Making the cut into consumer memory
The effect of online product placement and native advertising
Master’s Thesis
Graduate School of Communication Master’s programme Communication Science
Name: Lydia Papatheodorou Student ID number: 12456403
Supervisor: Hilde Voorveld Date: 20-06-2020
Abstract
The aim of the present study was to discover the differences in the way various types of branded content, namely online product placement and native advertising, affect the formation of consumer’s evoked set. Moreover, it addressed the role of perceived persuasive intent as the underlying mechanism that would explain the different responses of consumers towards online product placement vs. native advertising. The differential effect of the social media platform, through which the branded content was presented, was also taken into account. Results of a 2 (products placement vs. native advertising) X 2 (Instagram vs. TikTok) online experiment (N = 155) revealed that online product placement is slightly more effective than native advertising. Neither the type of platform nor perceived persuasive intent were found to affect this relationship. For marketers, these results imply that online product placement might be more effective than native advertising in conveying the desired message to their audience.
Introduction
The 21st century, along with the digital revolution it has brought about, has utterly changed the way in which brands approach consumers. The rise of digital media platforms has offered
marketers around the world new opportunities and challenges to reach their audiences.
Advertising has evolved into “branded content” and is used in two ways; a) in the form of a
commercial embedded within the newsfeed of social media platforms, also referred to as
“native advertising” (Campbell and Marks, 2015; Matteo and Dal Zotto, 2015) and b) as product placement embedded within media content, that is now delivered via closed-access,
online media platforms (e.g. Netflix) or short-video platforms (e.g. as Tiktok). The main
difference between the two methods lies in their delivery; native advertising directly
references the brand/ product (most commonly through an endorser/ influencer), while
product placement does not directly mention it.
Both methods are very expensive in terms of creation and delivery. However, research
comparing these two forms of online advertising is scarce, especially in terms of newly
created digital platforms. Thus, marketers are mostly guessing the form of branded content
and type of platform they should use to communicate their messages. In 2019, $105.94 billion
were spent on online advertising (PQ Media, 2020). Correct allocation of funds is of essence,
so as for the marketer’s message to have the strongest possible effect. Consequently, it is equally important that more research on the different effects of the aforementioned branded
content types is conducted.
As brands strive to remain relevant, their main goal is that their own products are the
first to come to consumers’ minds, when considering a purchase. In other words, one of their
main concerns is whether they are part of the consumer’s evoked set (Belch and Belch, 2018).
Steps have been taken into that direction with the integration of branding into social media
technique – in terms of top-of-mind-awareness and purchase intention - has yet to be
discovered.
Product placement, as content integrated branding has become increasingly popular
both in terms of academic research and professional practice. Many studies have already been
conducted on the effect of product placement on brand recall and brand attitudes (Gupta and
Gould, 2007; Van Reijmersdal, Neijens and Smit, 2007; Van Reijmersdal, 2011; Van
Reijmersdal, Tutaj and Boerman, 2013), but only a few on the effects of online product
placement, which might differ from traditional branded placement, due to the way in which
the content is consumed by the audience. Differences in the way each medium is used can
alter the effects product placement has on the audience (Matrix, 2014; Dagnino, 2018). This
might also extend to social media with different content and layout. To be more specific,
platforms such as TikTok allow the viewer to keep watching the content non-stop, whereas in
traditional media, the product placement is viewed at a specific time. New content platforms
allow for constant watching and, as such, constant exposure to the placement. Thus, the
placement becomes more natural and the viewer conceives it as less invasive.
Moreover, little is known about the effectiveness of product placement compared
to traditional advertising (Todor, 2016), and even less so for native advertising (Wojdynski
and Golan, 2016). Both tools are extensively used by advertisers. However, as the inherent
unobtrusive nature of online product placement may affect differently the evoked set of
consumers, than the comparatively more direct approach of native advertising, it is the
perceived persuasive intent (henceforth PPI) as an underlying mechanism of branded
content’s effectiveness that needs to be explored.
As social media are omnipresent and information flow is ambient, it is crucial to
assess the differences between the two methods. In line with Russel’s, Stern’s and Stern’s
branded product placement embedded in online networks, thus exploring different types of
new media. In this way, it aspires to shed light on “specific media related effects on consumer
behavior” (Russel, Stern and Stern, 2006, p. 16).
Marketing communications literature has, in general, focused on PPI of the source of a
message (Batada and Borzekowski, 2008; Kunkel, 2010; Tutaj and Van Reijmersdal, 2012;
Van Noort, Antheunis and Reijmersdal, 2012), without, however, examining the ways it
might influence the effect of native advertising and product placement. In her research on
radio-content embedded branding, Van Reijmersdal (2011) mentions that PPI is of importance
but measures the message’s persuasiveness per se and not whether the message was perceived
by the consumers as an attempt to induce a specific behavior.
What is more, as different social media platforms are used for different types of
content (photos, videos, opinion posts, etc.), it is possible that one type of branded content
may be more suitable for certain social media platforms than another. For example, short
video content engages users more deeply and by doing so users are more susceptible to the
content’s message. Thus, short video platforms, such as TikTok, may enhance the effect of branded content type on consumer’s evoked set, compared to other platforms, such as Instagram, which has a more varied content (Xiao, Wang and Wang, 2019).
Drawing from the above, the following research questions were formulated:
RQ1: To what extent does product placement on online platforms vs native advertising affect
the formation of consumer’s evoked set? Does PPI influence the effect of the different types of branded content?
RQ2: How does the type of platform (Instagram vs TikTok) affect the relationship between
Theoretical Background Defining Branded Content
The term “branded content” has been used in marketing communication literature to describe different aspects of what could be considered as the same concept. Horrigan (2009) defines
branded content as the merging of advertising with entertaining content created and delivered
by the brand, and uses the term as an alternate to “branded entertainment”. Branded
entertainment, in turn, is explained as product placement intertwined within entertainment
content, thus playing an important role in the evolution of the storyline (Hudson and Hudson,
2006). For other researchers, branded content goes hand in hand with social networking sites
(henceforth SNS) and refers to online content produced and “posted” by the brand, in the same way as consumer posts with the aim of stimulating interaction with prosperous
consumers (Sabate, Berbegal-Mirabent, Cañabate and Lebherz, 2014).
The confusion around the definition of branded content can be explained by the fact
that the concept was developed before the appearance of digital media. Digital media and
especially the opportunity for brands to promote themselves through them in a
non-interruptive and unobtrusive way is relatively new. Facebook, for example, which is the most
popular SNS with more than 2.6 billion active users per month (Statista, 2020a) first
introduced the option of advertising in November 2007 (Facebook Unveils Facebook Ads,
2007). Since then, digital media and the consequent SNSs evolved, leading to the need for a
new definition. Thus, Asmussen, Wider, Williams, Stevenson, Whitehead and Canter (2016)
re-introduced “branded content” as an overarching term describing any type of output, aligned
with the brand’s fundamentals with the aim to attract audiences in engaging with the brand. The primary reason branded content was adopted as a marketing tool is twofold;
firstly, it is the means of brand evolution into the prevalence of the digital world. Secondly, it
online environment has equipped consumers with, such as adblocking software (Jiang,
McKay, Richards and Snyder, 2017).
In general, “branded content” covers three types of created and distributed content: a) content delivered by the brand’s owned media (SNS accounts that belong to the brand), b) the so-called “native distribution of marketers’ paid content” (e.g. product placement) and c)
“material hosted by, or made by, publishers”, also mentioned as “native advertising”
(Asmussen et al., 2016; Hardy, 2018). The present study adopts the definition of branded
content as a term that covers any type of content that has an unobtrusive and brand-related
nature and compares the latter two different types of branded content, namely product
placement and native advertising, in terms of achieving top-of-mind-awareness on part of the
receivers. The decision to exclude “owned media” from the analysis is based upon the fact
that the topic has been thoroughly researched in terms of its effectiveness (Fournier and
Avery, 2011; Lovett and Staelin, 2016), reach and attitude formation (Boerman, Willemsen
and Van Der Aa, 2017; Zarouali, Ponnet, Walrave and Poels, 2017) with results that show that
solely relying on owned media in the digital era would be an ineffective marketing strategy.
Hence, this research focuses on the two other branded content categories.
The Evoked Set
A measure of effectiveness for online product placement and native advertising used by the
brands, is whether the brand became part of the consumer’s evoked set. Being amongst the
first products that come to the mind of a consumer considering a purchase from a specific
product category, is one of the ultimate goals for which marketers employ branded content
(Belch & Belch, 2018). Being part of the consumer’s evoked set means that every time the consumer is considering a purchase of a certain type of product, the individual will think of
the brand, and is thus a point between considering a product and buying it (McGuire, 1976).
associative network, which leads to brand loyalty (Solomon, 2004). By means of native
advertising and product placement the brand aims to be better remembered and preferred to its
competitors.
Cognitive Processes of Evoked Set Formation
As the evoked set depends on remembering the brand and intending to purchase its products,
it involves two stages: 1) making it into consumer’s memory and 2) persuading the consumer about its valuable or useful characteristics (May, 1979). However, the latter is not as
necessary, as – for low risk purchases - consumers have the tendency to eliminate cognitive
load, by automating the decision process and sufficing for what they can remember (Radder
and Huang, 2008).
Petty’s and Cacciopo’s (1986) Elaboration Likelihood Model (ELM), affirms that the information presented to an individual is processed through one of two processing routes; the
central (whereby the individual deliberately focuses on information) and the peripheral route
(whereby information processing is incidental and happens not due to elaborate cognitive
thinking, but through affective cues, e.g. through heuristics). Thus, processing of information
can also happen unintentionally.
Schroeder (2005) found that both routes result in learning. However, as information
storage in long-term memory depends on the attention that was paid to it, it is evident that
information stored through the central route creates stronger associations with pre-existing
knowledge and thus is easier retrievable. The evoked set is strongly associated with the
retrieval stage of the information processing model, in which a stimulus causes the
recollection of the mental representation of the information, in this case – the brand (Lang,
2000).
Being a part of the evoked set means that the consumer, apart from remembering the
towards the brand is also an important procedure. Attitudes have been found to be significant
predictors of purchase intention, which in turn has been found to be a significant predictor of
purchase behavior. In the early stage of getting acquainted with a brand attitude formation
also takes place, and the more positive the attitude is the stronger the intention becomes
(Fishbein and Cappella, 2006). What is more, attitudes created by active processing (central
route) are stronger and, as such, more influential than those formed merely through heuristics
(peripheral route) (Lang, 2000). According to the ELM, the cognitive paths that lead to
memory and attitude formation are closely related to each other.
The ELM and Native Advertising
The way information is processed explains why native advertising is an effective marketing
tool. Native advertising is defined as “the practice by which a marketer borrows from the
credibility of a content publisher by presenting paid content with a format and location that
matches the publisher’s original content” (Wojdynski and Golan, 2016, p. 1403). Though a lot has been said about the unethical issues that are involved with the specific marketing strategy,
as the commercial content is “disguised” in the users newsfeed (Matteo and Dal Zotto, 2015; Wojdynski and Golan, 2016; Taylor 2017), it has also been argued that it offers marketers
with a solution to the audience’s shift in attention from traditional to online media content, inviting them to engage with the brands (Campbell and Marks, 2015) and has proved to be a
necessary means against consumer avoidance (Zarouali et al., 2017).
However, there are distinctive characteristics that help recognize it as
paid-by-the-brand message. For example, such a post will usually be accompanied by of one the following
hashtags; #ad, #advert, #advertisements, #sponsored, etc. (Baker and Rojek, 2020). This
practice has been implemented for a while now and social media users are aware of the
specific marketing tool. In other words, although native advertising content is published in a
As a result, native advertising has a clear selling point and encourages cognitive
thinking. Either because a trusted person suggests a product (Van Noort, Antheunis and Van
Reijmersdal, 2012) – in this case an influencer or a SNS friend who has reposted the native
advertising post - or because its utilities and/or effects are described in the post, native
advertisement posts invite the user to actively think on the brand/ product. Thus, processing
through the central route is performed leading also to attitude formation.
It should be noted that, during the exposure to a native advertisement, peripheral route
processing is also taking place, as the audience’s positive attitudes towards the publisher of the content operate as a contextual cue that leads to positive attitudes towards the brand. Thus,
the incoming information is peripherally processed in a positive way, as posited by the
spillover effect (Chaturvedi Thota, Hee Song and Biswas, 2012). Consequently, as native
advertising is processed through both the central and peripheral routes, strong nodes are
created in consumers memory and strong positive attitudes are formed, leading to the addition
of the brand into their evoked set.
The ELM and Product Placement
Product placement is a relatively old term in communication science literature, first defined
by Balasubramanian (1994) as a branded message incorporated in such a way in a movie or
television program so that it can unobtrusively affect the viewers. The main advantage of this
method is that it sets the product in its context of use, thus creating associations between the
brand and everyday situations that individuals find themselves in (Baker, Hutchinson, Moore
and Nedungadi, 1986). Product placement, in general, has been found to positively affect
consumer’s memory of the brand and product, though sometimes negatively affecting the audience’s attitudes towards the company, as the inclusion of its product or brand name in their entertainment content of choice is perceived as intrusive and unethical (Van Reijmersdal,
also appears in other types of content, such as music videos, song lyrics, video games, radio
programs, book narratives (Gupta and Lord, 1998) and social media.
However, the most relevant medium of product placement for the present research
refers to short video platforms that are relatively new in the digital world – TikTok, for
instance, was created in late 2016 and now has more than 800 million users worldwide
(Clement, 2020), while Instagram stories where introduced in 2016 (Introducing Instagram
Stories, 2016). Marketing tools that are being created and used in these platforms are referred
to as short video marketing and share the same characteristics as tools used in other social
media platforms such as: a) influencers and b) entertainment content (Xiao, Wang and Wang,
2019). Videos produced on these platforms may be short, but they do present certain
characteristics: they tell a story. The advantage short video platforms have is that they provide
the best layout for product placement, as a brand can be an element of the presented video,
without it clearly addressing the consumer. Most common content includes challenges, satires
of sociopolitical context, how the users are spending their time, etc.. In many cases brands are
featured in them, without being explicitly referenced. Thus, the brands are considered less
intrusive in SNS user’s lives. The brands used are the props and their placement within the
content is perceived as natural.
Nowadays, product placement has evolved even more in the context of these platforms
in such a way that it ties a product or brand with the presented content, so that it plays a key
role in the story-line, in other words, the brand or product gives meaning to the viewed
content (Hardy, 2018). For example, in the beginning of 2020 during the Corona virus
(COVID-19) outbreak, many videos were created satirizing the reaction of many countries,
the optimistic idea that 2020 would be a “good year” or merely presenting with humor the “social distancing” practice, adopted worldwide. One of the trending videos featured a Corona beer bottle terrifying with its presence the rest, unnamed beers on a kitchen shelf. This
placement features the Corona beer and would not make sense with any other beer brand
whatsoever. However, it is appealing to the audience due to its humorous content.
Consequently, it is evident that online product placement is mostly dependent on
peripheral processing. The individual does not deliberately notice the brand and process its
message. It happens though heuristics, as the appearance of logo and the spillover effect of the
enjoyable content to the brand. As a result, nodes formed in long -term memory through this
marketing tool are not as strong, as those created through native advertising. The same applies
for the positive attitudes formed through the spillover effect, which, is mostly dependable on
contextual cues (Chaturvedi Thota, Hee Song and Biswas, 2012). In other words, product
placement is less likely to affect the evoked set than native advertising.
Drawing from the theories analyzed above and by bearing in mind that product
placement on short-video social media relies mostly on heuristics, whereas native advertising
posts incorporated in social media feeds also present some information on the product’s
attributes and characteristics, the present study suggests that:
H1: Native advertising is more successful in adding the marketed brand in the consumer’s
evoked set than product placement on digital social media.
Branded Content and Perceived Persuasive Intent (PPI)
Even though the brand-related information may be remembered by the consumer, it is
possible that it doesn’t end up in the evoked set. Some information may end up in the consumers inept set, which is a set of choices a consumer knows about and would not
consider purchasing, due to formation of negative attitudes, bad reputation, etc. (Narayana
and Markin, 1975). An underlying mechanism that can explain why branded content may
have resulted formation of negative attitudes is considered to be the PPI of the message. This
argument is presented in accordance with reactance theory, which posits that human beings
source poses a threat to it. Essentially, reactance is a reactive response to an external source
that is trying to affect the individual. (Miron and Brehm, 2006).
In the world of marketing and branded messages, the theory is found implemented as
follows; individuals realize that the message they are receiving (e.g. tv advertisements,
branded content, etc.) are brands’ attempts to convince them to purchase their products. Their
response, thereof, is to adopt the opposite behavior and not consider any purchase from the
brand.
Many studies have been conducted measuring this type of reactance. The most
astounding findings suggest that the purpose of marketing communications is, in general,
accepted. However, as selective avoidance to branded messages was developed by consumers
and online branded content was adopted by brands, the following paradox was reported: when
it isn’t clear that a message is intended to persuade consumers to make a purchase, the message is positively evaluated. However, consumers who are presented with the same
message and understand its underlying persuasive intent, evaluate it poorly (Van Noort,
Antheunis and Van Reijmersdal, 2012; Boerman, Willemsen and Van Der Aa, 2017; Zarouali,
Ponnet, Walrave and Poels, 2017).
When it comes to marketing tools, PPI seems to be the result of persuasion
knowledge. Persuasion knowledge refers to the level of the individual’s understanding of a
message as advertising (Friedstad and Wright, 1994). Persuasion knowledge is cultivated with
time and exposure to advertisements and native advertising is no exception. The more
experience an individual has with native advertising, the easier it is to understand its nature
(Jung and Heo, 2019), thus increasing the level of PPI.
As native advertising has long been incorporated in social media feeds, users now not
only expect to be exposed to them, but they can evaluate whether a post is trying to sell
words, SNS users have increased persuasion knowledge of this marketing tool and are able to
recognize native advertising as a message sponsored by the brand (Johnson, Potocki and
Veldhuis, 2019). Thus, it is believed that native advertisements result in high PPI on the
users’ part.
It should be noted, that as persuasion knowledge is a cognitive experience that
depends on the realization that the individual is exposed to a branded message, it involves
deliberate thinking, so central route processing. Boerman, Tessitore and Muller (2019) in their
research on long term effects of branded placement disclosure on persuasion knowledge and
its derivatives (ad skepticism, etc.) found that though in the long-run persuasion knowledge
was increased, its derivatives were not. This can be attributed to the fact that product
placement mostly relies on the peripheral route. The individual may not notice the featured
brand. Thus, when not explicitly invoked, persuasion knowledge is not activated and the PPI
of the message is low. As a result, it is believed that product placement leads to lower PPI
than native advertising.
Based on these premises, the following hypothesis was formulated:
H2: PPI mediates the relationship between branded content and inclusion in the evoked set in
such a way that native advertising’s PPI is higher than product placement’s in digital media
platforms, therefore, brands which use product placement have a higher chance of inclusion in
the consumer’s evoked set than brands which use native advertising. The Moderating Role of Platform Type
As PPI is the result of the activated persuasion knowledge, which depends on experience with
the branded content, it is possible that different platforms – with different age and branded
content integration – lead to different levels of it. More specifically, PPI might be dependent
The various social media platforms have different layouts and so it is possible that not
all branded content tools are (equally) compatible with all the platforms. In this research the
main focus will be on Instagram and TikTok due to the fact that a) they share certain
characteristics and b) TikTok is quite new and hasn’t been the focus of a lot of marketing communications research, so far, though it has recently caught researchers’ attention (Kumar,
2020; Wang, 2020). As far as the commonalities between the two platforms are concerned,
both allow for short videos to be shared. TikTok allows only the use of this feature, while
Instagram allows the user to do so in the section of “Instagram Stories” where a video or
picture is shown (for up to a minute) and then another one succeeds it. Both formats allow for
ephemeral reproduction. However, should the individual choose to do so, they can view the
shared content again. In both platforms, the “moving on to the next video” is similar
(vertically in TikTok, horizontally in Instagram). Instagram, however, also allows users to
post pictures and videos in a different form that is later presented it its news feed.
However, there is an essential difference between the two platforms; TikTok was
created in the end of 2016 (Clement, 2020), while Instagram had already been in circulation
for more than 6 years (Instagram, 2020). As a result, TikTok - being relatively new - has
drawn less attention in its selling perspective. Thus, users are less acquainted with its
advertising aspect, which in hand results in lower levels of persuasion knowledge on their
behalf. What is more, as it is a short-video platform, it can be used by marketers with the
following benefit; viewers are so focused on the content that they don’t realize that the brand
is trying to affect their purchase intentions. The brand is integrated in the content and its
appearance seems natural. Although the use of branded content as a marketing tool is known,
its inclusion in TikTok is new. Therefore, persuasion knowledge for the specific platform is
the other hand, Instagram advertising attempts are more likely to be recognized (Johnson,
Potocki and Veldhuis, 2019).
Following this rationale, it is possible that the platform used (namely Instagram and
TikTok) alters the effect of branded content type on PPI and evoked set. As a result, the
following explorative research question was formulated:
RQ3: How does the type of platform (Instagram vs TikTok) affect the same relationship when
also accounting for PPI?
Method Design and Sample
For the purposes of the present research a 2 (product placement vs. native advertising) x 2
(TikTok vs Instagram) between subjects online embedded experiment was conducted. An
experiment was chosen as the appropriate research method, as it allows control over the
conditions, which ensures strong internal validity (Bryman, 2012). The confirmation of the
causal relationship between the variables can then serve as a background to carry out studies
on the participant’s personal environment, allowing for generalization of the results. The online nature of the experiment was chosen as it is easy to distribute and hence, allows for a
bigger sample. Also, the experiment was carried out during the COVID-19 outbreak, with
most countries having taken lockdown and social distancing measures. As a result, a
laboratory experiment would not be feasible.
The sampling method adopted was convenience sampling combined with the snowball
method. More specifically, as the study was online embedded, the link to the experiment was
sent to the researcher’s network via their social media accounts (What’s app, Facebook, Messenger, Instagram), inviting other social media users to participate and to share the link
with their network as well. The latter was used to achieve a bigger sample and to reduce bias.
results, the main purpose of the study was to prove a relationship between the variables, as
mentioned above. Therefore, the sampling methodology followed is not considered a threat to
the results of the study in hand.
In order to achieve a homogenous sample, participants should all be social media users
between 18 – 35 years old. This age cohort was chosen to avoid ethical dilemmas concerning
under aged participants. Another reason is that the social media platforms used in this
experiment are popular among people of the cohort (Statista, 2020b; Statista, 2020c). The
experiment was conducted with 228 participants. However, 57 participants were removed
because they were either below 18 or over 35 years old (8), were not able to see or listen to
the stimuli (6) or did not complete the questionnaire (43). The remaining sample was N = 155
participants (61.3% female, 36.8% male and 1.9% indicated that they did not wish to answer
this question). The mean age of the sample was M = 24.82, SD = 3.28. The majority lived in
Greece (41.9%) and the Netherlands (32.3%). Some of the participants resided in the U.K.
(6.5%), Germany (4.5%) Spain (1.9%), U.S.A. (1.9%), Italy (2.5%) and the rest from various
countries around the world.
Stimuli
Branded content type was operationalized as a two-group categorical level variable through a
video per category. Both videos presented similar content and had the same duration. The
brand featured in the video was fictional, in order to ensure that pre-existing evoked sets
would not contain it (Geuens and De Pelsmacker, 2017). What is more, the researcher created
a “brand” on a product category, where there aren’t universal product category leaders, frequently encountered in advertising. However, involvement with the product should not be
low. On that account, the chosen product category was sanitary gloves. As the experiment was
conducted within the COVID-19 outbreak, sanitary gloves were a product that supermarkets
In order for the account holder likeability and familiarity to not influence the results of
the experiment, the researcher created an unknown account that belonged to a “meme” site. Such accounts attract many followers and raise to influencer status. What is more, an account
of a meme website instead of a random person ensured that individual perceptions of
attractiveness, etc. would not bias the results.
The video content for product placement condition featured an “online game” that was
trending at the time of the experiment, with a humorous content. When watching the video,
the viewer heard a well-known song performed live on a concert (the song used was
“Someone like you”, by British artist Adele). In the beginning, only the singer was heard. At that point, the viewer watched one glove, while hearing the first verse. Later on, the singer
would let the audience carry on with the song, at which point the viewer watched the camera
turn to the opposite side, were many gloves appeared, producing an illusion that they were the
singing audience. The brand name appeared on a box, on which the original glove was set,
referring to the stage and again in the end, referring to people who watched the concert from
above, without tickets.
For the native advertising stimulus, the video content was slightly different, in order to
match the usual feed of said social media. An ongoing meme trend was joking on people who
had their birthdays during the months April and May 2020. As worldwide quarantines were
implemented, people should stay at home and socializing centers were closed. Therefore, in
this case the video showed a hand setting in place the sanitary gloves in order to participate in
the trending game described above. The video showed first the hand taking a glove out of it
box – where the brand name appeared – then putting the glove on the ground next to many
other gloves (the exact same setting as the gloves of the product placement video) and lastly
putting them up where the “concert outsiders” would be. Special care was taken so as for the
under the video that usually accompanies such posts was “What to get people born in April and May to help them celebrate”. The song “Someone like you” was also featured in this
video, starting and finishing at the same point as in the product placement video. In this way
the spillover effect that could happen for the product placement video, due to its humorous
content, would also occur in the native advertising condition, which also conveyed a
humorous message. What is more, the brand name and the general video layout appeared in
the same song parts for both videos. Therefore, the conditions are considered comparable and
no other factors might have influenced the results. Hashtags referring to the brand, the product
and disclosure (#ad) were used in all videos, as they are normally used in the platforms’ posts. As far as the platform type was concerned, participants viewed either the native
advertising video, that was on an Instagram feed or TikTok, featured as a sponsored post, or
the product placement one, either as an Instagram story or as TikTok feed. The reason behind
this is that content videos are more likely to appear on the story section of Instagram, rather
than its feed, while TikTok only has one feed. Any other differences depended on the
platform (e.g. the place where the account holder’s name appeared). Procedure
As mentioned above, participants were approached through social media and were given
access to the experiment through a hyperlink. After accessing the experimental material, they
were confronted with a consent form, in which they were informed on the way their data
would be collected and analysed. After giving their consent they were asked to indicate
whether they were 18 years of age or older, for ethical reasons. During the experiment, they
were first confronted with an instruction sentence, informing them that they would see a video
created by a well-known satiric meme account and therefore they should make sure that they
had the sound on their device turned on. In this way, the researcher aimed at giving the SNS
exposed to the video-stimuli, to which they were randomly assigned by the experiment’s
platform.
After their exposure to the stimuli, they moved on to the questionnaire, which
consisted of four parts: a) the dependent variable measures, b) the mediator measures, c) the
manipulation check and d) the control variable measures. This order was chosen so as for the
stimulus effects to be as recent as possible on the consumers and for the scales of the mediator
and manipulation check to not predispose the participants’ answers regarding the dependent
variable measurement (Geuens and De Pelsmacker, 2017).
After answering all the questions, they were thanked for their participation and given a
concrete explanation on the aim of the experiment. On the same page the researcher’s contact
details appeared, in case they would like to receive information on the results of the
experiment.
Measures
Dependent variable: evoked set
The dependent variable was measured by three close-ended questions, adapted from LeBlanc
and Turley (1994). As the brand assessed in this experiment was fictional, participants were
asked to imagine themselves in different situations, in which they would have to buy products
and indicate which brand from a given list they would consider purchasing. The products
referred to were antiseptic gel, sanitary gloves and surgical masks. More products apart from
gloves that were included in the experiment were assessed in order to test whether the
response in the glove question is given as a result of the manipulation, and whether the mere
appearance of the brand in the videos created associations not only with gloves but with other
products as well. All the brands contained in the lists were fictional, in order to avoid
exclusion of the shown brand from their evoked set due to pre-disposition towards other
A principal axis factor analysis (PAF) was conducted showing that all items formed a
single uni-dimensional scale: only one component was found with an eigenvalue above 1
(eigenvalue 1.52), explaining 50.61% of the variance in the original items. The dependent
variable was computed into an interval level variable, by adding the participants’ scores of all
relevant items. A Pearson’s correlation was run between the items that showed a significant
relationship between all of them (rab = .24, pab = .003, rac = .27, pac. = 001 and rbc = .29, pbc
< .001).
Mediator: PPI
PPI was measured as a continuous level variable on a ten-item, seven-point Likert scale, on
which participants should indicate their level of agreement, ranging from “Totally Disagree” to “Totally Agree”. The items used in the measure were the six measures developed by Boerman, Van Reijmersdal, Rozendaal and Dima (2018) and four extra items that were used
as fillers, in accordance with the researchers’ recommendations. The items used were: “The reason AntiMic is shown in the video is:” a) “to stimulate people to want the advertised
brand”, b) “to encourage people to buy the brand”, c) “to sell products”, d) “to make people think positively about the brand”, e) “to attract attention to the brand”, f) “to make people remember the brand” and the fillers were: “The reason AntiMic is shown in the video is:” a)
“by accident”, b) “to make the video appealing”, c) “to help in the realistic presentation of the humorous video” and d) “For no reason”. The fillers were not assessed in the factor
construction. The rest of the items were incorporated into one variable for PPI based on their
mean ( M = 4.33, SD = 1.45, α = .94). Low scores indicate low PPI, and higher ones the
Control variables: age, gender, nationality, social media use, attitude towards the material
The researcher controlled for demographic variables, whose effect could threaten the internal
validity of the experiment. Their measurement was included in the final part of the
questionnaire, as demographic details.
Age. Age was measured by an open-ended question, asking participants to fill in their
age in an integral number, on a ratio level. To be sure that the participants would comprehend
how the number should be filled in, the researchers included an example.
Gender. Gender was measured on a categorical level, by a semi-open-ended question,
where participants were asked to indicate the gender they identify with among the following
options; a) male, b) female, c) Other (please indicate) and d) I would prefer not to say.
Residence. Place of residence was also measured on a categorical level, by a
close-ended question, in which participants should indicate by choosing from a drop-down menu.
Social Media Use: This variable was measured on an interval level, six-item,
seven-point Likert scale, adapted from Panek (2004). Participants were asked to indicate how often
they use their social media accounts ( a) Facebook, b) Twitter, c) Instagram, d) Snapchat, e)
TikTok and f) Youtube) on a scale from one to seven (where: “1” stands for “I do not use the application”, “2” stands for “less than once a day”, “3” stands for “once a day”, “4” stands for “two to three times a day”, “5” stands for “once an hour”, “6” stands for “two to three times an hour” and “7” stands for more than three times an hour”). Social media use was computed with the mean of the items, M = 2.96, SD = .75, α = .90.
Attitude towards the material. This variable was measured on a four item, seven-point
semantic scale adapted from Orús, Gurrea and Flavián (2017), in which participats had to
indicate how they found the video. The options were “unpleasant-pleasant”,
into one variable based on their mean ( M = 3.48, SD = 1.51, α = .91) after deleting the item
“unpleasant-pleasant”, in order to increase the reliability of the scale, α = .39.
Device. Participants were also asked to indicate what type of device they used to fill in
the questionnaire. They could choose from the following categories: a) mobile phone, b)
tablet, c) laptop, d) desktop computer and e) other.
Results Control Variables’ Randomization Check
To check whether the sample characteristics varied across the manipulated groups, namely
branded content type and platform type, randomization checks were run on the control
variables. As shown below, none of the variables were found to significantly differ between
the examined groups, and thus, there was no reason to treat them as covariates.
The randomization checks for age, social media use and attitude towards the stimuli
were carried out through two-way analyses of variance with branded content type and
platform type as predictors and each of the aforementioned variables as the outcome variable.
The results showed that age (F1(1,154) = 1.41, p1 = .238 and F2(1,154) = .001, p2 = .971),
social media use (F1(1,154) = .87, p1 = .354 and F2(1,154) = .324, p2 = .570) and attitude
towards the stimuli (F1(1,154) = 2.56, p1 = .111 and F2(1,154) = .684, p2 = .409) were all
evenly distributed between the groups of branded content type and platform type respectively.
Three chi-square tests were run for gender (Ficher’s exact p1 = .867, φ1 = .02 and
Ficher’s exact p2 = .738, φ2 = .04), residential area (, χ21(20,154) = 18.16, p = .577, Crammer’s
V2 = .34 and χ22(20,154) = 21.70, p = .357, Crammer’s V1 = .36 ) and device used (χ21(1,154) =
2.01, p = .366, Crammer’s V1 = .11 and χ22(1,154) = .004, p = .998, Crammer’s V2 = .01) that
didn’t reveal any significant differences between the groups of branded content type and platform type respectively. It should be noted that for “gender” to meet the chi-square test
from the analysis. To be certain that the removed values were not influencing the results, a
univariate analysis of variance was run, that showed that gender is not a predictor of evoked set
formulation, F(2,154) = .22, p = .801. Residential area also did not meet the chi-square test
assumptions. However, most countries mentioned appeared once, thus this variable was
excluded from further analysis. As the sample distribution of none of the variables mentioned
above was found to differ among the examined groups, they were not taken into account in the
main analysis.
Manipulation Check
A manipulation check was added in the end of the experiment, asking the participants
to indicate – on a range from 0 to 6 - the level upon which they believe that the shown video
content differs from the content they usually encounter on their social media feeds.
Main Analysis
The effect of branded content type on evoked set
To address the first hypothesis concerning the direct effect of branded content type on the
evoked set, an independent samples t-test was performed with branded content type as the
independent variable and evoked set as the dependent one. The results revealed a marginally
significant difference between the groups, t(147.1) = 1.71, p = .09, d = .32. The target brand
was incorporated slightly more often in the evoked set of the participants after being exposed
to brand placement (M = .59, SD = .88) than native advertising (M = .37, SD = .67).
Therefore, the first hypothesis was tentatively rejected.
The mediating role of PPI
The mediating role of PPI intent in the relationship between branded content type and evoked
set, described in the second hypothesis, was assessed by the use of Hayes’ PROCESS macro
v.3.2. As the described relationship refers to a simple mediation, Model 4 with bootstrapping
PPI as the mediator and inclusion in the consumer’s evoked set as the outcome variable. The
assessment revealed that no mediation effect takes place in the process, as the regression
model accounted for 2.12% of the variance in the dependent variable (R2 = .021) and was not
found to be significant, F(2,152) = 1.65, p = .196.
The direct effect of the predictor on the dependent variable was once again found to be
marginally significant, t = 1.67, p = .0973. Thus, all else held equal product placement is
slightly more likely to succeed in adding the brand/ product in the consumer’s evoked set that
native advertising by b = .21, SE = .13 in the population. However, no significance was found
in the mediation analysis for neither the first part of the mediated relationship, namely the
effect of branded content type on PPI ( t = .36, p = .717) nor for the second part, the effect of
PPI on the consumer’s evoked set ( t = .66, p = .508). It should be noted that the scores of
both groups on PPI were quite similar M = 4.34 for the native advertising group and M = 4.42
for the product placement group, correspondingly. These results were also confirmed by the
indirect effect assessment, b = .003, SE = .02, 95%CI [-.02, .03]. Consequently, the second
hypothesis was rejected as well.
The moderating role of platform type
To assess whether platform type moderated the relationship between branded content type and
evoked set, a two-way analysis of variance was conducted with branded content and platform
type as the predictors and evoked set as the outcome variable. The moderation effect was not
found to be significant F(1,154) = .06, p = .802. The mean differences for those who were
exposed to product placement on Instagram (M = .49, SD = .13) and TikTok (M = .67, SD =
.12) and for those who were exposed to native advertising on Instagram (M = .32, SD = .12)
and TikTok (M = .44, SD = .14) were similar. Neither the direct effect of branded content type
(F(1,154) = 2.50, p = .116), nor that of platform type (F(1,154) = 1.50, p = .222 ) were
.37, SD = .67) and product placement (M = .59, SD = .88) and ii) between TikTok (M = .57,
SD = .82) and Instagram (M = .40, SD = .76) was similar for all groups. Consequently,
platform type does not moderate the direct effect of branded content on evoked set and the
second research question was refuted.
Assessment of the moderated mediation model
The fourth research question concerned a moderated mediation effect of branded content type
on evoked set with PPI as the mediator and platform type as the moderator variable. Hayes'
PROCESS macro v.3.2 was once again employed, with the use of model 8 and bootstrapping
set at 5000 to address it.
The moderated mediation model described above was not significant as b = .02, SE =
.03, 95%CI [-.03, .11]. This can also be confirmed by the results concerning the moderating
role of platform type on the relationship of branded content on PPI (R2 = .03, F(3,151) = 1.31,
p = .127.) and on relationship of branded content type on evoked set (R2 = .03, F(4,150) = 1.29, p = .275.). As a result, the final research question was also refuted.
Discussion
The purpose of this study was to assess whether different types of branded content affected
the inclusion of the promoted brand in a consumer’s evoked set in a different way. The role of platform was also considered as its interaction with the type of branded content could produce
different results on the effect on evoked set. The role of PPI as an underlying mechanism that
could explain the differences in the evoked set formation caused by branded content and
platform type was also assessed. The results revealed that the aforementioned relationships,
addressed by the study in hand, do not exist, even though they suggest that product placement
might be slightly more effective than native advertising in adding the promoted brand to the
consumer’s evoked set, regardless of the PPI of the placement and the platform it was displayed on.
It was hypothesized that native advertising would be more successful in adding the
product in the consumer’s evoked set than product placement, as, according to the ELM, attitudes and memories formed through the central route are stronger than those formed
through the peripheral one (Lang, 2000; Petty and Cacioppo, 1986). The rationale behind this
hypothesis refers to the fact that while product placement on social media lies merely within
the scope of heuristics, native advertising affects both processing routes by informing the
potential consumer about the brand’s attributes and placing it within an appealing context. The present study also took reactance theory into account, considering that when
branded content was noticed as such, and thus the PPI of the message increased, inclusion of
the brand in the evoked set would fail, as consumers have the tendency to avoid the influence
of external factors on their choices and behaviors (Miron and Brehm, 2006). As product
placement is reliant upon heuristics, it is believed to cause less reactance and its persuasive
intent to be mostly unnoticed, and in this case be more successful than native advertising.
Neither product placement nor native advertising were found to lead to great
differences in the formulation of the consumers’ evoked set. Nonetheless, it should be noted
that product placement was revealed to be slightly superior to native advertising with regards
to in its effectiveness in evoked set formation, which is quite promising on the differential
effects of various branded content types on consumer memory and purchase intention. A
possible explanation for the disconfirmation of the hypothesized relationship between the two
branded content types and the evoked set formulation is that they weren’t going through their
own or another person’s social media page when coming across the stimuli. Even though the present experiment was online, so for participants to answer in their “natural habitat”,
disrupting as little as possible its ecological validity, their exposure to the stimuli was not
natural, in that sense. This, in combination with the fact that they were aware of their
In this way, it is possible that both the central and peripheral processing route of the ELM
were activated, which wouldn’t have happened in everyday circumstances, hindering the differences in effects of each type of branded content.
As far as PPI is concerned, its role as an underlying mechanism of the relationship
between the examined branded content types on evoked set formulation, was not proven. As
both groups perceived the intent of the presented videos in a similar way, it comes as no
surprise that it was not found to mediate the presented relationships. Once again this can be
explained by the fact that the participants were aware of their inclusion in a study, so they
probably were more skeptical towards the presented brand, since they had never heard of it
before. What is more, even though the questions were set in such a way so as for their order to
not bias the answers to the scales (Geuens and De Pelsmacker, 2017), the evoked set’s
measures might have increased their skepticism. So even though the PPI was, in, fact lower
during the measurement of the dependent variable (which was presented first), their
perception of the stimuli changed after exposed to the aforementioned scale, that hinted them
on the purpose of the video, resulting in a higher level of PPI. This is an issue that arises in
many cases with self-reports, as PPI was not measured in the time of the presentation of the
stimuli, but later, after being exposed to items asking for their purchase intentions. As a result,
their reactance grew, and a relationship between the variables was not found.
As far as the different effects of platform type on the relationship of branded content
type with consumer evoked set are concerned, once again no differential role was uncovered.
A reason why this may have happened is that even though the presented videos were
recordings of posts on the corresponding social media platforms, the platform layout was only
presented in the video. On the other hand, in actual circumstances people are aware of which
social media platform they will use, they are exposed to its logo first and start scrolling
as the video was shown after showing introductory guidelines. Thus, it might not have been
enough to represent the platform and assess its effect. In a real-life environment, consumers
come across various social media posts before and after their exposure to branded content. In
this case, especially for native advertising, the experiment depended on their already formed
perceptions of what social media content should look like, as there was no other post before
and after the branded content viewing. It was not possible to overcome this barrier in the
design of the present experiment, as the guidelines had to be given to the participants at all
phases.
Finally, the present study addressed a moderated mediation model, representing an
overall effect of branded content type on evoked set, through PPI that depends on social
media platform type. As none of the suggested independent relationships were found to be
statistically significant, it came as no surprise that the proposed relationship was not verified.
However, as the research plan of this study was based on previous empirical studies on the
ELM and reactance theory and was in line with their claims, it should not be definitive upon
the proposed relationships. On the contrary, future studies could focus on unveiling the
reasons behind the non-significant results of the present one.
Limitations
A limitation of the present study was its sample size and consistency. Even though more than
220 participants entered, it ended up with 155, as responses were removed due to technical
issues regarding the stimuli, questionnaire incompleteness, not meeting the set age
requirements or not agreeing with the use of their collected data. The specified “rules” of
empirical studies and statistical tests that call for at least 30 participants per condition were
met, but as a moderated mediation model was proposed, it would be suitable to have about 60
What is more, since the present experiment was not carried out in a lab, it remains
unclear to what degree the participants paid attention to the study. Multitasking was not
controlled for. However, if the participants were also listening to music or active on their own
social media accounts, it is possible that their attention to the study was diminished and the
hypothesized processing of the native advertising condition through the central route may
have not taken place, explaining the lack of statistically significant results.
Finally, the time frame in which the present study was conducted, did not allow for
repeated exposure of the branded content. Many studies in the field of advertising in general,
and especially internet advertising have found that repeated exposure to the ad increases brand
awareness and recall, as well as attitude towards the brand and, therefore purchase intention -
Fishbein and Capella, 2006) - which are both essential to the evoked set formation
(Ehrenberg, 2000; Drèze and Hussherr, 2003; Schmidt and Eisend, 2015). However, only one
exposure was measured in the present study, which can account for the weak effect of
branded content type on consumer evoked set.
Drawing from above, future research should address similar relationships allowing for
repeated exposure to stimuli. This way, it will be closer to everyday life settings and natural
exposure, as social media advertising is repetitive. Repetitive exposure could also have
different effects on PPI, in line with the proposed hypothesis, as repetitive exposure to
branded content that is processed through the central route may have different effects than
branded content type that is processed through the peripheral route both PPI and evoked set
formation. It should be noted that this proposition does not refer to the addition of another
variable to the proposed model, but to the reassessment of the model in closer to real-life
circumstances.
Furthermore, future research could be employed with developing a new measurement
evoked set formation is mostly inferential addressing brand recall and attitude. A scale
combining the two or a new one would benefit many studies to come, as it would measure the
variable directly and more in line with recent technological developments in the area of
branded content and marketing. Though the definition of evoked set has not changed, as it still
refers to the top brands or products that come in mind when a person is considering a
purchase (Belch and Belch, 2018), its measurement should. As the information flow is
ambient, the new measurement should be able to capture more delicate differences in the way
consumers consider a brand for purchasing purposes.
Practical Implications
Even though the relationships examined in this study were not found to be significant, the
slight difference in the effectiveness of online product placement against native advertising on
evoked set formation, suggests that its results should not be discarded. There are various
reasons that might have affected the described relationships that are thoroughly addressed
above. Therefore, the present study can be used as groundwork for future study conduct. The
slightly stronger effect of online product placement compared to native advertising suggests
that repetition of the present study or conduct of other similar studies can lead to more
significant results, that will help in the development of relevant guidelines for practitioners,
on the way the specific marketing tool is used.
Marketers and advertisers especially would benefit a lot from such research as it
would become apparent which branded content type better suits their need of promotion.
Also, they could differentiate the branded content format they use, taking into account the
individual differences of social media platforms. Especially when it comes to Instagram and
TikTok, their effectiveness on evoked set formation was found to be the same, thus they could
be used interchangeably or complementarily, for the age cohort 18 – 35. From the results of
agencies should choose product placement to deliver their message and be fondly recalled by
consumers, when the latter are faced with purchase decisions. In this way, their message
References
Asmussen, B., Wider, S., Williams, R., Stevenson, N., Whitehead, E. & Canter, A. (2016).
Defining branded content for the digital age. The industry experts’ views on branded content as a new marketing communications concept. A collaborative research project
commissioned by the BCMA and conducted by Oxford Brookes University and Ipsos
MORI.
Baker, S. A., & Rojek, C. (2020). Lifestyle Gurus: Constructing Authority and Influence
Online. Cambridge: Polity Press.
Baker, W., Hutchinson, J., Moore, D., & Nedungadi, P. (1986). Brand familiarity and
advertising: effects on the evoked set and brand preference. ACR North American
Advances, 13, 637-642.
Balasubramanian, S. K. (1994). Beyond advertising and publicity: Hybrid messages and
public policy issues. Journal of advertising, 23(4), 29-46.
https://doi.org/10.1080/00913367.1943.10673457
Batada, A., & Borzekowski, D. (2008). Snap! Crackle! What? Recognition of cereal
advertisements and understanding of commercials' persuasive intent among urban,
minority children in the US. Journal of Children and Media, 2(1), 19-36.
https://doi.org/10.1080/17482790701733179
Belch G.E., & Belch M.A. (2018). Advertising and Promotion, an Integrated Marketing
Communications Perspective. McGraw Hill-Education.
Boerman, S. C., Willemsen, L. M., & Van Der Aa, E. P. (2017). “This post is sponsored”:
Effects of sponsorship disclosure on persuasion knowledge and electronic word of
mouth in the context of Facebook. Journal of Interactive Marketing, 38, 82-92.
Boerman, S. C., Tessitore, T., & Müller, C. M. (2020). Long-term effects of brand placement
disclosure on persuasion knowledge and brand responses. International Journal of
Advertising, 1-23. https://doi.org/10.1080/02650487.2020.1775036
Boerman, S. & Van Reijmersdal, E. (2016), "Informing Consumers about “Hidden”
Advertising: A Literature Review of the Effects of Disclosing Sponsored Content", De
Pelsmacker, P. (Ed.) Advertising in New Formats and Media, Emerald Group
Publishing Limited, pp. 115-146. https://doi.org/10.1108/978-1-78560-313-620151005
Boerman, S. C., van Reijmersdal, E. A., Rozendaal, E., & Dima, A. L. (2018). Development of
the persuasion knowledge scales of sponsored content (PKS-SC). International Journal
of Advertising, 37(5), 671-697. https://doi.org/10.1080/02650487.2018.1470485
Bryman, A. (2012). Social research methods. N.Y.: Oxford University Press.
Campbell, C., & Marks, L. J. (2015). Good native advertising isn’t a secret. Business
Horizons, 58(6), 599-606. https://doi.org/10.1016/j.bushor.2015.06.003
Chaturvedi Thota, S., Hee Song, J., & Biswas, A. (2012). Is a website known by the banner ads it hosts? Assessing forward and reciprocal spillover effects of banner ads and host websites. International journal of advertising, 31(4), 877-905.
https://doi.org/10.2501/IJA-31-4-877-905
Clement, J. (2020). Topic: TikTok. Statista. Retrieved from
https://www.statista.com/topics/6077/tiktok/
Dagnino, G. (2018). Regulation and co-regulation of product placement for OTT SVODs: The
case of Netflix. International Journal of Digital Television, 9(3), 203-218.
https://doi.org/10.1386/jdtv.9.3.203_1
Drèze, X., & Hussherr, F. X. (2003). Internet advertising: Is anybody watching?. Journal of
Ehrenberg, A. S. (2000). Repetitive advertising and the consumer. Journal of Advertising
Research, 40(6), 39-48. https://doi.org/10.2501/JAR-40-6-39-48
Facebook Unveils Facebook Ads. (2007). About Facebook. Retrieved from
https://about.fb.com/news/2007/11/facebook-unveils-facebook-ads/
Fishbein, M., & Cappella, J. N. (2006). The role of theory in developing effective health
communications. Journal of communication, 56, S1-S17.
https://doi.org/10.1111/j.1460-2466.2006.00280.x
Fournier, S., & Avery, J. (2011). The uninvited brand. Business horizons, 54(3), 193-207.
https://doi.org/10.1016/j.bushor.2011.01.001
Friestad, M., & Wright, P. (1994). The persuasion knowledge model: How people cope with
persuasion attempts. Journal of consumer research, 21(1), 1-31.
https://doi.org/10.1086/209380
Geuens, M., & De Pelsmacker, P. (2017). Planning and conducting experimental advertising research and questionnaire design. Journal of Advertising, 46(1), 83-100.
https://doi.org/10.1080/00913367.2016.1225233
Gupta, P. B., & Lord, K. R. (1998). Product placement in movies: The effect of prominence
and mode on audience recall. Journal of Current Issues & Research in Advertising,
20(1), 47-59. https://doi.org/10.1080/10641734.1998.10505076
Hardy, J. (2018). Branded Content: Media and Marketing Integration In H. Powell, J. Hardy,
S. Hawkin & I. MacRury (4). The advertising handbook (pp. 102-117). Routledge.
https://doi.org/10.4324/9781315558646
Horrigan, D. (2009). Branded content: A new model for driving tourism via film and branding
strategies. TOURISMOS: An International Multidisciplinary Refereed Journal of
Hudson, S., & Hudson, D. (2006). Branded entertainment: a new advertising technique or
product placement in disguise?. Journal of Marketing Management, 22(5-6), 489-504.
https://doi.org/10.1362/026725706777978703
Instagram (2020). About us. Instagram. Retrieved from https://www.instagram.com/about/us/
Introducing Instagram Stories. (2016). About Instagram. Retrieved from
https://about.instagram.com/blog/announcements/introducing-instagram-stories
Jiang, M., McKay, B. A., Richards, J. I., & Snyder, W. (2017). Now you see me, but you
don't know: Consumer processing of native advertisements in online news
sites. Journal of Interactive Advertising, 17(2), 92-108.
https://doi.org/10.1080/15252019.2017.1399839
Johnson, B. K., Potocki, B., & Veldhuis, J. (2019). Is that my friend or an advert? The
effectiveness of Instagram native advertisements posing as social posts. Journal of
Computer-Mediated Communication, 24(3), 108-125.
https://doi.org/10.1093/jcmc/zmz003
Jung, A. R., & Heo, J. (2019). Ad disclosure vs. ad recognition: How persuasion knowledge
influences native advertising evaluation. Journal of Interactive Advertising, 19(1),
1-14. https://doi.org/10.1080/15252019.2018.1520661
Kumar, G. (2020). Tik Tok Advertising policies in India: An Overview. Studies in Indian
Place Names, 40(69), 297-305.
Kunkel, D. (2010). Commentary Mismeasurement of children's understanding of the
persuasive intent of advertising. Journal of Children and Media, 4(1), 109-117.
https://doi.org/10.1080/17482790903407358
Lang, A. (2000). The limited capacity model of mediated message processing. Journal of
LeBlanc, R. P., & Turley, L. W. (1994). Retail influence on evoked set formation and final
choice of shopping goods. International journal of retail & distribution management,
22(7), 10-17. https://doi.org/10.1108/09590559410069882
Lovett, M. J., & Staelin, R. (2016). The role of paid, earned, and owned media in building
entertainment brands: Reminding, informing, and enhancing enjoyment. Marketing
Science, 35(1), 142-157. https://doi.org/10.1287/mksc.2015.0961
Matrix, S. (2014). The Netflix effect: Teens, binge watching, and on-demand digital media
trends. Jeunesse: Young People, Texts, Cultures, 6(1), 119-138. DOI:
10.1353/jeu.2014.0002
Matteo, S., & Dal Zotto, C. (2015). Native advertising, or how to stretch editorial to
sponsored content within a transmedia branding era. In Handbook of media
branding (pp. 169-185). Springer, Cham.
https://doi.org/10.1007/978-3-319-18236-0_12
May, F. E. (1979). Evoked set formation and composition: the learning and information
processing hypotheses. Advances in Consumer Research, 6(1), 222-226.
McGuire, W. J. (1976). Some internal psychological factors influencing consumer
choice. Journal of Consumer research, 2(4), 302-319. https://doi.org/10.1086/208643
Miron, A. M., & Brehm, J. W. (2006). Reactance theory-40 years later. Zeitschrift für
Sozialpsychologie, 37(1), 9-18. https://doi.org/10.1024/0044-3514.37.1.9
Narayana, C. L., & Markin, R. J. (1975). Consumer behavior and product performance: An
alternative conceptualization. Journal of Marketing, 39(4), 1-6.
https://doi.org/10.1177/002224297503900401
Orús, C., Gurrea, R., & Flavián, C. (2017). Facilitating imaginations through online product
