ABSTRACT
The key characteristic of native advertising is that it simulates the format of the editorial content. The current study examines how format similarity—the extent to which the format of an advertisement is similar to the format of the editorial content—affects consumers' responses to the advertisement and whether the effects differ by format novelty, the extent to which the format of the editorial content is novel or familiar to consumers. Results showed that format similarity increased perceived deceptiveness, and format novelty reduced advertising recognition and perceived irritation while inducing greater click intentions. Additionally, there was a significant interaction effect between format similarity and format novelty such that the negative impact of format similarity on perceived irritation was found when format novelty was low but not when format novelty was high. Theoretical and practical implications are discussed.
MANAGEMENT SLANT
Format similarity induces greater deceptive perceptions.
Format novelty induces lower advertising recognition, lower perceived irritation, and greater click intention.
There is a significant interaction between format similarity and format novelty such that the negative impact of format similarity on perceived irritation was observed only when format novelty was low, not when it was high.
INTRODUCTION
Native advertising is a popular form of advertising. According to eMarketer, native advertising was expected to account for approximately 60 percent of all U.S. digital-display advertising spending in 2019 (eMarketer, 2018). Native advertising is found on various sites, including social-networking sites such as Facebook, Twitter, and Instagram in the form of social native advertising, such as sponsored posts and promoted Tweets; and on news publishers' websites (e.g., The New York Times and The Huffington Post) in the form of article-style advertising.
Native advertising is an advertisement that is “cohesive with the page content, assimilated into the design, and consistent with the platform behavior” (Interactive Advertising Bureau, 2019, p. 11). Native advertising simulates editorial content using a similar format, similar content, and similar function and platform behavior. Although native advertising could be different from nonnative advertising in terms of its format similarity, content similarity, and functional similarity to editorial content, the most fundamental difference between native advertising and nonnative advertising is format similarity—that is, whether an advertisement simulates the format of editorial content.
Various definitions of native advertising highlight the importance of format similarity, compared with content similarity or functional similarity. Native advertising has been defined as “any paid advertising that takes the specific form and appearance of editorial content from the publisher itself” (Wojdynski and Evans, 2016, p. 157, emphasis added). Other definitions have emphasized “the similarity between the format of native advertising and its surrounding media content” (Lee, Kim, and Ham, 2016, p. 1427, emphasis added). Native advertising alternatively can be described as “a form of paid media where the commercial content is delivered within the design and form of editorial content” (Conill, 2016, p. 904, emphasis added). All of these definitions mention format similarity as the core component of native advertising.
Format similarity can have either positive or negative effects. Format similarity can induce positive responses to an advertisement by reducing advertising recognition or activating content schema. In contrast, format similarity can induce negative responses by generating deceptive or manipulative perceptions. Previous research that tested the effectiveness of native versus nonnative advertising has reported inconsistent findings regarding the impact of format similarity (Kim, Choi, and Kim, 2019; Kim, Youn, and Yoon, 2018; Park, Kim, and Lee, 2018). Because the role of format similarity is not yet clear, the current study examines the effect of format similarity.
The current study further examines whether the effect of format similarity might vary by format novelty. In particular, the extent to which a given format is novel or familiar can moderate the impact of format similarity. This is because format novelty can reduce the amount of cognitive resources available for critical processing of advertisements and suppress the activation of persuasion knowledge, which could be explained on the basis of the limited-capacity model of information processing (Lang, 1995, 2000). Format novelty thus either can increase the positive effects of format similarity or can reduce the negative effects of format similarity. The current study therefore tests the effect of format similarity as well as the moderating role of format novelty. This study can contribute to research on native advertising by illuminating and conceptualizing the two format aspects of native advertising: similarity and novelty.
LITERATURE REVIEW
Effects of Format Similarity
As mentioned, a key characteristic of native advertising is format similarity, the similarity between the format of an advertisement and the format of the surrounding media content. Unlike traditional online advertising, such as banner advertisements, native advertising does not have a fixed format. Instead, it adopts the format used by the surrounding media content.
The specific format that is adopted by a native advertisement depends on the type of native advertising. Social native advertising, such as sponsored posts or Tweets, for example, resembles a social-media post, whereas article-style native advertising takes the form of a news article. The specific appearance of a native advertisement differs by the type of website where it is presented. An article-style native advertisement presented on CNN's website, for example, looks different from that presented in The Huffington Post, because CNN and The Huffington Post use different formatting style to present their news articles. Although there are some differences across the different types of native advertising, all types of native advertising share a common feature, which is to simulate or imitate the format of the surrounding media content, whether that is a social-media post or a news article (Wojdynski and Golan, 2016).
Format similarity could have both positive effects and negative effects. On the one hand, format similarity may induce positive responses because it may reduce advertising recognition or activate editorial-content (nonadvertising) schema (Wojdynski, 2016a). Given the similar format shared with editorial content, consumers might not recognize easily that a native advertisement is an advertisement; if they do recognize it as an advertisement, activated content schema can lead it to be processed similarly to editorial content. In one study, fewer than half of the study participants were able to recognize native advertisements when they engaged in naturally browsing a website (Jiang, McKay, Richards, and Snyder, 2017). The reduction in advertising recognition and the activation of content schema can result in positive responses to an advertisement by suppressing the activation of persuasion knowledge.
The persuasion-knowledge model (Friestad and Wright, 1994) suggests that the target consumer's persuasion knowledge can play an important role in persuasion, such that the activation of persuasion knowledge can reduce persuasion. In particular, the persuasion-knowledge model suggests that a target consumer copes with persuasion attempts by interpreting, evaluating, and responding to an agent's persuasion intent (Friestad and Wright, 1994). When the target perceives that an agent's action has a persuasive intent, a change of meaning occurs (Friestad and Wright, 1994), which could cause disruptive message processing, such as discounting. The persuasion-knowledge model thus suggests that the activation of persuasion knowledge can inhibit persuasion (Campbell and Kirmani, 2000; Cowley and Barron, 2008; Hwang and Jeong, 2016; Nelson, Wood, and Pack, 2009; Wei, Fischer, & Main, 2008). Because format similarity can reduce the activation of persuasion knowledge, format similarity may result in positive responses to an advertisement.
On the other hand, format similarity may induce negative responses because of its potential deceptiveness or manipulativeness. Researchers have noted that native advertising can be perceived as deceptive (Amazeen and Wojdynski, 2018b; Schauster, Ferrucci, and Neill, 2016; Taylor, 2017; Wojdynski, 2016a, 2016b), because consumers may believe that advertisers intentionally present advertisements that look similar to noncommercial content to conceal that they are advertisements.
Deception can include two types of strategies: dissimulation and simulation (Boush, Friestad, and Wright, 2009; Whaley, 1982). Whereas dissimulation refers to “hiding the real,” simulation refers to “showing the false” (Johnson, Grazioli, and Jamal, 1993, p. 470). Native advertising with high format similarity can hide the fact that it is actually an advertisement by using editorial-content-like appearance. Perceived deceptiveness can induce perceived manipulativeness (Lee et al., 2016), which refers to “consumer inferences that the advertiser is attempting to persuade by inappropriate, unfair, or manipulative means” (Campbell, 1995, p. 228).
When consumers focus on the potential deceptive or manipulative aspect of format similarity, format similarity may induce greater activation of persuasion knowledge. This is because, according to the persuasion-knowledge model (Friestad and Wright, 1994), consumers have persuasion knowledge regarding the appropriateness of persuasive tactics. Persuasion knowledge includes consumers' beliefs about marketers' tactics, consumers' own coping tactics, the effectiveness and appropriateness of persuasion tactics, and the psychological mediators of tactic effectiveness. Because consumers have their own standard regarding the appropriateness of persuasion tactics, perceived inappropriateness can activate persuasion knowledge. When individuals think that native advertising attempts to deceive them using format similarity, critical processing of the advertisement ensues, and their responses to native adverting can be negative.
It is not yet clear whether format similarity has positive or negative effects. Given the potential positive and negative effects of native advertising, one study suggested that native advertising can be “a double-edged sword,” with nonintrusiveness on the one hand and manipulativeness on the other hand (Lee et al., 2016). The researchers noted that the nonintrusiveness feature of native advertising can play a positive role in the effectiveness of native advertising, whereas the manipulativeness feature of native advertising can play a negative role.
Despite the potentially interesting role of format similarity, little research has tested the impact of format similarity. Although some research has examined the impact of relevance or congruence between media content and advertising, these studies focused on content similarity of the advertisement and the surrounding media content, rather than format similarity. Past research has tested the effect of topical congruity or relevance between television program and advertisement product (Furnham, Gunter, and Richardson, 2002), between website content and banner-advertisement product (Moore, Stammerjohan, and Coulter, 2005; Shamdasani, Stanaland, and Tan, 2001), and between news article and banner advertisement (Chun, Song, Hollenbeck, and Lee, 2014).
The psychological process for the impact of format similarity, however, could be different from that for the impact of content similarity. Whereas the effect of content similarity could be explained on the basis of priming, the effect of format similarity could be explained on the basis of advertising recognition and persuasion knowledge. Research on the impact of format similarity thus warrants scholarly attention.
Although some research has compared the effectiveness of native versus nonnative advertising by varying format similarity (Kim et al., 2018, 2019; Park et al., 2018), the results are not entirely consistent. Researchers compared the effects of native versus banner advertisements and showed that there were no significant differences in perceived persuasive intent and attitude toward the advertisement between the two types of advertisement (Kim et al., 2018). The researchers found that a native advertisement induced greater advertising credibility compared with a banner advertisement, however. Another group showed that an advertisement with a native, versus nonnative, design induced more favorable attitudes toward the advertisement, more favorable attitudes toward the brand, and greater purchase intention (Kim et al., 2019).
Also, some researchers compared the effects of native versus nonnative advertisements and showed that there were no significant differences between the two types of advertisements in perceived advertisement intrusiveness, perceived advertisement irritation, attitudes toward the advertisement, attitudes toward the brand, and intention to touch the advertisement (Park et al., 2018). There were significant interaction effects, however, between native format and thinking styles on perceived advertisement irritation, attitudes toward the advertisement, and intention to touch the advertisement. The native advertisement induced greater perceived advertisement irritation and lower intention to touch the advertisement than the nonnative advertisement among analytic thinkers but not among holistic thinkers. Moreover, the native advertisement resulted in less-favorable attitudes toward the advertisement among analytic thinkers, but it induced morefavorable attitudes toward the advertisement among holistic thinkers. These researchers also showed that there was a significant interaction effect between native format and congruency on perceived advertisement irritation, such that perceived advertisement irritation increased for the native advertisement, versus the nonnative advertisement, when there was brand-application incongruency but not when there was brand-application congruency (Park et al., 2018).
In sum, previous research on the effect of native versus nonnative advertising has shown mixed results. Some studies have shown that native format is more effective, whereas other studies have shown that native format is not more effective or less effective. These results suggest that more research is needed to explore the effects of format similarity. The current study thus examines whether format similarity has positive effects or negative effects.
Effects of Format Novelty
Although native advertising adopts a format that is used by the surrounding editorial content, different publishers use different formatting styles to list their editorial content. The mobile page of The Huffington Post, for example, displays a list of its top stories by presenting an image on the left and text on the right, whereas the mobile web page of CNN provides a list of its news articles by presenting a big image box over a small text box. The different formatting styles could be more or less familiar depending on whether an individual has visited the publisher's website. Even if an individual has visited the publisher's website, he or she may find the format novel if there has been a change in formatting style.
The extent to which the formatting style of the surrounding editorial content is novel can influence people's response to the advertisement. This is because editorial content presented with a novel format can attract more attention and reduce cognitive resources required for advertising recognition and processing. The role of format novelty can be explained on the basis of the limited-capacity model (Lang, 1995, 2000), which suggests that humans have a limited amount of cognitive resources to process information. When some elements in media messages receive more attention or cognitive resources, cognitive resources available for other elements are reduced.
Given previous research suggesting that novelty attracts attention (Berlyne, 1950; Smith, Chen, and Yang, 2008; Smith and Yang, 2004; Wu and Huberman, 2007; Yang and Smith, 2009), information presented in a novel format could attract more attention and leave fewer cognitive resources available for other elements. Information presented in a novel format requires greater cognitive resources for processing, whereas information presented in a familiar format requires fewer cognitive resources. This is because prior exposure is believed to increase perceptual fluency (Bornstein and D'Agostino, 1994; Jacoby and Kelley, 1987), which refers to “the ease with which a person perceives and identifies the physical characteristics of a stimulus” (Lee and Labroo, 2004, p. 152). Compared with a familiar format, a novel format requires extra cognitive resources because of the lack of schema regarding the format.
Given the limited cognitive capacity in human information processing, format novelty could reduce cognitive resources available for critical processing of advertisements and subsequently could suppress the activation of persuasion knowledge. This is because the activation of persuasion knowledge itself requires cognitive resources. One study found that explicit sponsorship disclosure activated persuasion knowledge only when cognitive load was low, not when it was high (Wen, Kim, Wu, and Dodoo, 2020, Study 1).
Additionally, although not in the context of native advertising, past research has shown that cognitive overload can reduce the activation of persuasion knowledge and induce more positive responses to persuasion efforts. Researchers, for example, showed that cognitive load induced by multitasking reduced the negative effect of intrusive product placement (Yoon, Choi, and Song, 2011). Other researchers also showed that media multitasking increased cognitive overload and decreased counterarguing (Jeong and Hwang, 2012, 2015). A review of a large body of research on multitasking has shown that cognitive overload can suppress negative responses to advertising and enhance persuasion (Jeong and Hwang, 2016).
Given that format novelty can suppress the activation of persuasion knowledge, format novelty is expected to moderate the effects of format similarity. As mentioned earlier, format similarity may have either positive effects, by reducing the activation of persuasion knowledge, or negative effects, by increasing the activation of persuasion knowledge. Because format novelty can suppress the activation of persuasion knowledge, format novelty either can increase the positive effects of format similarity or can reduce the negative effects of format similarity. Either way, format novelty was expected to play a beneficial role in the effects of format similarity.
RESEARCH QUESTIONS
The current study examines the effects of format similarity and format novelty on advertising recognition, advertising evaluations (perceived deceptiveness, perceived irritation, and attitude toward the advertisement), and click intention. There has been much research on factors influencing individuals' recognition or identification of native advertising (e.g., Amazeen and Wojdynski, 2018a, 2018b; Wojdynski, 2016a; Wojdynski and Evans, 2016; Wojdynski, Bang, Keib, Jefferson, et al., 2017; Wu, Huang, Li, Bortree, et al., 2016). Little research, however, has examined the impact of format similarity and novelty on advertising Although past research has raised the issue of potential deceptiveness (Mudge and Shaheen, 2017; Schauster et al., 2016; Wojdynski, 2016a) and manipulativeness (Lee et al., 2016) of native advertising, little empirical research compares native advertising and nonnative advertising in terms of the deceptiveness and manipulativeness. Finally, past research on native advertising has examined the effects of format similarity on perceived irritation (Park et al., 2018), attitude toward the advertisement (Kim et al., 2018; Park et al., 2018), and click intention (Kim et al., 2018; Park et al., 2018) by comparing native and banner advertisements. Past research has not examined the role of format novelty, however. The current study thus examines the effects of format similarity and format novelty on consumers' responses to advertisements:
RQ1: How does format similarity influence consumers' responses to an advertisement (i.e., advertising recognition, advertising evaluations, and click intention)?
RQ 2: How does format novelty influence consumers' responses to an advertisement (i.e., advertising recognition, advertising evaluations, and click intention)?
RQ 3: Are the effects of format similarity moderated by format novelty?
METHODS
Design
This study employed a 2 (format similarity: high versus low) × 2 (format novelty: high versus low) between-groups experimental design. Study participants were assigned randomly to one of the four experimental conditions. Sample sizes for each group ranged from 74 to 80.
Participants
In an online experiment, 311 Korean adults participants were exposed to experimental advertisements and then completed a survey. Study participants could access the online link for the experiment using their computers or mobile phones. The online link was created and maintained by Macromill Embrain, which is one of the largest research firms in South Korea.
The participants were recruited from a panel of more than one million individuals maintained by Macromill Embrain. From the panel, 521 individuals were selected randomly and invited to participate in the study, and 311 respondents completed the task (response rate = 60 percent). Participants' ages ranged from 25 to 43 years (M = 33.63, SD = 5.50), and 50.2 percent of the participants were female.
Manipulations
Format similarity was manipulated on the basis of the extent to which the target advertisement used a format that was similar to the format of the surrounding editorial content. In other words, the authors manipulated format similarity by presenting an advertisement with a format that was similar to (high similarity) or different from (low similarity) the format of the surrounding content. The high-similarity (native) advertisements adopted the exact same format used by editorial content surrounding the advertisement. This is because the format of native advertisements is “context-specific,” reflecting the editorial environments (Wang and Li, 2017, p. 915).
If editorial content used a format showing a thumbnail image on the left and some text on the right, the high-similarity advertisement therefore used the same format (a thumbnail image on the left and text on the right). If editorial content used a format showing a big image box over a small text box, the high-similarity advertisement used the same format (a big image box over a small text box). The low-similarity advertisements, conversely, used a banner format.
The authors manipulated format novelty by varying the format of the surrounding editorial content (news articles). Participants in the high-novelty condition were shown editorial content in a novel format (i.e., CNN format), whereas those in the low-novelty condition were shown editorial content in a familiar format (i.e., Naver format). For a high-novelty format, the authors presented news articles with the format used by CNN mobile web (i.e., a big image box over a small text box), which is novel and unfamiliar to most Koreans. No major Korean websites have adopted the style used by CNN. For a low-novelty format, the authors presented news articles with the format used by Naver (i.e., a thumbnail image on the left and some text on the right), which is familiar to most Koreans. Naver is one of the most popular Korean news platforms, and, according to Nielsen Koreanclick (n.d.), Naver has the highest reach rate of all Korean websites.
Across all experimental conditions, the authors presented an image of a mobile page containing a list of six news articles and one advertisement. The advertisement was placed after the fourth article. The specific formats of the advertisement and the surrounding news articles varied in each condition.
Participants in the low-novelty and high-similarity condition saw a Naver-style native advertisement surrounded by Naver-style news articles. Participants in the low-novelty and low-similarity condition saw a banner advertisement surrounded by Naver-style news articles. Participants in the high-novelty and high-similarity condition saw a CNN-style native advertisement surrounded by CNN-style news articles. Last, participants in the high-novelty and low-similarity condition saw a banner advertisement surrounded by CNN-style news articles. The surrounding six articles covered various topics (i.e., “foods that cause inflammation,” “health benefits of bananas,” “at-home workouts,” “foods that boost the immune system,” “the first civilian pilot,” and “characteristics of people who maintain good relationship with others”).
Across all experimental conditions, the advertisement featured a hypothetical mouthwash brand, Gargle Fresh, and had the same advertisement copy claiming that the mouthwash can prevent bad breath. The advertisement contained the same visual in which a woman suggests that something smells bad by pinching her nose. The advertisements in all four experimental conditions contained the same sponsorship disclosure cue, with “AD” written in a white box on the right side of the advertisement.
Pretest
The authors conducted a pretest to ensure the comparability of the four advertisements. Sixty Korean adults participated in the pretest. Their age ranged from 31 to 55 years (M = 40.95, SD = 5.58), and 57 percent were men.
The authors created a Google survey link, which was sent out to individuals who agreed to participate in the pretest. In the Google survey link, each participant was shown one of the four different types of stimulus advertisements in the first page and then were asked to respond to questions about the advertisement and demographic information (age and gender) in the next page. The questionnaire specifically assessed participants' various perceptions of the stimulus advertisements.
Results showed that there were no significant differences across the four advertisements in terms of the following perceptions: The advertisement was attention getting, F(3, 56) = 0.44, p = 0.72; persuasive, F(3, 56) = 0.27, p = 0.85; believable, F(3, 56) = 0.40, p = 0.76; good, F(3, 56) = 0.52, p = 0.67; and enjoyable, F(3, 56) = 0.03, p = 0.99. Note that the same hypothetical brand (Gargle Fresh) was used across the four advertisements; the four advertisements thus seemed to be comparable.
Measures
Advertising recognition was measured on the basis of the item adapted from previous research (Rozendaal, Opree, and Buijzen, 2016). Respondents were asked whether there was an advertisement in the previous page they had just viewed. Four response options were provided: (1) “definitely yes,” (2) “probably yes,” (3) “probably no,” and (4) “definitely no.” Responses were recoded so that a higher score indicated greater recognition (M = 3.13, SD = 0.78). After participants' level of advertising recognition was assessed, study respondents were informed that the previous page contained an advertisement for a mouthwash brand and were asked about their perceptions regarding the advertisement.
Perceived deceptiveness was measured on the basis of items used in previous research (Wojdynski, 2016a; Wojdynski, Evans, and Hoy, 2018). The following three items were presented: “The mouthwash ad was trying to fool consumers into thinking it was not advertising,” “The mouthwash advertiser tried to obscure the fact that this was an ad,” and “The mouthwash ad tried to deceive the viewer about the fact that it was advertising.” Response options ranged from “strongly disagree” (1) to “strongly agree” (5), and the responses to the three items were averaged (Cronbach's α = 0.88; M = 3.30, SD = 0.95).
The authors measured perceived irritation by asking participants to report the extent to which the mouthwash advertisement was “irritating,” “annoying,” and “insulting people's intelligence” (Ducoffe, 1995, 1996). Response options ranged from “strongly disagree” (1) to “strongly agree” (5), and the responses to the three items were averaged (Cronbach's α = 0.89; M = 2.65, SD = 0.94).
The authors measured attitudes toward the advertisement by asking participants to report the extent to which the mouthwash advertisement was “bad” (1)–“good” (5), “negative” (1)–“positive” (5), and “unfavorable” (1)–“favorable” (5) on the basis of past research (Mitchell, 1986). Responses to the three items were averaged (Cronbach's α = 0.91; M = 2.83, SD = 0.89).
Click intention was measured on the basis of items used to measure behavioral intention in past research (Fishbein and Ajzen, 1975). The following two items were used: “I intend to click the mouthwash ad,” and “I will click the mouthwash ad.” Response options ranged from “strongly disagree” (1) to “strongly agree” (5), and the responses to the two items were averaged (r = 0.89; M = 2.33, SD = 1.05).
For manipulation checks, perceived format similarity and perceived format novelty were assessed. The authors measured perceived format similarity by asking participants to report the extent to which “the mouthwash ad is presented in a format that is similar to [resembles/is consistent with] the format used by the news articles surrounding the ad.” Response options ranged from “strongly disagree” (1) to “strongly agree” (5), and the responses to the three items were averaged (Cronbach's α= 0.91; M = 3.35, SD = 0.78).
Perceived format novelty was measured by the extent to which study participants thought that they were “familiar with [acquainted with] the format that news articles were presented.” Responses to the two items were reverse-coded and averaged (r = 0.84; M = 2.51, SD = 0.82). The authors also measured study participants' age, gender, and prior mouthwash usage for randomization checks.
RESULTS
Manipulation and Randomization Check
Results of manipulation checks indicated that the manipulations of format similarity and format novelty were successful. There was a significant difference in perceived format similarity, F(1, 309) = 5.32, p < 0.05, η2p = 0.02, between the high- and low-similarity conditions. Consistent with the manipulation, compared with respondents in the low-similarity condition (M = 3.38, SD = 0.72), those in the high-similarity condition (M = 3.59, SD = 0.87) reported greater perceived similarity between
the format in which the advertisement was presented, and
the format in which the news articles were presented.
There was also a significant difference in perceived format novelty, F(1, 309) = 8.36, p < 0.01, η2p = 0.03, between the high- and low-novelty conditions. Consistent with the manipulation, compared with respondents in the low-novelty condition (M = 2.38, SD = 0.80), those in the high-novelty condition (M = 2.65, SD = 0.83) reported greater perceived novelty with regard to the format in which the news articles were presented.
Results of randomization checks showed that there were no significant differences in the distribution of participants' gender, χ2(3, N = 311) = 0.90, p = 0.83; age, F(3, 307) = 0.14, p = 0.94; and prior mouthwash usage, F(3, 307) = 0.53, p = 0.66, across the four groups.
Results for Research Questions
Advertising Recognition. Results of a two-way analysis of variance (ANOVA) showed that there was a significant main effect of format novelty on advertising recognition, F(1, 307) = 35.26, p < 0.001, η2p = 0.10. (The authors conducted analysis of covariance (ANCOVA) with age as a control variable. Results of ANCOVA did not differ from results of ANOVA in terms of significance testing. The authors thus report ANOVA results.) The high-novelty condition (M = 2.88, SD = 0.77) resulted in lower advertising recognition than the low-novelty condition (M = 3.38, SD = 0.71). There was no significant main effect of format similarity, F(1, 307) = 0.89, p = 0.35, or an interaction effect between format similarity and format novelty, F(1, 307) = 0.13, p = 0.72.
Deceptiveness. There was a significant main effect of format similarity on perceived deceptiveness, F(1, 307) = 5.98, p < 0.05, η2p = 0.02 (See Table 1). The high-similarity condition (M = 3.43, SD = 0.91) resulted in greater perceived deceptiveness compared with the low-similarity condition (M = 3.17, SD = 0.96). There was no main effect of format novelty, F(1, 307) = 0.99, p = 0.32, or an interaction effect between format similarity and format novelty, F(1, 307) = 1.85, p = 0.17.
Irritation. There was no significant main effect of format similarity on perceived irritation, F(1, 307) = 2.57, p = 0.11. There was, however, a significant main effect of format novelty, F(1, 307) = 6.44, p < 0.05, η2p = 0.02. The high-novelty condition (M = 2.52, SD = 0.92) resulted in lower perceived irritation than the low-novelty condition (M = 2.79, SD = 0.94).
There also was a significant interaction effect between format similarity and format novelty, F(1, 307) = 6.83, p < 0.01, η2p = 0.02, such that the impact of format similarity on perceived irritation was observed when format novelty was low, but not when it was high. When format novelty was low, results of Scheffe's post-hoc tests showed that there was a significant difference in perceived irritation (p < 0.05) between the high- (M = 3.00, SD = 0.92) and low-similarity (M = 2.56, SD = 0.92) conditions. When format novelty was high, however, there was no significant difference in perceived irritation (p = 0.92) between the high- (M = 2.47, SD = 0.91) and low-similarity (M = 2.57, SD = 0.93) conditions (the results are presented in Figure 1).
Attitudes toward the Advertisement. There were no significant main effects of format similarity, F(1, 307) = 1.11, p = 0.29, or format novelty, F(1, 307) = 0.03, p = 0.87, on attitudes toward the advertisement. There also was no significant interaction effect between format similarity and format novelty, F(1, 307) = 1.62, p = 0.20, on attitudes toward the advertisement.
Although there was no direct effect of format similarity on attitudes toward the advertisement, a supplementary analysis showed that there was an indirect effect of format similarity on attitudes toward the advertisement via its impact on perceived deceptiveness and irritation (total indirect effect 95 percent confidence interval [CI] [−0.24, −0.03]). A mediation analysis using a bootstrapping analysis (Hayes, 2018) showed that format similarity induced greater perceived deceptiveness, which, in turn, led to less-favorable attitudes toward the advertisement (indirect effect via perceived deceptiveness 95 percent CI [−0.20, −0.02]). Format similarity also induced greater perceived deceptiveness, which led to greater perceived irritation, which in turn resulted in less favorable attitudes toward the advertisement (indirect effect via perceived deceptiveness and then perceived irritation, 95 percent CI [−0.04, −0.001]).
Click Intention. There was no significant main effect of format similarity, F(1, 307) = 1.25, p = 0.27, or an interaction effect between format similarity and format novelty, F(1, 307) = 0.30, p = 0.58, on click intention. There was, however, a significant main effect of format novelty on click intention, F(1, 307) = 4.89, p < 0.05, η2p = 0.02. Participants in the high-novelty condition (M = 2.46, SD = 0.98) showed a higher level of click intention compared with those in the low-novelty condition (M = 2.19, SD = 1.11).
DISCUSSION
This study showed that format similarity induced greater perception of deceptiveness, and format novelty induced lower advertising recognition, lower perceived irritation, and greater click intention. More important, the effect of format similarity was moderated by format novelty, such that a negative impact of format similarity on perceived irritation was observed only when format novelty was low, not when it was high.
The effect of format similarity on greater deceptiveness perception suggests that the manipulativeness feature of format similarity can have negative effects. Although the potentially negative impact of native advertising has been suggested in previous research (Lee et al., 2016), little empirical research directly has tested the negative impact of format similarity. The current study suggests the deceptiveness feature of format similarity can induce negative responses. Given that some previous research (Kim et al., 2019) has suggested that native advertising is more effective than nonnative advertising, presumably because of the nonintrusiveness feature of format similarity, future research needs to explore further the conditions under which the deceptiveness feature or the nonintrusiveness feature of native advertising might have greater effects.
Moreover, the current study focused on the role of format novelty as an important factor that can moderate the effects of format similarity. The results of this study show that the negative impact of format similarity on perceived irritation was found only when format novelty was low, not when it was high. The results suggest that when the format of the editorial content (surrounding the advertisement) was familiar, format similarity resulted in negative responses. When the format of editorial content (surrounding the advertisement) was novel, however, format similarity did not cause negative responses. This could be because format novelty induces cognitive overload, which in turn reduces advertisement recognition or critical processing of advertising. Because greater cognitive resources are allocated to the processing of editorial content presented with a novel format, a lesser amount of cognitive resources would be available to process critically the advertisement.
The moderating role of format novelty can be explained on the basis of differences in schema development. Consumers already have a well-developed schema for a familiar format, and they easily could distinguish an advertisement from editorial content and can respond critically to native advertising. For a novel format, however, consumers might not have a developed schema, and they may have difficulty distinguishing between advertising and editorial content. The results resonate with an earlier study (Jung and Heo, 2019) that showed that persuasion knowledge regarding social-media advertising tactics influenced the evaluation of native advertising. Across two studies, the researchers found that greater knowledge regarding social-media advertising tactics resulted in greater awareness of selling and persuasive intent regarding the native advertisement.
Furthermore, the moderating role of format novelty found in this study raises an interesting research question. The current study showed that native advertising with a novel format did not induce negative responses, whereas native advertising with a familiar format resulted in negative responses (greater perceived irritation). The result suggests a possibility that the effectiveness of native advertising may decrease as consumers become more familiar with the particular type of format. Future research thus could test whether the effectiveness of native advertising may decrease as consumers become more familiar with the format, on the basis of longitudinal data.The main effects of format novelty found in this study are noteworthy as well. Format novelty resulted in lower advertising recognition, lower perceived irritation, and greater click intentions. The effects of format novelty on advertising recognition and perceived irritation could be explained on the basis of the inhibition of critical advertising processing as a result of cognitive overload, and the effect of format novelty on greater click intentions could be explained by the role of novelty in inducing greater attention and curiosity. These results together suggest that format novelty could result in greater persuasion.
Theoretical Contributions
The current study has important theoretical contributions. First, this study can contribute to research on native advertising by conceptualizing and testing format characteristics, such as format similarity and format novelty, of native advertising. Much research on native advertising has tested the effects of sponsorship disclosure in native advertising (e.g., Amazeen and Wojdynski, 2018a; An, Kang, and Koo, 2018; Sweetser, Ahn, Golan, and Hochman, 2016; Wojdynski, 2016a; Wojdynski and Evans, 2016; Wojdynski et al., 2017), and this line of research has advanced the understanding of the role of sponsorship disclosure in native advertising. The current study further focused on format similarity as an important aspect of native advertising, and shows the roles played by format characteristics in the effects of native advertising.
The results of the current study do not provide a definitive answer to the question of whether native advertising is more effective than traditional, nonnative advertising. Although the current study tested the effect of the most important characteristic of native advertising, format similarity, and showed that format similarity can have negative effects, the effectiveness of native advertising might be influenced not only by format similarity but also by content similarity and functional similarity. To understand the effectiveness of native (versus nonnative) advertising, therefore, one must examine the independent and interactive effects of each aspect (format similarity, content similarity, and functional similarity).
This is because, although the current study found that format similarity could have some negative effects, content similarity and functional similarity could have positive effects. For example, content similarity could have positive effects because native advertising with high content similarity tends to contain a greater amount of information compared with nonnative advertising with low content similarity. A previous study compared a banner advertisement with sponsored content (news article) and found that the sponsored content (versus banner advertisement) was perceived as having greater information and amusement value (Tutaj and van Reijmersdal, 2012). Because native advertising can be different from nonnative advertising in terms of various attributes, such as format and content, the relative effectiveness of native advertising versus nonnative advertising should be answered on the basis of a series of research that tests the independent and interaction effects of these attributes.
The current study also can contribute to the native-advertising literature by explaining the effects of native advertising on the basis of various theoretical models, such as the persuasion-knowledge model (Friestad and Wright, 1994) and the limited-capacity model (Lang, 1995, 2000). The persuasion-knowledge model has been used in previous research on native advertising (e.g., Amazeen and Wojdynski, 2018a, 2018b); however, the current study applied the limited-capacity model to explain format-novelty effects in native advertising. Native advertising is a fairly novel type of advertising, and therefore critical processing of native advertising requires greater cognitive resources. Future research could manipulate cognitive load directly and examine the role it plays in the effects of native advertising.
Additionally, the current study can contribute to advertising research on contextual effects by examining format-based contextual effects. Whereas past research on contextual effects has focused on the impact of topic-based or content-based contextual effects (Chun et al., 2014; Moore et al., 2005), the current study examined the impact of format-based contextual effects. Given that new types of advertising, including native advertising and branded content, tend to adopt the format of noncommercial content, examining the impact of the format characteristics of the surrounding media content can be an important research topic. Future research further can examine the impact of format-based contextual effects by examining other factors, such as format complexity, and can explore the interactive effects between format-based contextual effects and content-based contextual effects.
Practical Implications
The findings have important practical implications. This study suggests that format similarity may not always induce positive responses, and thus native advertising might not be always more effective than nonnative advertising. This finding is somewhat unexpected given the growing popularity of native advertising.
Although the current study does not provide a definitive answer regarding the relative effectiveness of native (versus nonnative) advertising, it does suggest that format similarity can induce negative responses. Native advertising thus should be planned and executed carefully. Given that format similarity might increase irritation, particularly when the format of the surrounding content is familiar (or not novel), the effectiveness of native advertising might be maximized when the advertisement is placed in a novel publisher format rather than in a familiar publisher format. What's more, given the positive effects of format novelty, publishers can attract more attention to their editorial content as well as native advertising by updating its format regularly.
Limitations
This study is not without limitations. First, the manipulations of this study were not completely free from potential confounders. There were some differences in color (shading) of the advertisements across the conditions, for example. In the high-novelty conditions, there was no difference in the shading of the high- and low-similarity advertisements, whereas in the low-novelty conditions, there was some difference in the shading of the high- and low-similarity advertisements, and this might explain the significant interaction effect on perceived irritation.
The authors note, however, that the difference in shading (in the low-novelty conditions) was inevitable for a successful manipulation of format similarity. Without shading, the banner advertisement in the low-novelty condition appeared too similar to the surrounding news articles, and thus format similarity could not be manipulated successfully. The authors therefore decided to add shading to the banner advertisement in the low-novelty condition. The color of shading in the banner advertisement in the low-novelty condition was selected carefully to be consistent with the background (versus person) color of the image used in the native advertisement in the low-novelty condition. Additionally, the banner advertisement in the low-novelty condition was used as a prototype of the banner and native advertisements in the high-novelty conditions to minimize potential confounding.
[Note: The authors acknowledge the possibility of potential confounding due to shading and have collected additional data to address this issue. The authors collected additional data comparing a nonshaded (white) banner advertisement and a shaded native advertisement in the high-novelty conditions to see whether shading (and salience) might have caused the effect found in this study (i.e., the effect on perceived irritation). The authors examined whether there was a difference in perceived irritation when the banner advertisement in the high-novelty format condition did not have the shaded area whereas the native advertisement in the high-novelty format condition had the shaded area. Results based on 31 respondents showed that there was no significant difference in perceived irritation between the two types of advertisement, t(29) = 0.13, p = 0.90. Given that there was no significant difference in perceived irritation when there was a difference in shading (in the additional data), the significant effect of similarity in the original main data unlikely was caused by the difference in shading.]
Second, the current study did not measure positive outcomes of advertisement exposure; instead, it focused on negative outcomes, such as deceptiveness and irritation. This was because the current study expected that format similarity could have positive effects by reducing advertising recognition or negative effects by inducing greater deceptiveness and irritation perceptions. Because of this expectation, the authors focused on more negative outcomes (deceptiveness and irritation) instead of more positive outcomes. The authors also focused on more proximal and direct outcomes, such as attitudes toward the advertisement or click intention, and have not examined attitude toward the brand. Future research needs to examine more diverse outcomes, including attitude toward the brand.
Finally, this study employed two types of formats (CNN style and Naver style) to examine the effects of format similarity and novelty. Future research can use more diverse formats to generalize the findings of this study. The current study also used a banner advertisement to represent a low-similarity advertisement; however, low similarity does not necessarily have to take the form of banner advertisements. Future research thus can employ different types of advertisements and compare them with native advertisements to test the impact of format similarity.
CONCLUSION
The current study examined the effectiveness of native advertising by investigating the roles of format similarity and novelty. This study found that format similarity induced greater deceptive perceptions, whereas format novelty induced lower advertising recognition, lower perceived irritation, and greater click intention. Also, there was a significant interaction between format similarity and format novelty such that the negative impact of format similarity on perceived irritation was observed only when format novelty was low, not when it was high.
The current study can contribute to research on native advertising by conceptualizing format characteristics (e.g., format similarity and format novelty), and it can contribute to advertising research on contextual effects by examining the effects of format-based contextual effects. Future research can examine further the effectiveness of native advertising by focusing on other format characteristics (e.g., format complexity) and other attributes of native advertising (e.g., functional similarity).
ABOUT THE AUTHORS
Yoori Hwang is professor at the Department of Digital Media, Myongji University, in Seoul, Korea. Her research interests include branded content and marketing communication. Hwang's research has been published in the Journal of Advertising Research, Journal of Advertising, Journal of Communication, and Media Psychology, among other journals.
Se-Hoon Jeong is professor at the School of Media and Communication, Korea University, in Seoul, Korea. His research interests include media uses and effects in the context of strategic and marketing communication. Jeong's work can be found in the Journal of Advertising Research, Journal of Advertising, Journal of Communication, and Media Psychology, among other journals.
- Received April 4, 2019.
- Received (in revised form) December 4, 2019.
- Accepted January 2, 2020.
- Copyright© 2020 ARF. All rights reserved.
REFERENCES
ARF MEMBERS
If you are a member of the Advertising Research Foundation, you can access the content by logging in here
Log In
Pay Per Article - You may access this article (from the computer you are currently using) for 30 days for US$20.00
Regain Access - You can regain access to a recent Pay per Article purchase if your access period has not yet expired.