Reijmersdal, and P.C. Neijens. 2015. Using eye tracking to understand the
effects of brand placement disclosure types in television programs. Journal of Advertising 44,
no. 3:196207.
INTERNATIONAL JOURNAL OF ADVERTISING 25
Boerman, S.C., L.M. Willemsen, and E. Van der Aa. 2017. This post is sponsored: Effects of spon-
sorship disclosure on persuasion knowledge and electronic word of mouth in the context of
Facebook. Journal of Interactive Marketing, no. 38:8292.
Campbell, M.C., and A. Kirmani. 2000. Consumersuse of persuasion knowledge: The effects of
accessibility and cognitive capacity on perceptions of an influence agent. Journal of Consumer
Research 27, no. 1:6983.
Campbell, M.C., and A. Kirmani. 2008. I know what youre doing and why youre doing it: The
use of the persuasion knowledge model in consumer research. In The handbook of consumer
psychology, ed. C. P. Haugtvedt, P. Herr, and F. Kardes,549571 (NJ: Lawrence Erlbaum).
Campbell, M.C., G. Mohr, and P. Verlegh. 2013. Can disclosures lead consumers to resist covert
persuasion? The important roles of disclosure timing and type of response. Journal of
Consumer Psychology 23, no. 4:483495.
Carr, C.T., and R.A. Hayes. 2014. The effect of disclosure of third-party influence on an opinion
leaders credibility and electronic word of mouth in two-step flow. Journal of Interactive
Advertising 14, no. 1:3850.
Christiansen, T., and S.S. Tax. 2000. Measuring word of mouth: The questions of who and when?
Journal of Marketing Communications 6, no. 3:185199.
Crowley, A.E., and W.D. Hoyer. 1994. An integrative framework for understanding two-sided per-
suasion. Journal of Consumer Research, no. 20:561574.
De Veirman, M., V. Cauberghe, and L. Hudders. 2017. Marketing through Instagram influencers:
The impact of number of followers and product divergence on brand attitude. International
Journal of Advertising 36, no. 5:798828.
Eisend, M. 2006. Two-sided advertising: A meta-analysis. International Journal of Research in
Marketing 23, no. 2:187198.
Eisend, M. 2007. Understanding two sided persuasion: An empirical assessment of theoretical
approaches. Psychology & Marketing 24, no. 7:615640.
Erdogan, Z. 1999. Celebrity endorsement: A literature review. Journal of Marketing Management
15, no. 4:291314.
Evans, N.J., B.W. Wojdynski, and M. Grubbs Hoy (2018) How sponsorship transparency mitigates
negative effects of advertising recognition, International Journal of Advertising, DOI: 10.1080/
02650487.2018.1474998.
Evans, N.J., J. Phua, J. Lim, and L. Jun. 2017. Disclosing Instagram influencer advertising: The
effects of disclosure language on advertising recognition, attitudes, and behavioral intent.
Journal of Interactive Advertising 17, no. 2,138149.
Fransen, M.L., P.W.J. Verlegh, A. Kirmani, and E.G. Smit. 2015. A typology of consumer strategies
for resisting advertising, and a review of mechanisms for countering them. International
Journal of Advertising 34, no. 1:616.
Friestad M., and P. Wright. 1994. The persuasion knowledge model: How people cope with per-
suasion attempts. Journal of Consumer Research 22, no. 1:131.
Federal Trade Commission (FTC). 2015. Native advertising: A guide for businesses. https://www.
ftc.gov/tips-advice/business-center/guidance/native-advertising-guide-businesses. (accessed
August 22, 2018).
Gilbert, D.T., and P.S. Malone. 1995. The correspondence bias. Psychological Bulletin, no. 117:
2138.
Goldsmith, R.E., and R.A. Clark. 2008. An analysis of factors affecting fashion opinion leadership
and fashion opinion seeking. Journal of Fashion Marketing and Management 12, no. 3:
308322.
Hayes, A.F. (2018). Introduction to mediation, moderation, and conditional process analysis. (2nd
Ed.). New York: The Guilford Press.
Hennig-Thurau, T., K.P. Gwinner, G. Walsh, and D.D. Gremler. 2004. Electronic word-of-mouth via
consumer-opinion platforms: What motivates consumers to articulate themselves on the inter-
net? Journal of Interactive Marketing 18, no. 1:3852.
Hong, J., and B. Sternthal. 2010. The effects of consumer prior knowledge and processing strat-
egies on judgments. Journal of Marketing Research 47, no. 2:301311.
26 M. DE VEIRMAN AND L. HUDDERS
Hovland, C., and W. Weiss. 1951. The influence of source credibility on communication effective-
ness. Public Opinion Quarterly 15:635650.
Huang, L., and C. Lin. 2009. The effects of message sidedness and commercial sponsorship on
consumersevaluations towards web forums. Proceedings of Business and Information 6, no. 1.
Hudders, L., P. De Pauw, V. Cauberghe, K. Panic, B. Zarouali, and E. Roosendaal. 2017. Shedding
new light on how advertising literacy can affect childrens processing of embedded advertis-
ing formats: A future research agenda. Journal of Advertising 46, no. 2:333349.
Hwang, Y., and S. Jeong. 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 62:528535.
Interactive Advertising Bureau (IAB). 2016. IAB Europe Guidance: How to comply with EU rules
applicable to online native advertising. https://www.iabeurope.eu/wp-content/uploads/2016/
12/IAB-Europe-Online-Native-Advertising-Guidance.pdf (accessed August 22, 2018).
Janssen, L., B. van Sprang, and M. Fransen. 2017. Het effect van sponsorvermeldingen op merke-
valuaties en geloofwaardigheid van bloggers: De rol van tweezijdige boodschappen.
Tijdschrift voor Communicatiewetenschap 45, no. 1:1936.
Kaikati, A.M., and J.G. Kaikati. 2004. Stealth marketing: How to reach consumers surreptitiously.
California Management Review 46, no. 4:622.
Kardes, F.R., S.S. Posavac, and M.L. Cronley. 2004. Consumer inference: A review of processes,
bases, and judgment contexts. Journal of Consumer Psychology 14, no. 3:230256.
Kelley, H. 1973. The processes of causal attribution. American Psychologist 28:107128.
Kelley, L., G. Kerr, and J. Drennan. 2010. Avoidance of advertising in social network sites: The
teenage perspective. Journal of Interactive Advertising 10, no. 2:1627.
Lane, V., and R. Jacobson. 1995. Stock market reactions to brand extension announcements: The
effects of brand attitude and familiarity. Journal of Marketing 59, no. 1:6377.
Lu, L., W. Chang, and H. Chang. (2014). Consumer attitudes toward bloggerssponsored recom-
mendations and purchase intention: The effect of sponsorship type, product type, and brand
awareness. Computers in Human Behavior, 34,258266.
Main, K.J., D.W. Dahl, and P.R. Darke. 2007. Deliberative and automatic bases of suspicion:
Empirical evidence of the sinister attribution error. Journal of Consumer Psychology 17, no. 1:
5969
Morgan, R.M., and S.D. Hunt. 1994. The commitment-trust theory of relationship marketing.
Journal of Marketing 58, July:2038.
Nelson, M.R., M.L. Wood, and H. Paek. 2009. Increased persuasion knowledge of video news
releases: Audience beliefs about news and support for source disclosure. Journal of Mass
Media Ethics 24:220237.
Obermiller, C., and E.R. Spangenberg. 1998. Development of a scale to measure consumer skep-
ticism toward advertising. Journal of Consumer Psychology 7, no. 2:159186.
Ohanian, R. 1990. Construction and validation of a scale to measure celebrity endorsersper-
ceived expertise, trustworthiness, and attractiveness. Journal of Advertising 19, no. 3:3952.
Ohanian, R. 1991. The impact of celebrity spokespersonsperceived image on consumersinten-
tion to purchase. Journal of Advertising Research 31, no. 1:4654.
OKeefe, D.J. 2003. Message properties, mediating states, and manipulation checks: Claims, evi-
dence, and data analysis in experimental persuasive message effects research. Communication
Theory 13, no. 3:251274.
Petty, R.D., and J.C. Andrews. 2008. Covert marketing unmasked: A legal and regulatory guide
for practices that mas advertising messages. Journal of Public Policy & Marketing 27, no. 1:
718.
Phua, J.J., S.A. Jin, and J.H. Kim. 2017. Gratifications of using Facebook, Twitter, Instagram, or
Snapchat to follow brands. Telematics & Informatics 34, no. 1:412424.
Rozendaal, E., M. Buijzen, and P. Valkenburg. 2011. Childrens understanding of advertisersper-
suasive tactics. International Journal of Advertising 30, no. 2:329350.
van Reijmersdal, E.A., M.L. Fransen, G. Van Noort, S.J. Opree, L. Vandenberg, S. Reusch, S.C.
Boerman (2016). Effects of disclosing sponsored content in blogs: how the use of resistance
INTERNATIONAL JOURNAL OF ADVERTISING 27
strategies mediates effects on persuasion. American Behavioral Scientist, 60, no. 12:
14581474. DOI: 10.177/000276421666041
Schindler, R.M., and B. Bickart. 2005. Published word of mouth: Referable, consumer-generated
information on the Internet. In Online consumer psychology: Understanding and influencing
behavior in the virtual world, ed. C.P. Haugtvedt, K.A. Machleit, and R.F. Yalch,3561. Mahwah,
NJ: Lawrence Erlbaum Associates.
Shoham, M., S. Moldovan, and Y. Steinhart. 2017. Positively useless: Irrelevant negative informa-
tion enhances positive impressions. Journal of consumer psychology 27, no. 2:147159.
Spears, N., and S.N. Singh. 2004. Measuring attitude toward the brand and purchase intentions.
Journal of Current Issues and Research in Advertising 26, no. 2:5366.
Statista. 2018a. Number of brand sponsored influencer posts on Instagram from 2016 to 2019
(in millions). https://www.statista.com/statistics/693775/instagram-sponsored-influencer-con-
tent/ (accessed August 22, 2018).
Statista. 2018b. Most famous social network sites worldwide as of July 2018, ranked by number
of active users (in millions). https://www.statista.com/statistics/272014/global-social-networks-
ranked-by-number-of-users/ (accessed August 22, 2018).
Statista. 2018c. Distribution of Instagram users worldwide as of July 2018, by age and gender.
https://www.statista.com/statistics/248769/age-distribution-of-worldwide-instagram-users/
(accessed August 22, 2018).
Uribe, R., C. Buzeta, and M. Vel
asquez. 2016. Sodedness, commercial intent and expertise in blog
advertising. Journal of Business Research 69, no. 10,44034410.
van Reijmersdal, E.A., M.L. Fransen, G. Van Noort, S.J. Opree, L. Vandeberg, S. Reusch, F. van
Lieshout, and S.C. Boerman. 2016. Effects of disclosing sponsored content in blogs. How the
use of resistance strategies mediates effects on persuasion. American Behavioral Scientist 60,
no. 12:14581474.
van Reijmersdal, E.A., N. Lammers, E. Rozendaal, and M. Buijzen. 2015. Disclosing the persuasive
nature of advergames: Moderation effects of mood on brand responses via persuasion know-
ledge. International Journal of Advertising 34, no. 1:7084.
van Reijmersdal, E.A., P.C. Neijens, and E.G. Smit. 2010. Customer magazines: Effects of commer-
ciality on readersreactions. Journal of Current Issues and Research in Advertising 32, no.
1,5967.
Vonk, R. 1998. The slime effect: Suspicion and dislike of likeable behavior toward superiors.
Journal of Personality and Social Psychology 74:849864.
World Federation of Advertisers. 2018. Survey on influencer marketing. https://www.wfanet.org/
tools/global-knowledge-base/#!/item/558 (accessed August 22, 2018)
Winter, S., and N. Kr
amer. 2012. Selecting science information in web 2.0: How source cues, mes-
sage sidedness, and need for cognition influence usersexposure to blog posts. Journal of
Computer-Mediated Communication 18, no. 1:8096.
Winter, S., N.C. Kr
amer, L. R
osner, and G. Neubaum. 2015. Dont keep it (too) simple: How textual
representations of scientific uncertainty affect laypersonsattitudes. Journal of Language and
Social Psychology 34, no. 3:251272.
Wojdynski, B.W., and N.J. Evans. 2016. Going native: Effects of disclosure position and language
on the recognition and evaluation of online native advertising. Journal of Advertising 45, no.
2,5768.
Wu, P. C., and Y.C. Wang. 2011. The influences of electronic word-of-mouth message appeal and
message source credibility on brand attitude. Asia Pacific Journal of Marketing and Logistics
23, no. 4:448472.
28 M. DE VEIRMAN AND L. HUDDERS
Appendix. Methodological details
Pertinent details about the procedure and a priori screening rules for excluding
subjects from data collection and rationale
For both the pretest and main study, the data were collected in March 2018. The links to the
online experiments were provided through Amazons Mechanical Turk accompanied by the fol-
lowing description: Give us your opinion about an Instagram postand the following keywords:
survey, demographics. Qualification requirements were HIT Approval Rate (%) for all
RequestersHITs greater than or equal to 85and Location is US. The minimum age to partici-
pate was 18 years. The reward per assignment was $0.85. The average session length was
515.22 seconds.
The experiments were created in Qualtrics, which made it possible to assign each participant
to one of the conditions. 419 responses were collected. 5 participants who reported having no
Instagram account after reading the introduction, were directed to the end of the experiment,
which brought back the sample to 414 participants. 59 responses were deleted from the data
file as these participants failed to correctly answer one or both control questions. This brought
the sample back to 355 participants.
Overview and full phrasing of all questions used in the reported research
Pretest
1. Introduction
Thank you for taking time to participate in our survey!
The survey should take about 15 minutes, and your responses are completely anonymous. You can only take the
survey once. At the end of the survey you will receive a code which you paste into the box on the Mechanical
Turk platform to receive credit for taking part in our survey.
If you have any questions about the survey, please email us: <e-mail>
We really appreciate your input!
2. Do you have an Instagram account? 0 Yes
0No
Directed to end of survey
3. What is your age?
4. Are you ?0 Male
0 Female
5. Please, read the text below carefully
Imagine you are scrolling through your Instagram feed and you come to the following post of Stephen /
Stephanie Jones, an influencer you follow.
Please, carefully take a look at the Instagram post below.
6. Manipulation Stimuli Cfr. Figure 2
11. Manipulation Check Sidedness
What kind of information did Stephen / Stephanie Jones include in his / her Instagram post? (tick the correct
box
Only opinions in favor of Innerme.
Both opinions in favor as against Innerme.
12. Manipulation Check Disclosure Recognition
Which of the following text entries did you notice in Stephen / Stephanie JonesInstagram post? (tick the correct
box)
O #sponsored
O #spon
O #advertisement
O #ad
O #notsponsored
O #nospon
O none of the above
13. Manipulation Check Material Connection
Read the statements below carefully and indicate to what extent you agree (1 ¼totally disagree, 7 ¼totally
(continued)
INTERNATIONAL JOURNAL OF ADVERTISING 29
Main study
agree).
Stephen/Stephanie Jones received free samples of Innerme in return for promotion.
Stephen/Stephanie Jones received financial compensation in return for promoting Innerme.
Stephen/Stephanie Jones did not receive any compensation for posting about Innerme on Instagram.
15. Socio-Demographical Details
To finish, we would like to ask you some questions
about yourself.
16. What is the highest degree or level of school you have
completed? If currently enrolled, highest degree
received?
0 None
0 High/Secondary school
0 Bachelor
0 Master
0 PhD
17. Are you currently ? 0 Student
0 Full time employed
0 Part time employed
0 Unemployed
0 Retired
18. What is your marital status? 0 Single / never been married
0 Married or domestic partnership
0 Widowed
0 Divorced
0 Separated
19. What is your race? 0 White
0 African American
0 Other (please specify)
20. Thank you for participating in this study. Your validation code for m-Turk is Please press the NEXTbutton in
order to receive your payment!
The gender of the participant was matched to the gender of the manipulation stimuli (i.e. the influencers
Instagram account). Thus, male participants were exposed to the Instagram account of Stephen Jones, female par-
ticipants to the one of Stephanie Jones. Accordingly questions containing info about the influencers gender were
also adapted.
1. Introduction
Thank you for taking time to participate in our survey!
The survey should take about 15 minutes, and your responses are completely anonymous. You can only take the
survey once. At the end of the survey you will receive a code which you paste into the box on the Mechanical
Turk platform to receive credit for taking part in our survey.
If you have any questions about the survey, please email us: <e-mail>
We really appreciate your input!
2. Do you have an Instagram account? 0 Yes
0No
Directed to end of survey
3. What is your age?
4. Are you ?0 Male
0 Female
5. Please, read the text below carefully
Imagine you are scrolling through your Instagram feed and you come to the following post of Stephen /
Stephanie Jones, an influencer you follow.
Please, carefully take a look at the Instagram post below.
6. Manipulation Stimuli Cfr. Figure 2
7. Brand Attitude (Spears & Singh2004)
Do you find Innerme
bad 00000 good
unfavorable 00000 favorable
dislikeable 00000 likeable
unappealing 00000 appealing
unpleasant 00000 pleasant
8. Source Credibility (Ohanian1990)
Do you find Stephen / Stephanie Jones
(continued)
30 M. DE VEIRMAN AND L. HUDDERS
unattractive 00000 attractive
not classy 00000 classy
ugly 00000 beautiful
plain 00000 elegant
insincere 00000 sincere
undependable 00000 dependable
dishonest 00000 honest
unreliable 00000 reliable
untrustworthy 00000 trustworthy
not an expert 00000 expert
inexperienced 00000 experienced
unknowledgeable 00000 knowledgeable
unqualified 00000 qualified
unskilled 00000 skilled
9. Ad Recognition (Van Reijmersdal et al.2016) Read the statements below carefully and indicate to what extent
you agree (1 ¼strongly disagree, 7 ¼strongly agree). The Instagram post
is advertising.
is commercial.
contains advertising.
10. Ad Skepticism (Lu, Chang & Chang2014) Read the statements below carefully and indicate to what extent you
agree (1 ¼strongly disagree, 7 ¼strongly agree).
I think Stephen/Stephanie JonesInstagram post tells the truth. (R)
I dont believe what Stephen/Stephanie Jones wrote in his / her Instagram post.
I can learn real product information from Stephen / Stephanie JonesInstagram post. (R)
Please select Strongly disagree.
After viewing Stephen / Stephanie JonesInstagram post, I have been correctly informed about the product
information. (R)
Control question
11. Manipulation Check Sidedness Read the statements below carefully and indicate to what extent you agree (1 ¼
strongly disagree, 7 ¼strongly agree).
Stephen / Stephanie JonesInstagram post only gives opinions in favor of Innerme.
Stephen / Stephanie JonesInstagram post gives both opinions in favor as against favor of Innerme.
The information in Stephen / Stephanie JonesInstagram post is balanced.
12. Manipulation Check Disclosure Recognition
Which of the following text entries did you notice in Stephen / Stephanie JonesInstagram post? (tick the correct
box)
O #sponsored
O #spon
O #advertisement
O #ad
O #notsponsored
O #nospon
O none of the above
13. Manipulation Check Material Connection Read the statements below carefully and indicate to what extent you
agree (1 ¼totally disagree, 7 ¼totally agree).
Stephen/Stephanie Jones received free samples of Innerme in return for promotion.
Stephen/Stephanie Jones received financial compensation in return for promoting Innerme.
Stephen/Stephanie Jones did not receive any compensation for posting about Innerme on Instagram.
Please select Somewhat disagree
Control question
14. Involvement with a Healthy Lifestyle (Van Reijmersdal, Neijens & Smit2010) Read the statements below carefully
and indicate to what extent you agree (1 ¼strongly disagree, 7 ¼strongly agree).
A healthy lifestyle is important to me.
Information about a healthy lifestyle interests me.
15. Socio-Demographical Details To finish, we would like to ask you some questions about yourself.
16. What is the highest degree or level of school you have completed? If cur-
rently enrolled, highest degree received?
0 None
0 High / Secondary school
0 Bachelor
0 Master
0 PhD
17. Are you currently ? 0 Student
0 Full time employed
0 Part time employed
0 Unemployed 0 Retired
(continued)
INTERNATIONAL JOURNAL OF ADVERTISING 31
Exploratory analyses
Correlation table
Main effect of disclosure type
We conducted a one-way analysis of variance (ANOVA) with the different disclosure types as
independent variable and ad recognition, ad skepticism, source credibility and brand attitude as
dependent variables. There were only significant differences between the four disclosure types
for ad recognition (F ¼28.34, p <.001). A scheffe multiple comparisons test showed that only
the explicit non-sponsorship condition significantly differed from all other disclosure types (p <
.001) in that it is significantly less recognized as advertising than the no disclosure condition,
the #sponsored material compensation and #sponsored financial compensation condition.
Results one-way ANOVA
Mean scores for ad recognition
Main effect of message sidedness
We conducted an independent samples t-test with message sidedness as independent variable
and ad recognition, ad skepticism, source credibility and brand attitude as dependent variables.
The results showed that compared to two-sided posts, one-sided posts resulted in significantly
higher source credibility (t (353) ¼3.12, p ¼.00) and brand attitude (t (353) ¼3.33, p <.001).
18. What is your marital status? 0 Single / never been married
0 Married or domestic partnership
0 Widowed
0 Divorced
0 Separated
19. What is your race? 0 White
0 African American
0 Other (please specify)
20. Thank you for participating in this study. Your validation code for m-Turk is Please press the NEXTbutton in
order to receive your payment!
The gender of the participant was matched to the gender of the manipulation stimuli (i.e. the influencers
Instagram account). Thus, male participants were exposed to the Instagram account of Stephen Jones, female par-
ticipants to the one of Stephanie Jones. Accordingly questions containing info about the influencers gender were
also adapted.
Ad Recognition Ad Skepticism Source Credibility Brand Attitude
Ad Recognition 1 r ¼.40, p <.001 r ¼.24, p <.001 r ¼.26, p <.001
Ad Skepticism 1 r ¼.72, p <.001 r ¼.68, p <.001
Source Credibility 1 r ¼.73, p <.001
Brand Attitude 1
df F Sig.
Ad Recognition 3 28,35 <.001
Ad Skepticism 3 1.86 .13
Source Credibility 3 .57 .63
Brand Attitude 3 .37 .25
MSD
No disclosure 5.82 1.19
Explicit non-sponsorship 4.74 1.50
#sponsored material compensation 6.26 .81
#sponsored financial compensation 5.96 1.07
32 M. DE VEIRMAN AND L. HUDDERS
No significant differences between one-sided and two-posts were found for ad recognition (t
(353) ¼1.02, p ¼.31) nor ad skepticism (t (353) ¼1.19, p ¼.24).
Mean scores for source credibility and brand attitude
Interaction effect of disclosure type and message sidedness
We checked the interaction effects between disclosure type and message sidedness with a two-
way ANOVA. Results only revealed a significant interaction effect on advertising recognition (F
(3, 346) ¼.38, p <.001) and ad skepticism (F (3, 346) ¼1.53, p ¼.01). Pairwise comparisons
showed that only the explicit non-sponsorship condition significantly differed from all other dis-
closure types both when a one-sided and a two-sided message was used in that it leads to sig-
nificant lower ad recognition (Mone-sided ¼4.74, SDone-sided ¼.17, p <.001; Mtwo-sided ¼
4.68, SDtwo-sided ¼.18, p <.001) compared to all other disclosure types, i.e. the no disclosure
(Mone-sided ¼5.96, SDone-sided ¼.17, p <.001; Mtwo-sided ¼5.65, SDtwo-sided ¼.18, p <
.001), the #sponsored material compensation (Mone-sided ¼6.25, SDone-sided ¼.17, p <.001;
Mtwo-sided ¼6.31, SDtwo-sided ¼.18, p <.001) and #sponsored financial compensation
(Mone-sided ¼6.04, SDone-sided ¼.18, p <.001; Mtwo-sided ¼5.92, SDtwo-sided ¼.17, p <
.001) conditions. Concerning ad skepticism, pairwise comparisons showed only significant differ-
ences between the disclosure types when a one-sided message was used. More specific, the no
disclosure condition (M ¼3.81, SD ¼.20) led to significant lower ad skepticism compared to
the #sponsored gift (M ¼4.37, SD ¼.19, p ¼.04) and #sponsored paid (M ¼4.34, SD ¼.20, p
¼.05) condition. Also, the explicit non-sponsorship condition (M ¼3.83, SD ¼.19) led to signifi-
cant lower ad skepticism than the #sponsored gift condition (M ¼4.37, SD ¼.19, p ¼.05).
MSD
Source Credibility One-sided 4.41 1.10
Two-sided 4.05 1.13
Brand Attitude One-sided 4.71 1.38
Two-sided 4.22 1.42
INTERNATIONAL JOURNAL OF ADVERTISING 33
Hypothesis testing
Statistical analyses process macro model 6
Comparing the No Disclosure with the #sponsored condition
Coefficient(SE) t 95% Confidence Interval
Outcome: Advertising Recognition
Constant 4.91(.38) 13.05 [4.17; 5.65]
Disclosure .31(.13) 2.32 [.05; .58]
Involvement with a healthy lifestyle .16(.06) 2.53 [.03; .28]
Outcome: Skepticism
Constant 3.29(.56) 5.87 [2.18; 4.39]
Disclosure .07(.16) .44 [.24; .38]
Advertising Recognition .48(.07) 6.79 [.34; .63]
Involvement with a healthy lifestyle 36(.07) 4.93 [.50; .22]
Outcome: Credibility
Constant 6.74(.40) 16.84 [5.95; 7.53]
Disclosure .13(.11) 1.19 [.33; .08]
Advertising Recognition .03(.05) .53 [.07; .13]
Skepticism .63(.04) 15.28 [.71; .55]
Involvement with a healthy lifestyle .19(.11) 1.70 [.09; .11]
Outcome: Brand Attitude
Constant 3.35(.70) 4.78 [1.97; 4.73]
Disclosure .07(.13) .51 [.19; .32]
Advertising Recognition .03(.06) .41 [.15; .10]
Skepticism .36(.07) 5.23 [.50; .23]
Credibility .58(.07) 7.68 [.43; .72]
Involvement with a healthy lifestyle .05(.06) .83 [.07; .17]
Total effect of X on Y
.17(.19) .91 [.54; .20]
Direct effect of X on Y
.07(.13) .51 [.19; .32]
(continued)
34 M. DE VEIRMAN AND L. HUDDERS
Comparing the No Disclosure with the #notsponsored condition
Comparing the #notsponsored, #sponsored material compensation and #sponsored financial com-
pensation conditions
Indirect effects of X on Y
M¼advertising recognition & skepticism .05(.03) [.12; .00]
M¼advertising recognition, skepticism & credibility .05(.03) [.12; .01]
Partially standardized indirect effects of X on Y
M¼advertising recognition & skepticism .04(.02) [.08; .00]
M¼advertising recognition, skepticism & credibility .04(.02) [.08; .00]
Non-significant indirect effects are not reported
Coefficient(SE) t 95% Confidence Interval
Outcome: Advertising Recognition
Constant 5.37(.64) 8.32 [4.09; 6.63]
Disclosure 1.09(.20) 5.37 [1.49; .69]
Involvement with a healthy lifestyle .08(.11) .72 [.13; .29]
Outcome: Skepticism
Constant 3.08(.67) 4.59 [1.75; 4.41]
Disclosure .50(.19) 2.57 [.12; .88]
Advertising Recognition .50(.07) 7.51 [.37; .63]
Involvement with a healthy lifestyle 34(.09) 3.55 [.53; .15]
Outcome: Credibility
Constant 6.79(.42) 16.00 [5.85; 7.50]
Disclosure .15(.11) 1.36 [.39; .07]
Advertising Recognition .06(.04) 1.43 [.02; .15]
Skepticism .64(.04) 14.35 [.73; .55]
Involvement with a healthy lifestyle .01(.06) .17 [.12; .10]
Outcome: Brand Attitude
Constant 2.40(.79) 3.03 [.84; 3.97]
Disclosure .03(.14) .23 [.31; .24]
Advertising Recognition .07(.05) 1.35 [.18; .03]
Skepticism .23(.08) 2.89 [.39; .07]
Credibility .76(.09) 8.31 [.58; .94]
Involvement with a healthy lifestyle .03(.07) .43 [.11; .17]
Total effect of X on Y
.09(.21) .43 [.51; .33]
Direct effect of X on Y
.03(.14) .23 [.82; .31]
Indirect effects of X on Y
M¼advertising recognition & skepticism .13(.05) [.04; .23]
M¼skepticism & credibility .24(.10) [.45; .06]
M¼advertising recognition, skepticism & credibility .27(.07) [.14; .42]
Partially standardized indirect effects of X on Y
M¼advertising recognition & skepticism .09(.03) [.03; .16]
M¼skepticism & credibility .17(.07) [.31; .04]
M¼advertising recognition, skepticism & credibility .19(.05) [.10; .29]
Non-significant indirect effects are not reported
Coefficient(SE) t 95% Confidence Interval
Outcome: Advertising Recognition
Constant 3.69(.42) 8.80 [2.86; 4.52]
X1 1.58(.18) 8.96 [1.23; 1.92]
X2 1.27(.17) 7.37 [.93; 1.61]
Involvement with a healthy lifestyle .18(.07) 2.61 [.04; .31]
Outcome: Skepticism
Constant 3.44(.48) 7.11 [2.49; 4.40]
X1 .36(.20) 1.77 [.76; .04]
X2 .35(.19) 1.84 [.73; .02]
(continued)
INTERNATIONAL JOURNAL OF ADVERTISING 35
Statistical Analyses Process Macro Model 7
Comparing the No Disclosure with the #sponsored condition
Coefficient(SE) t 95% Confidence Interval
Outcome: Credibility
Constant 3.75(.44) 8.56 [2.89; 4.61]
Disclosure .54(.21) 2.61 [.94; .13]
Sidedness .70(.24) 2.94 [1.17; .23]
int_1 .56(.29) 1.91 [.02; 1.13]
Involvement with a healthy lifestyle .18(.07) .07 [.04; .31]
Outcome: Brand Attitude
Constant .19(.42) .44 [.64; 1.01]
Disclosure .05(.13) .41 [.21; .32]
Credibility .88(.66) 15.72 [.77; .99]
Involvement with a healthy lifestyle .09(.06) 1.45 [.03; .21]
Direct effect of X on Y
.05(.13) .41 [.21; .32]
Conditional indirect effects of X on Y at values of the moderator
One-sided .47(.18) [.82; .13]
Two-sided .02(.19) [.35; .40]
Index of Moderated Mediation
Sidedness .49(.26) [.00; 1.02]
Advertising Recognition .46(.06) 7.36 [.34; .58]
Involvement with a healthy lifestyle 28(.07) 4.08 [.42; .15]
Outcome: Credibility
Constant 6.48(.36) 17.80 [5.76; 7.19]
X1 .10(.14) .68 [.18; .37]
X2 .04(.13) .28 [.22; .30]
Advertising Recognition .02(.05) .48 [.07; .11]
Skepticism .60(.04) 14.09 [.68; .51]
Involvement with a healthy lifestyle .00(.05) .10 [.09; .10]
Outcome: Brand Attitude
Constant 3.24(.62) 5.21 [2.01; 4.46]
X1 .12(.16) .77 [.19; .44]
X2 .01(.15) .09 [.28; .31]
Advertising Recognition .02(.05) .34 [.12; .09]
Skepticism .34(.06) 5.19 [.46; .21]
Credibility .61(.07) 8.53 [.47; .75]
Involvement with a healthy lifestyle .02(.06) .38 [.09; .13]
Relative total effects of X on Y
X1 .08(.21) .38 [.49; .33]
X2 .13(.20) .64 [.54; .27]
Relative direct effects of X on Y
X1 .12(.16) .77 [.19; .44]
X2 .01(.15) .10 [.28; .31]
Relative indirect effects of X on Y
M¼advertising recognition & skepticism X1
X2
.24(.08)
.20(.07)
[.40; .11]
[.34; .08]
M¼advertising recognition, skepticism & credibility X1
X2
.26(.07)
.21(06)
[.42; .14]
[.34; .11]
Partially standardized indirect effects of X on Y
M¼advertising recognition & skepticism X1
X2
.18(.05)
.14(.05)
[.29; .08]
[.25; .06]
M¼advertising recognition, skepticism & credibility X1
X2
.19(.05)
.15(.04)
[.30; .10]
[.24; .08]
Non-significant indirect effects are not reported
36 M. DE VEIRMAN AND L. HUDDERS
Comparing the No disclosure with the #notsponsored condition
Comparing the #notsponsored, #sponsored material compensation and #sponsored financial com-
pensation conditions
Coefficient(SE) t 95% Confidence Interval
Outcome: Credibility
Constant 3.82(.53) 7.22 [2.77; 4.86]
Disclosure .37(.22) 1.64 [.81; .07]
Sidedness .70(.22) 3.13 [1.15; .26]
int_1 .35(.32) 1.09 [.27; .33]
Involvement with a healthy lifestyle .16(.08) 1.92 [.00; .33]
Outcome: Brand Attitude
Constant .15(.46) .33 [1.07; .76]
Disclosure .11(.13) .82 [.15; .37]
Credibility 1.01(.06) 16.57 [.89; 1.13]
Involvement with a healthy lifestyle .05(.07) .65 [.09; .19]
Direct effect of X on Y
.11(.13) .82 [.15; .37]
Conditional indirect effects of X on Y at values of the moderator
One-sided .37(.22) [.81; .06]
Two-sided .02(.23) [.48; .44]
Index of Moderated Mediation
Sidedness .35(.32) [.26; .98]
Coefficient(SE) t 95% Confidence Interval
Outcome: Credibility
Constant 3.65(.42) 8.75 [2.83; 4.48]
X1 .15(.23) .67 [.61; .30]
X2 .20(.23) .85 [.65; .26]
Sidedness .35(.23) 1.51 [.81; .11]
int_1 .13(.33) .38 [.53; .79]
int_2 .26(.33) .79 [.39; .91]
Involvement with a healthy lifestyle .13(.06) 1.98 [.00; .26]
Outcome: Brand Attitude
Constant .49(.40) 1.22 [.30; 1.28]
X1 .00(.15) .02 [.30; .29]
X2 .07(.14) .46 [.35; .22]
Credibility .88(.05) 16.16 [.77; .99]
Involvement with a healthy lifestyle .05(.06) .88 [.06; .16]
Relative direct effects of X on Y
X1 .00(.15) .02 [.30; .29]
X2 .07(.14) .46 [.35; .22]
Relative conditional indirect effects of X1 on Y at values of the moderator
One-sided .14(.20) [.53; .27]
Two-sided .02(.22) [.45; .41]
Index of Moderated Mediation X1
Sidedness .11(.30) [.47; .70]
Relative conditional indirect effects of X2 on Y at values of the moderator
One-sided .17(.19) [.53; .20]
Two-sided .12(.20) [.33; .45]
Index of Moderated Mediation X2
Sidedness .23(.28) [.31; .79]
INTERNATIONAL JOURNAL OF ADVERTISING 37
... (Lou & Yuan, 2019, p. 59) Los anunciantes le entregan en cierta manera el control al influencer de las redes sociales en la creación de contenido, lo que permite obtener un mensaje que es considerado más auténtico que cuando lo presenta directamente la marca (Childers et al., 2018). A su vez, los consumidores utilizan cada vez más las redes sociales para obtener información sobre productos y guiar sus decisiones de compra, valorando y confiando en las opiniones de sus pares en internet (De Veirman & Hudders, 2019). Así, los influencers de las redes sociales transmiten su opinión sobre las marcas en sus publicaciones, teniendo el potencial de intervenir en el comportamiento de la audiencia . ...
... Sin embargo, los influencers de las redes sociales pueden permanecer imprecisos sobre su relación comercial con la marca patrocinadora (De Veirman & Hudders, 2019). Por lo tanto, el marketing de influencer puede clasificarse como una forma de publicidad nativa, por la naturaleza de la relación pagada entre la marca y el influencer de las redes sociales, que puede no ser del todo clara para la audiencia (Evans et al., 2017). ...
... No obstante, aunque se reconoce la relevancia de la transparencia en la relación entre el influencer de las redes sociales y sus seguidores, la literatura del marketing de influencer, se ha centrado en la visión de los seguidores. Específicamente, referente a la percepción de la audiencia sobre la transparencia del patrocinio en las publicaciones de los influencers de redes sociales y el reconocimiento de los anuncios en las respuestas hacia la marca, como las actitudes de la marca y las intenciones de compra por parte de los seguidores (por ejemplo, Boerman, 2020;De Jans et al., 2018;De Veirman & Hudders, 2019;Evans et al., 2017). ...
Article
Full-text available
Las redes sociales han contribuido a una nueva estrategia de marketing donde las marcas forjan alianzas con los influencers de las redes sociales para generar contenido publicitario. Esta estrategia conocida como marketing de influencer se encuentra en crecimiento y ganando un interés sustancial en la literatura. Sin embargo, la comprensión actual de la percepción de los influencers de las redes sociales frente a la transparencia del contenido que publicitan sigue siendo limitada, donde prácticamente todos los estudios se centran en la percepción del seguidor y en países desarrollados donde existe mayor regulación frente a la transparencia publicitaria. Para abordar esta problemática, la presente investigación utilizó la teoría de credibilidad de la fuente y la teoría institucional con el objetivo de explorar las percepciones de los influencers de las redes sociales frente a la transparencia del contenido que publicitan. Sobre la base de un estudio cualitativo que incorporó entrevistas semiestructuradas con ocho influencers de las redes sociales en Chile. Los resultados de este estudio muestran que los influencers de las redes sociales consideran relevante en su credibilidad de la transparencia publicitaria; la confiabilidad, la experiencia, la auto-presentación en línea y los aspectos normativos. Estos hallazgos contribuyen a la literatura de marketing de influencer y también tiene importantes implicancias prácticas para el amplio y creciente campo de la publicidad como son los profesionales de marketing, las marcas y los agentes reguladores encargados de proteger al consumidor.
... These opinions are perceived to be more genuine and reliable than traditional advertisements [3,4] because consumers trust social media influencers as they would trust their friends [5]. Furthermore, social media influencers' opinions effectively influence consumers' brand attitudes and guide purchasing decisions [6]. ...
... A clear advertising disclosure is useful for consumers to promptly identify sponsored posts in an advertising-intensive environment on social media. However, some marketers and influencers are still reluctant to disclose sponsorship because they do not want to reveal their commercial motivations [1], which may decrease their credibility [6]. A decrease in an influencer's credibility can subsequentially decrease consumers' brand attitudes [8,9]. ...
... According to the information richness theory, visual presentation is richer than text because it conveys more information to consumers [10] and has the potential to affect influencer credibility and brand attitudes. However, prior research has predominantly focused on the effects of advertising-disclosure language on influencer credibility and brand attitudes [6]. Little attention has been paid to the ways in which influencers present and promote products visually. ...
Conference Paper
Full-text available
The impacts of advertising-disclosure language on influencer marketing have increasingly attracted researchers’ attention. However, little research has focused on the role of pictorial types on social media influencers and brands. Thus, this study explores how different combinations of advertising-disclosure conditions and pictorial types can affect social media influencer credibility and consumers’ brand attitudes. By using an experimental 3x2 between-subject design, we collected data from 264 followers of beauty influencers on Weibo. Results revealed that a post that disclosed sponsorship significantly decreased influencer credibility and consumers’ brand attitudes compared to a post without any disclosure. Results also showed that a post with no-sponsorship disclosure significantly increased influencer credibility and consumers’ brand attitudes than a post without any disclosure or a post disclosing sponsorship. With regard to pictorial types, the influencer-with-product type led to higher influencer credibility and more positive brand attitudes in consumers than the product-only type in all three disclosure conditions. This study highlights the role of pictorial types in strengthening the power of advertising-disclosure language, which can further affect influencer credibility and consumers’ brand attitudes. This study contributes to the visual complexity literature for influencer marketing.
... Some recent studies have also examined the effect of sponsorship disclosure on consumers' attitude toward the brand in the context of other forms of eWOM, and found that disclosure negatively affected brand attitude (Boerman, Willemsen, and van der Aa 2017;Campbell, Mohr, and Verlegh 2013;de Veirman and Hudders 2020;van Reijmersdal et al. 2016). There are also studies examining disclosure effects on ad recall or recognition (Boerman, Willemsen, and van der Aa 2017;Campbell, Mohr, and Verlegh 2013), purchase intent (Liljander, Gummerus, and S€ oderlund 2015;Mayrhofer et al. 2020;van Reijmersdal et al. 2016), source credibility (de Veirman and Hudders 2020; Liljander et al. 2015), interest in the blogger (Liljander et al. 2015), and the likelihood of posting eWOM content (Boerman, Willemsen, and van der Aa 2017;Liljander et al. 2015). ...
Article
Addressing the problem of increasing sponsored eWOM and diverse and confusing disclosure practices, this study examined the effects of different types of sponsorship disclosure messages on (1) consumers’ trust in the sponsored product reviewer and (2) attitudes toward the reviewer and the sponsoring brand. An online experiment revealed several key effects of sponsorship disclosure and disclosure message types. Overall, sponsorship disclosure messages generated lower consumer trust in the reviewer and attitudinal responses. In terms of the effects of different types of disclosed commercial gain, reviews disclosing the receipt of a free product were perceived to be equally acceptable as reviews without sponsorship disclosure. Disclosure revealing that the reviewer received payment or a sales commission led to lower trust in a reviewer, and disclosure of a sales commission generated more negative attitude toward the sponsoring brand. The disclosure message type effects on attitude were mediated by trust.
Article
Full-text available
Objetivo: Esta investigação objetivou identificar os atributos positivos e negativos das publicações do Instagram e suas formas de ativação do consumo.Método: Um estudo de caso foi realizado com os usuários do Instagram em duas etapas: entrevista coletiva e entrevistas individuais. Ao todo, 33 sujeitos participaram das entrevistas semiestruturadas, que foram examinadas por meio da análise de conteúdo com apoio do software ATLAS.ti.Originalidade/Relevância: Uma investigação sobre o Instagram como uma ferramenta online para influenciar o comportamento do consumidor é necessária, pelo fato de serem ambientes de construção e manutenção do relacionamento social que influencia a sociedade contemporânea.Resultados: Os resultados revelam que as publicações com imagens de elevada qualidade, pessoas felizes, boa apresentação do produto/serviço, experiências reais e positivas e informações complementares constituem os atributos positivos. Por outro lado, as publicações com imagens não atrativas e sem qualidade, que não correspondem à realidade do produto/serviço ou do usuário, com propaganda disfarçada e sem informações formam os atributos negativos. Ainda, nota-se que as publicações ativaram o consumo de diferentes formas, entretanto, isso apenas ocorreu devido às representações mentais pré-existentes. Contribuições teóricas/metodológicas: Este artigo trouxe a discussão sobre a temática da ativação do consumo no Instagram, ainda pouco explorada na literatura do comportamento do consumidor. Ademais, retrata características das publicações que conduzem a efeitos positivos e negativos na visão de entrevistados que representaram cada Estado do Brasil.
Article
This review provides insight into the research on the strategic use of social media influencers. A search in the Scopus database yielded a total of 154 peer-reviewed academic publications focussing on influencer marketing, published mostly in the last three years (2018–2020). Most of these studies were empirical and used survey research, an experimental design, content analysis, or interview methodology. While a few studies examined how influencers can incite behavioural change or affect public opinion, the majority of studies focussed on the use of influencers as a commercial marketing tactic. Using Stern’s Revised Communication Model for Advertising as a theoretical framework, the studies were classified according to three research characteristics: source, message, and audience. The first research stream focussed on the perspectives of influencers and communication professionals. The second stream was concerned with the content strategies used by influencers in their sponsored posts. The third research stream focussed on the appeal of influencers and the efficacy of their sponsored recommendations. Several studies in the third stream were concerned with the transparency of this covert marketing tactic. The paper concludes with a future research agenda and implications for marketing practice and public policy.
Article
Full-text available
According to the law of the European Union (EU), broadcasters are obligated to inform consumers about the presence of product placements (PP) through disclosures. To ascertain whether disclosures are able to improve consumers’ understanding of persuasive intent, researchers have examined the impact of multiple disclosure presentations. This can be attributed to the fact that all EU Member States currently apply a variety of different disclosures. No study so far, however, has assessed how PP disclosures are implemented across the EU and whether PP disclosures investigated in advertising research relate to disclosures used in practice. In the present study, we thus conducted both a systematic assessment of practical disclosures of leading EU broadcasters and a comparison of the current practical disclosure presentations with the empirically tested versions. Results revealed that especially brand-unspecific and repetitive disclosures are under-investigated forms of disclosures although commonly used in practice. Implications for both science and policy, and future research are discussed.
Article
Influencers can disclose vlog advertising by implementing a platform-generated disclosure or by generating their own disclosure. This study aims to investigate whether and how these two types of disclosures differentially affect children's responses toward the sponsored content and the influencer. To do so, a two (platform-generated [PG] disclosure: no PG disclosure versus PG disclosure) by three (influencer-generated [IG] disclosure: no IG disclosure versus no commercial interference disclosure versus commercial interference disclosure) between-subjects experimental study was conducted among 190 children (aged ten to twelve). The results show that both types of disclosures increased children's recognition of vlog advertising; however, children's advertising literacy (i.e. their understanding of influencer marketing and skepticism toward the ad) was less activated when the influencer specifically indicated that there was no commercial interference of the brand. In addition, besides being more beneficial for influencers (by not negatively affecting children's evaluations of the influencer), an IG disclosure also showed to be more favorable for brands, as it increased brand effects among children, whereas a PG disclosure decreased these brand effects.
Article
Background: Nutrition science is currently facing issues regarding the public’s perception of their credibility, with Social Media Influencers increasingly becoming key sources of nutrition-related information with high engagement rates. Source credibility, and to an extent, authenticity, have been widely studied in marketing and communications, but have not yet been considered in the context of nutrition or health communication. Thus investigation into the factors that impact perceived source and message credibility and authenticity is of interest to inform health communication on social media. Objective: To explore the factors that impact message and source credibility (which includes trustworthiness and expertise) or authenticity judgements on social media platforms, to better inform nutrition science social media communication best practice. Methods: Six databases across a variety of disciplines were searched during March 2019. Inclusion criteria were experimental studies, studies focusing on microblogs, healthy adult populations, and studies focusing on either source credibility or authenticity. Exclusion criteria were studies of those under 18 years old, clinical populations, grey literature, blogs, WeChat, online reviews, non-English papers, or those not involving participants’ perceptions. Results: Overall 22 eligible papers were included, giving a total of 25 research studies. Among these studies, Facebook and Twitter were the most common social media platforms investigated. The most effective communication style differed depending on the social media platform. Factors reported to impact credibility included: language used online, expertise heuristics, and bandwagon heuristics. No papers were found that assessed authenticity. Conclusions: Credibility and authenticity are important concepts studied extensively in the marketing and communications disciplines, however, further research is required in a health context. Instagram is a less-researched platform in comparison to Facebook and Twitter.
Article
Full-text available
In this study we examined the effect of disclosure language (control/no disclosure, “SP,” “Sponsored,” and “Paid Ad”) in Instagram-based influencer advertising on ad recognition, brand attitude, purchase intention, and sharing intention among a sample of 237 students. In line with prior research, results indicated that disclosure language featuring “Paid Ad” positively influenced ad recognition, which subsequently interacted with participants' memory of a disclosure and mediated the effect of disclosure language on attitude toward the brand and sharing intention. The findings offer a significant contribution to the literature on consumers' information processing and understanding for new and developing native advertising executions. Theoretical and managerial implications are discussed.
Article
Full-text available
Social media, such as Facebook, offer brands the opportunity to reach their target audience in a less obtrusive way than traditional media, through sponsored posts. Regulations require marketers to explicitly inform consumers about the commercial nature of these posts. This study addresses the effects of sponsorship disclosures by means of a 2 (no disclosure vs. the sponsorship disclosure ‘Sponsored’) × 2 (source: celebrity endorser vs. brand) experiment. Results suggest that a sponsorship disclosure only influences the use of persuasion knowledge when the post is disseminated by a celebrity. Moreover, a disclosure starts a process in which the recognition of advertising (i.e., the activation of conceptual persuasion knowledge) causes consumers to develop distrusting beliefs about the post (i.e., higher attitudinal persuasion knowledge), and in turn, decreases their intention to engage in electronic word of mouth.
Article
Full-text available
This research examines the impact of irrelevant information and its valence (positive or negative) on consumers' evaluations, choices, and post-choice satisfaction, within the context of online reviews. We demonstrate that seemingly irrelevant online reviews can enhance positive impressions, but only if they are labeled with a negative valence (e.g., with a one-star rather than a five-star rating). A series of studies provides support for this positive effect of negatively valenced irrelevant information; namely, the inclusion of a negatively valenced irrelevant review alongside positive reviews leads to greater product preferences, as consumers feel confident that the information they have about the product is more complete. We also demonstrate the moderating role of review source.
Article
Brand equity components, brand attitude and brand name familiarity, influence not only the positive benefits but also the adverse consequences that follow brand leveraging. Making use of event study methods, the authors assess whether and how the stock market return—a measure of the change in expected future cash flows—associated with a brand extension announcement depends on these brand equity components. The empirical analysis indicates that stock market participants’ responses to brand extension announcements, consistent with the trade-offs inherent in brand leveraging, depend interactively and nonmonotonically on brand attitude and familiarity.
Article
The proliferation of covert online advertising formats such as advergames has raised concerns about consumers’ ability to recognize such content as advertising and about how recognition affects evaluative outcomes. The present research utilized an online experiment (N = 179) to examine differences between covertness of advertising format (advergame vs. online video commercials) on advertising recognition, and whether sponsorship transparency mediates – and mitigates – the negative effects of advertising recognition on attitude toward the ad, attitude toward the brand, and purchase intent. The results show that covertness of advertising format, recognition, and sponsorship transparency all shape consumers’ responses to online ads. Specifically, the predicted negative indirect effect of covertness of advertising format on attitudes and behavioral intention via advertising recognition reversed valance when sponsorship transparency was included in the serial mediation model.
Article
Findings of two experimental studies show that Instagram influencers with high numbers of followers are found more likeable, partly because they are considered more popular. Important, only in limited cases, perceptions of popularity induced by the influencer's number of followers increase the influencer's perceived opinion leadership. However, if the influencer follows very few accounts him-/herself, this can negatively impact popular influencers’ likeability. Also, cooperating with influencers with high numbers of followers might not be the best marketing choice for promoting divergent products, as this decreases the brand's perceived uniqueness and consequently brand attitudes.
Article
Advertisers are continuously searching for new ways to persuade children, which includes fully integrating commercial content into media content, actively engaging children with the commercial content, and increasing the amount of commercial messages a child is confronted with at one moment in time. This poses a challenge for how children cope with embedded advertising. This conceptual paper aims to develop a theoretically grounded framework for investigating how children process embedded advertising. More precisely, it sheds light on previous research and conceptualizations of advertising literacy and provides suggestions for future research. The paper examines conceptual and methodological issues and discusses the need for research on how to improve children’s coping with embedded advertising by emphasizing the value of persuasive intent priming and implementation intentions. To conclude, future research directions are discussed regarding strategies to strengthen children’s coping skills and their dispositional (i.e., associative network consisting of cognitive, moral, and affective beliefs related to advertising) and situational (i.e., actual recognition of and critical reflection on advertising) advertising literacy.