Brand- Self Congruence on Attitudes Towards Ads, Click- Through Intention,
and OBA Acceptance: An Experimental Study About Personally Targeted Ads
on Facebook
Renée Bajić 11821191
Master’s Thesis
Graduate School of Communication
Master’s programme Communication Science Supervisor: James Slevin
Abstract
Facebook Ads is a product for individuals and businesses that offers several advertising strategies to target audiences. Consumers see the product on the Facebook newsfeed as personally targeted ads. Besides the traditional location and demographics strategy,
retargeting and lookalike are the most successful ones. However, interest strategy sometimes delivers incongruent information, which can be improved by matching more congruent ads to the consumer’s interests.
Interest strategy uses Online Behavioral Advertising (OBA), which is the method responsible for delivering personally targeted ads that are formed on consumer’s data that was collected about their online activities. Because it is depicted as an invasive method of targeting, consumers are avoiding it and generally have inconsistent attitudes towards the ads that appear on their newsfeed.
This study proposes that more favorable attitudes towards ads, higher click intention, and higher acceptance of OBA depend on the congruence between the ad and consumer’s interests. As a result, the theory of brand- self congruity was applied as a theoretical
framework for this study and an experimental research design was conducted in an attempt to solve these issues.
However, the quantitative approach used did not contribute to the theoretical and conceptual body of knowledge of this research. Brand-self congruence theory seems not to explain consumer’s attitudes towards personally targeted ads, their intention to click on the ad, and neither does increase the likelihood of accepting OBA.
The study results deliver practical advices to business managers about what should they focus on when using Facebook interest strategy and thus, is of interest for businesses that want to
increase revenues with online advertising. It delivers recommendations to advertisers as well, which can contribute in creating more acceptable audience targeting instruments for the consumer. Additionally, it contributes to the literature of self- congruence theory by suggesting new theoretical developments.
Keywords: personally targeted ads; interest advertising strategy; online behavioral advertising
Brand- Self Congruence on Attitudes Towards Ads, Click- Through Intention, and OBA Acceptance: An Experimental Study About Personally Targeted Ads on Facebook
Nowadays, when it comes reaching the desired target online, advertisers are mostly dealing with heavily fragmented audiences. Communication is not addressed to the masses anymore, but is getting more individual and personalized. (Black, Asadorian & Dunnett, 2017; Southgate, 2017). That is why, the importance of personally targeted ads is
exponentially growing. Even though they are relevant and useful for the consumer (Ur, Leon, Cranor, Shay & Wang, 2012; Pandey, Bagherjeiran, Hatch, Ciccolo, Ratnaparkhi &
Zinkevich, 2011), there were instances where consumers would ask themselves why are they being displayed that particular ad. Sometimes this question is asked because it is an
extremely accurate representation the consumer’s interests, but other times it is because the advertised product or service is completely unrelated to the them, as it came out of the blue. Still, understanding the consumer is an important prerequisite for effective advertising (Stipp, 2016) and that is why delivering highly relevant information for them is crucial (Kallweit, Spreer & Toporowski, 2014).
Facebook, as a social media service, offers a product to individuals and businesses called Facebook Ads, which lets them build an audience they are determined to target
(Audience Targeting Options on Facebook, n.d.). One of these available advertising strategies is Facebook interest strategy, which is based on matching consumer’s interest activities with the product or service they relate the most. This matching process sometimes appears as unsuccessful, mostly because there is an incongruence between the product and the
negative attitudes towards the ads (Low & Lamb, 2000), and in turn, a lower click intention (Yun Yoo, 2011). Because of this, the interest advertising strategy needs to be improved.
In order to create personally targeted ads, a certain method is working behind the screens: OBA. OBA stands for Online Behavioral Advertising and it is responsible for
delivering ads which are formed on user’s data that was collected about their online behaviors (Boerman, Kruikemeier & Zuiderveen Borgesius, 2017). The individual part is important because this practice concentrates on targeting individuals that have specific preferences (Beales, 2010). This represents a great opportunity for advertisers because, when using online behavioral data, it is possible to make finer distinctions between individuals, which means delivering more relevant ads to them (ibid.). However, OBA is generally considered as
controversial because has raised a lot of privacy concerns (Moore, Moore, Shanahan & Mack, 2015; McDonald & Cranor, 2009). Since OBA represents the future of online advertising (Bosset, Frankel, Friedman & Satterfield, 2011), it is important to make it more acceptable both for the consumers and for the benefit of future businesses’ advertising ventures.
To put in simpler words, there are two research problems that need to be addressed. Firstly, because of inconsistent attitudes towards personally targeted ads, advertising strategies need to be improved. Facebook interest advertising strategy is at the centre of attention in this study, since it has shown to deliver incongruent ads to consumers. Secondly, the method responsible for delivering personally targeted ads, OBA, is generally being avoided because of privacy concerns. Because of its importance in online advertising, OBA needs to become more acceptable for the consumer.
Considering the research problems mentioned above, the first aim of this study is to address the problem of individual’s attitudes towards personally targeted ads on Facebook
and to observe whether those attitudes are more favorable when there is more congruency between the brand advertised and the individual’s personal brand interests. Therefore, the brand- self congruence theory is proposed as the theoretical framework for addressing the problem of incongruent ads that are delivered to the consumer on the basis of interest advertising strategy. The second aim of the study is see whether brand- self congruence theory will create a greater likelihood of the consumer accepting OBA. Hence, the research question:
To what extent does brand-self congruency theory explain attitudes towards personally targeted ads, click- through intention, and the likelihood of an individual
accepting online behavioral targeting method that is responsible for delivering such ads on social media?
A quantitative research method was used for this study: experimental survey design. After answering the research question, the study will be able to contribute to the advertiser’s practical knowledge, which is specifically aimed at improving their online targeting
solutions. This study results will deliver practical advices to business managers about what should they focus on when using Facebook interest advertising strategy. Moreover, the study will deliver recommendations to Facebook advertisers that can contribute in creating more acceptable audience targeting instruments for the consumer.
Conceptual framework
The following paragraphs incorporate definitions and explanations of all the concepts that are integrate part of the research question. Firstly, Facebook advertising strategies are listed together and defined. Secondly, the brand-self congruity theory is explained as why it can be considered as a solution for more favorable attitudes and higher
click- through intention. Thirdly, online behavioral advertising (OBA) is defined and
explained as why is it considered controversial. And finally, thanks to the model proposed by the theory of planned behavior, two mediation hypotheses are proposed.
Interest advertising strategy
Social media services are advancing the way advertisers are reaching the consumer, which is consequently revolutionizing online advertising overall (Gangadharbatla, 2008). Therefore, advertising strategies on social media have visibly progressed over the past few years (Lafferty, 2014). The new generation of ‘smart advertising’ is allowing the advertisers to quickly adapt and to target the consumers needs (Lukka & James, 2014).
Facebook, as a free social media networking site, mostly generates its revenue through advertising (Roberts, 2010). Businesses make use of Facebook’s features, such as Facebook Ads, to reach specific audiences online (ibid.). It consists of five different types of strategies: location, demographics, retargeting, lookalike, and interest targeting (Facebook ads 101: Why Am I Seeing This?, 2018). The location and demographic method are the least sophisticated and effective on Facebook, based on consumer’s location, age, gender,
educational level, etc. Retargeting is a practice that uses information from past online browsing on a company’s website to improve advertising on external websites, in this case Facebook (Lambrecht & Tucker, 2013). This is extremely important for consumers when they are already in an advanced purchase decision journey (ibid.). The lookalike strategy is based on targeting users that are similar to the company’s known customers.
Both retargeting and lookalike are great tools businesses can use for advertising and the two most powerful advertising features that are available on Facebook (Facebook Ads 101: Why Am I Seeing This?, 2018). However, when the method is targeting users based on
their interests, some questions arise. Sometimes the consumer’s interests are completely incongruent with the brand that is advertised (ibid.). Therefore, there is need to improve interest targeting so it can compete with the lookalike and retargeting method.
Brand- self congruity theory
One way to improve interest advertising strategy can be found in the brand- self congruence theory. Sirgy, Grewal, Mangleburg, Park, Chon, Claiborne, Johar, and Berkman (1997) introduced the personality approach which postulates that there is a strong emotional bond between the brand and the consumer. The bond is portrayed as an interactive
relationship where brand and consumer exchange symbolic benefits. The consumer finds the exchange beneficial when it contributes to the construction and reflection of their self- identity. The exchange consists in matching the brand image with the self- image. Sjödin and Törn (2006) depict high congruity as a match between the brand and self- image. The
congruence is categorized as high when these images are consistent between each other, when one image reflects the other, and when one image is the mirror image of the other (ibid.). When this matching is fulfilled, it begins to act as the driving force that affects consumer’s intention to engage with the brand (Sirgy et al., 1997).
Brand associations, such as marketing activities surrounding the brand, are being used by consumers in order to create images about the brand. Hence, when a brand that is advertised on social media, consumers link it to the images they have stored about the brand and that can further trigger matches or mismatches between the brand and their self-image (Sirgy et al., 1997). Mismatches, or incongruences, have been found to stimulate less favorable attitudes towards ads (Low & Lamb, 2000; Pechmann & Stewart, 1990; Czellar, 2003). Moreover, the study of Celebi (2015) has found that when the perceived self- brand
congruity of Facebook ads is higher, the attitudes towards them are more favorable. High congruence therefore should have a more positive impact on attitudes towards ads when compared to low congruence. Moreover, more relevant (Yun Yoo, 2011) and more finely tailored ads (Tene & Polonetsky, 2012) are found to improve click-through intentions (Wang & Sun, 2010). In Tucker’s study (2014), the results showed a higher click-through rate especially for Facebook ads delivered by interest advertising strategy when comparing to demographic strategy. Based on these findings, two hypotheses are stipulated:
H1a: Individuals are more likely to click on a personally targeted ad when the brand- self congruence is high or moderate when compared to low brand-self congruence.
H1b: Individuals are more likely to have favorable attitudes towards personally targeted ads when when brand- self congruence is high or moderate, which in turn will predict higher click-through intention when compared to low brand- self congruence.
Online behavioral advertising (OBA)
Personally targeted ads on Facebook and other social media are delivered to the users thanks to the advertising networks which work as intermediaries (Beales, 2010). The
H1a H1b H1b Brand- self congruence Click- through intention Ad attitudes
advertising networks connect publishers, such as social media platforms, with advertisers that want to reach a specific online audience. There are three types of strategies that ad networks use for matching advertisers with online users that look for content and services: contextual, vertical, and behavioral network strategies (ibid.). Contextual networks base their matching on keyword search, vertical networks group similar publishers and offer them to advertisers, while behavioral networks are delivering ads to users based on their online browsing
behaviors and afterwards categorize them based on their interests (Beales, 2010; McDonald & Cranor, 2009). The behavioral activities that are evaluated include consumer’s keyword searches, clicking ads, clicking on pages, liking pages on social media (Yan, Liu, Wang, Zhang, Jiang & Chen, 2009), which is all usually set in motion by installing ‘cookies’ (Smit, Van Noort & Voorveld, 2014).
Besides providing positive and relevant outcomes for advertisers and businesses (Tene & Polonetsky, 2012; Sanje & Senol, 2012; Beales, 2010; Yan et al., 2009), OBA has been found to have some positive effects on the consumers as well. Personally targeted ads that are delivered by OBA method were mostly useful, relevant, and beneficial as a source of information to the consumer (Bleier & Eisenbeiss, 2015; Schumann, Wangenheim & Groene, 2014; Ur et al., 2012; McDonald & Cranor, 2009). On the other hand, OBA was found to have particular set of negative effects for the consumer. The method has raised privacy concerns and suspicions about user’s data collection (Sanje & Senol, 2012; Beales, 2010; Ur et al., 2007). Because of it, OBA is being often avoided by the consumers (Summers, Smith & Reczek, 2016; Baek & Marimoto, 2012).
Previous studies have tried to decrease the avoidance of OBA by manipulating the level of personalization the ad has (Van Doorn & Hoekstra, 2013; Baek and Morimoto,
2012). The marketing strategy of personalization implicates the collection of the consumers’ online information done by OBA, which is in turn is used to display customized ads to them at the right time (Aguirre, Mahr, Grewal, de Ruyter & Wetzels, 2015; Smit, 2014). However, the personalization strategy takes into account the right timing for the consumers, which is more related to the advertising strategy of retargeting (Lambrecht & Tucker, 2013).
Consumer’s interests can be regarded as more persistent over time and do not change as often as their needs (Aguirre et al., 2015). This is why this study, instead of using the level of personalization, uses the theory of brand- self congruency to explain consumer’s process of matching their interests with the brand’s image. Therefore, brand- self congruence should create as well more acceptance towards OBA:
H2a: Individuals are more likely to accept OBA when the brand- self congruence is high or moderate when compared to low brand-self congruence.
OBA seems to first trigger affective responses, which consequently affect behavior (Boearman, Kruikemeier & Zuiderveen Borgesius, 2017). The main behavior in question is the likelihood of accepting OBA as the practice of delivering personally targeted ads on social media. Higher perceptions of relevance, utility, and level of personalization have been found to generate more favorable outcomes regarding ad attitudes and OBA (Baek &
Morimoto, 2012; Schumman et al., 2014; Cho, 2004) The theory of planned behavior
postulates that a certain behavioral intention is partly predicted by a the attitude towards that specific behavior (Ajzen, 1991). The behavioral intention in question is defined in this study as OBA acceptance. Normally, the stronger the attitude towards the behavior, the stronger is the intention to perform that behavior (ibid.). Therefore, the next and last hypothesis is stipulated:
H2b: Individuals are more likely to have favorable attitudes towards personally targeted ads when when brand- self congruence is high or moderate, which in turn will predict higher OBA acceptance when compared to low brand- self congruence.
Methods Sample and procedure.
A quantitative research method was required in order to contribute answering the previously established research question. More specifically, a between subject experimental design was used with four different conditions. The respondents were either presented with high, middle or low brand congruity as well as one control group. The target population were males and females with an active Facebook account, age 18 or older, and with a reasonable knowledge of English. The target sample size was 120 because of the minimum requirement of 30 respondents per group supported by the Central Limit Theorem (Bryman, 2016). To reach the target population, a non-random and convenient sampling method was used. A snowball survey was executed by sending the link among Facebook users via message, on student groups, and survey exchange groups. It was an individual anonymous self-report and the participants could drop out at any time. There was an informed consent that stated the
Ad attitudes
H2b H2b
Brand- self
congruence OBA acceptance
participant's information will be safeguarded. The data collection was done from 21-12-18 until 05-01-19.
From the initial number of responses (N= 228), respondents who did not meet the criterion of having an active Facebook account were taken out (N= 8). Moreover, 38
responses were not completed. In the end, the sample consisted of 182 cases, where 60% of them were female. The majority of the respondents were from Croatia (63%), while the others were from Netherlands (6%), United Kingdom (4%), etc. The majority held a Bachelor’s degree (38%), more than half of them had a full-time job (55%), and 23% were still studying. The mean age was 31.51 and it ranged from 18 to 62 years old.
Measures
Independent variable
Brand congruency. The independent variable was operationalized by randomly assigning respondents to four conditions: high, medium, low congruency, and neutral
condition. The congruency was assessed between the brand that was advertised and the brand that was liked on Facebook by the consumer. The manipulated variable featured an ad of the new product launch from the sportswear and footwear brand PUMA. On the top right corner of the ad, a drop down menu was opened and it showed the respondent why that certain ad was displayed to him or her. Usually, Facebook displays more advertising strategies that were used, but for the purpose of this study, only interest based targeting was displayed. This means that only pages of brands the user liked on Facebook were shown. The brand liked by the user that has a high congruence with PUMA was Adidas, since they are both sportswear and footwear retailers. The fast fashion retailer H&M was selected as a medium congruence with PUMA, because they are both clothing retailers but different styles. The restaurant chain
Domino’s Pizza was selected for a low congruence because these brands are from different industries, fashion and food. The neutral condition did not show any brands liked on Facebook, just the plain ad without the drop down menu.
In order to check whether the conditions significantly differed between them, a single item question was presented to each respondent: “To what extent do find the brand PUMA and the brand Adidas/ H&M/ Domino’s pizza similar?” with a response scale from (1) Extremely dissimilar to (10) Extremely similar.
A manipulation check question was provided by the end of the survey as a single item question: “Which brand was displayed as liked as an explanation on why the PUMA ad was presented to you?”.
Dependent variables
OBA acceptance. To measure to what extent the respondents were willing to let the platform (Facebook) evaluate their online activities, a 7-point Likert scale from 1 (“Strongly disagree”) to 7 (“Strongly agree”) with 3 items was used, such as: „I would probably allow Facebook to evaluate my surfing behavior.“, „It is likely that I would consent to an analysis of my surfing behavior“, „I would be willing to agree to an evaluation of my surfing
behavior“. The scale items were modified from the study of Schumman et al. (2014) to fit the context of Facebook.
Click-through intention. The intention to click on an online ad was measured with one item from 1 (Strongly disagree) to 7 (Strongly agree): “I would like to click on the advertisement to get further information.” The item was taken from the studies of Aguirre et al. (2015) and Yun Yoo (2011).
Mediating variable
Ad attitudes. An earlier and highly reliable developed scale from Muehling (1987), consisted of only two items that measured attitudes towards ads. However, besides general positive or negative attitudes towards them, it was important to take other dimensions into account because of the personal nature of the ad. Attitudes towards personally targeted ads therefore were measured with 4 7-point semantic differential scales: bad – good; harmful – beneficial; not useful – useful; and uninformative – informative (Ducoffe, 1996).
Control variables
OBA knowledge. The knowledge about online behavioral targeting was measured on a subjective dimension (Ham & Nelson, 2016) and it consisted of 6 items on a 7-point Likert scale scale from 1 (“Strongly disagree”) to 7 (“Strongly agree”). Subjective knowledge of OBA of an individual is a self-assessed perception about how this kind of persuasion tactic from advertisers works (e.g., “I know how OBA displays personalized ads to me”). It has been found to influence both the individual itself and a third person positively and negatively, unlike the objective knowledge (Ham & Nelson, 2016).
Familiarity. Respondents that are more familiar with personally targeted ads on Facebook might have different effects on the dependent or mediating variables. Familiarity was operationalized with a single item on a 7 point Likert scale from 1 (“Not familiar at all”) to 7 (Extremely familiar): “How familiar are you with personally targeted ads on Facebook?” Product involvement. The degree of involvement in a certain product can have an effect on the attitudes towards the ad. Respondents that are more involved with the product in
question might have greater effect when evaluating the ad (Petty, Cacioppo & Schumann, 1983).
Gender. The PUMA ad displayed two pairs of shoes that were both for women and men. In addition, Adidas and H&M both have male and female clothing. However, there is a possibility that gender will have an effect on the response to Facebook ads (Rehman, Ilyas, Nawaz & Hyder, 2014).
Age. Smit et al. (2014) have found that older individuals had less knowledge about OBA and more negative attitudes about personally targeted ads when compared to younger ones. On the other hand, Turow, King, Hoofnagle, Bleakley and Hennessy (2009) found that younger people are more likely to accept personally targeted ads than older ones.
Results
Before the analyses, assumptions for normality were checked for all the variables. There were no outstanding outliers that could compromise the test results. A Saphiro-Wilk test of normality for small sample sizes (< 50 samples) was used. The results were almost entirely significant across all conditions, which means that variables were mostly not normally distributed, except for the attitudes towards ads. Besides the Pearson’s correlation table, Spearman’s rho as a nonparametric correlation test was run to ascertain the normality of distributions which showed that the significance of correlations did not deviate. However, the results from the upcoming analyses are interpreted with caution. Below, Table 1 shows means, standard deviations, correlations, and Cronbach’s alphas of the variables. Two one-way ANOVAs were run in order to test the direct effect hypotheses. To test the indirect effects, on the other hand, two mediation analyses with PROCESS using model number 4 were done.
Chi-square tests and a oneway ANOVA test for randomization check ensured that there was an equal distribution of respondents across conditions based on gender, χ2 (3, N= 182) = 7.38, p = .061, country, χ2 (69, N= 182) = 68.46, p = .496, occupation, χ2 (15, N= 182)= 9.87, p = .828, education, χ2 (12, N=182)= 8.47, p = .748, and age, F (3, 181)= 0.042, p = .989.
The Chi- square test showed that there was a significant association between the four
conditions and the recall of the brand that was displayed as liked on the Facebook ad , χ2 (12, N= 182)= 203.67, p < .001. To about half of the respondents across all the conditions the manipulation was successful, high (53%), medium (63%), low congruity (48%), and control (33%).
The oneway ANOVA test, F (2, 181)= 690.22, p < .000, established that the respondents evaluated the categories of similarity between brand liked to the brand advertised (high, medium, low) as significantly different from each other.
Table 1.
Correlations and descriptive statistics.
Notes: N = 182. Values on the diagonal in bold represent reliabilities (α). * Significant values at p < .05. ** Significant values at p < .01. ª high congruence= 1, medium congruence= 2, low congruence= 3, neutral= 4. ᵇ male= 1, female= 2.
Even though there were no significant correlations found for the two direct effect hypotheses (H1a, H2a), two one-way ANOVAs were still performed to see the differences between the group means. The correlations between the variables for the indirect effect were also non significant, except for the relationship between the mediator (ad attitudes) and the two dependent variables (click-through intention and intention to accept OBA). The mediation effect analyses were performed as well.
The first hypothesis proposed that when there is high or moderate congruity between the brand liked and the brand advertised, the click-through intention will be higher when comparing to low congruity. In order to test the hypothesis, oneway ANOVA test was run, with click-through intention as dependent variable, brand congruity as independent and familiarity, product involvement, age, and gender as control variables.
Variable M (SD) 1 2 3 4 5 6 7 8 9 1. Brand congruencyª 2.48 (1.11) -2. Ad attitudes 4.05 (1.31) -.01 .81 3. Click-through intention 3.30 (1.66) -.07 .39** -4. OBA acceptance 2.84 (1.47) .01 .35** .39** .93 5. OBA knowledge 4.43 (1.37) .01 .13 .17* .15* .90 6. Product involvement 5.24 (2.39) -.08 .27** .22** .29** .24** .82 7. Familiarity 2.90 (1.25) -.02 .19** .17 .02 .59** .14 -8. Age 31.51 (10.62) .01 .04 .05 .05 -.14 -.14 -9. Genderᵇ 1.60 (0.49) .15* .09 .11 .03 .03 .03 .04
-Firstly, Levene’s test was run to see whether the respondents varied equally across the conditions. It resulted insignificant, F (3, 178) = .69, p = .554, which means equal variances were established. The ANOVA results showed that there was no statistically significant difference between the four conditions, F (3, 178) = 0.63, p = .599, η2 = .002. The means across the four conditions were almost the same. However, the condition with high brand congruity had a slightly higher intention to click on the ad (M = 3.43, SD =1.61) when compared to medium (M = 3.30, SD = 1.74), low congruity (M = 3.42, SD = 1.71), and control condition (M = 3.00, SD = 1.61). Furthermore, the control variable product
involvement had an extremely small significant effect on the click-through intention, F (1, 181) = 8.37, p = .004, η2 = .008.
Based on these results, the first hypothesis can be rejected and the assumption that there is no difference of the effect of brand congruency on individual’s click-through intention in the population can be accepted.
The second hypothesis (H1b) proposed a mediation effect of attitudes towards ads between the independent variable brand congruency and the dependent variable click-through intention.
A regression model number 4 was used in PROCESS with the dependent variable click-through intention, mediator variable ad attitudes, independent variable brand congruency, and covariates familiarity, product involvement, age, and gender.
The first model with ad attitudes as outcome variable was significant, F (7, 174) = 3.45, p = . 002. The variables explained only 12% of the variance in the outcome variable (R² = .12). Medium congruity b =.14, t = 0.52, p = .601, CI 95% [-0.39, 0.67], low congruity b = -.03, t = -0.13, p = .895, CI 95% [-0.55, 0.48], and neutral condition b = 0.09, t = 0.35, p = .727, CI
95% [-0.43, 0.62], when compared to high congruity did not have a significant direct effect on ad attitudes. However, familiarity with personally targeted ads, b = 0.19, t = 2.44, p = . 016, CI 95% [0.04, 0.35], and product involvement, b = 0.14, t = 3.63, p < .001, CI 95% [0.07, 0.22], had a significant direct effect on attitudes towards ads.
The model with the outcome variable click-through intention was significant, F (8, 173) = 4.93, p < .001, and the variables explained only 19% of the variance in the dependent variable (R² = .19). As the ANOVA test already showed, brand congruity did not predict click-through intention. Attitudes towards ads, however, significantly predicted it, b = 0.43, t = 4.63, p < .001, CI 95% [0.25, 0.61]. The more favorable the attitudes towards a personally targeted ad are, the more it is likely that the individual will click on it.
There is a non significant partially standardized indirect effect of brand-self congruity on the click-through intention, medium congruity b = .04, CI 95% [-0.09, 0.19], low congruity b = -. 01, CI 95% [-0.15, 0.13], and neutral condition b = 0.02, CI 95% [-0.13, 0.17], when
compared to high congruity.
Given the results, the second hypothesis (H1b) can be rejected and the fact that there is no mediating effect of ad attitudes between brand congruity on click-through intention in the population can be assumed.
Notes: N = 182. ** Significant correlations at p < .01 Ad attitudes 0.39** Brand- self congruence Click- through intention - .07 - .01
The third hypothesis (H2a) stipulated that individuals will have a higher likelihood of accepting OBA if brand-self congruity is high or moderate when compared to low
congruity. An additional covariate was taken into consideration: OBA knowledge. Oneway ANOVA results show that there was no significant and almost non existent effect of the levels of brand-self congruity when looking at the dependent variable OBA acceptance, F (3, 181) = 0.08, p = .972, η² = .001. The covariate product involvement had a significant positive
moderate effect on the dependent variable, F (1, 181) = 14.12, p < .001, η² = .07. The more the individual is involved in the product, the more likely they will accept OBA. The Levene’s test was not significant, F (3,178) = 1.46, p = .228.
Given the results, the third hypothesis can be rejected. It can be assumed that there is no significant effect in the population of brand-self congruity on the individual’s likelihood to accept OBA.
The fourth and last hypothesis (H2b) proposed that there is a mediation effect of attitude towards ads between the independent variable brand congruity and the dependent variable OBA acceptance. The effect on attitudes towards ads was the same as from the second hypothesis (H1b), so no significant direct effect of brand-self congruity.
The second model representing the effects on OBA acceptance was significant, F (9, 172) = 4.43, p < .001, and the variables explained 19% of the variance in the dependent variable (R² = .19). Attitudes towards ads had a large significant positive main effect on the outcome variable, b = 0.34, t = 4.14, p < .001, CI 95% [0.18, 0.50]. The more positively the individual evaluates the ad, the more it is likely that he or she will accept OBA. Individual’s product involvement still had a medium significant positive main effect on OBA acceptance, b = 0.12,
The partially standardized indirect effect of brand congruity on OBA acceptance through attitudes towards ads was non significant: medium congruity b = .03, CI 95% [-0.09, 0.17], low congruity b = -.01, CI 95% [-0.15, 0.11], and neutral condition b = 0.02, CI 95% [-0.12, 0.15], when compared to high congruity.
Therefore, the fourth hypothesis (H2b) can be rejected. The assumption that there is no significant indirect effect of brand congruity on OBA acceptance in the population can be accepted.
Notes: N = 182. ** Significant correlations at p < .01
Conclusion
The research problem stipulated in the introduction questioned to what extent does brand-self congruity theory explain attitudes towards personally targeted ads, click-through intention, and the likelihood of an individual accepting online behavioral targeting method that is responsible for delivering such ads on social media.
Ads that appear on the Facebook newsfeed are denominated in this study as personally targeted ads. There are several advertising strategies on Facebook that deliver these ads, however the highlight of this study is the interest targeting strategy, which has not been found as successful as the retargeting and lookalike strategy (Facebook Ads 101: Why
Ad attitudes
- .01. 0.35**
Brand- self
Am I Seeing This?, 2018). One of the reasons for the interest strategy to stagnate can be because of this ads displaying incongruent pieces of information to the consumer. Because of the incongruence, they cannot reflect their self to the brand that is advertised and therefore evaluate the ad less favorably (Low & Lamb, 2000; McDonald & Cranor, 2009). Hence, the theory of brand-self congruence was applied as the theoretical framework for addressing the problem of negative attitudes towards personally targeted ads on Facebook and lower click-through intention.
Besides the problem addressed above, there was another issue acknowledged in this study. Consumers are generally found to have inconsistent attitudes towards personally targeted ads (e.g. Van Doorn & Hoekstra, 2013). These ads are delivered to the consumer thanks to online behavioral advertising (OBA) method, which is done by collecting data about their online activities (e.g. McDonald & Cranor, 2009). Smit et al. (2014) and Beales (2010) have found that a lot of consumers are avoiding OBA because of privacy concerns. Because of that, consumers have inconsistent attitudes towards the displayed ads (e.g. Smit et a.l, 2014). Avoidance and inconsistent attitudes towards ads is something that advertisers should prevent from occurring. This is why, in order to generate more favorable attitudes and in turn, higher acceptance of OBA, brand-self congruity was again proposed as the theoretical framework for this issue.
The main results have shown that the experimental research method design did not contribute answering the proposition of brand-self congruence theory explaining consumer’s attitudes towards personally targeted ads, their intention to click on the ad, and the likelihood of letting Facebook analyze their browsing behaviors. Nonetheless, the results showed that even though people notice when ads are more congruent with their interests, that does not
seem to affect their attitudes they have towards them. But, whenever their attitudes were favorable, it was more likely for them to click on the ad as well as accept OBA method of targeting.
This study has several contributions. Theoretical contributions consist in developing a richer understanding of the brand-self congruence concept, which is based on the
congruence between consumer’s brand interests and the brands advertised on social media. Furthermore, regarding practical implications, the study proposes better targeting solutions to advertisers on Facebook. The recommendations include displaying ads that are moderately congruent with the consumer. Advertisers should also focus on other successful Facebook advertising strategies, such as lookalike and retargeting strategy, until further research has investigated other possibilities on how to improve interest targeting. Lastly, in order to benefit both the consumer and business managers, advertisers should deliver ads to consumers that are already highly involved in certain product types similar to the brand that is advertised.
Discussion
The theory of self-brand congruity explains that consumers always try to match their attributes to the brand’s attributes when evaluating a product or service (Sirgy et al., 1997; Sjödin & Törn, 2006). In this study, the explanation of consumers’ evaluation and their intention to click on personally targeted ads on Facebook based on matching their image to the brand’s image has been shown to have some downfalls. Consumers might not experience the process of matching their own interests in brands to brands that are being advertised on social media. This is not in line with the self-identity theory which proposes that people are motivated by their need to maintain their constant behaviors (Sirgy, 1997). Moreover,
the brand associations, which is congruent with the study of McDonald and Cranor (2009) where participants in interviews mostly said that online ads are simply ‘a fact of life’ and had indifferent feelings towards them.
Another point regarding the theory of self congruence is that based on the results of this study, it does not seem to provide any significant explanations of whether individuals would accept OBA if the ad matched their interests. A possible reason behind it might be that consumers generally do not accept OBA, and therefore it does not matter whether it matches their interests or not. A similar result was found in the study of Schumman et al. (2014). Their findings suggest that consumers do not accept OBA because of the relevant ads it delivers, but they rather accept it because of the fact that ads provide free online content for them.
The theory of planned behavior has proved that the more favorable attitudes towards a certain behavior the individual has, the more he or she will have the intention to perform it (Ajzen, 1991), which is in this case, is both the intention to click on the personally targeted ad as well as the intention of letting Facebook evaluate consumers’ online behaviors. These results show that individuals are willing to click on the ad and accept the OBA method when they find that the ad is overall informative, good, useful, and beneficial.
There are several implications that can be delivered based on these results. Firstly, the concept of congruence between consumer’s interest and the brand advertised should be considered as an additional ‘distant family member’ of concepts that will contribute to a richer understanding of the brand-self congruence theory (Sjödin & Törn, 2006). This can be a relevant addition to the literature of advertising on social media, mainly because of the continuous increase in quality of behavioral targeting methods (Bosset et al., 2011).
shown that the more personalized ad is, the more likely is that they will accept it (Li &
Huang, 2016; Baek & Morimoto, 2013), individuals still hold negative attitudes towards them (e.g. Ur et al., 2007). On the other hand, when the ad is incongruent with the consumer’s self, it is still negatively evaluated (Sjödin & Törn, 2006). Advertisers should be careful in these situations because both the extremes lead to negative outcomes. Therefore, in the light of brand-self congruence theory, the best strategy for advertisers is to ‘play it safe’ and show moderately congruent ads with consumer’s interests.
Thirdly, advertisers should also seek to favor other types of audience targeting on Facebook that are more successful, such as retargeting and lookalike strategy, until further research has investigated ways to improve interest targeting strategies.
Finally, individuals that are highly involved with a certain product should be displayed ads with similar products. That is in line with the self identity theory where the consumer builds up on their ideal self by reinforcing certain behaviors (Sirgy, 1997).
There are some limitations to consider in this study, as well as several interesting suggestions for future research. Firstly, because of the experimental nature of the design, the controlled environment decreases the external validity of the responses. Moreover, the internal validity could have been compromised since only the respondents’ product category involvement was controlled for. Some people hold strong attitudes towards certain brands as well, which could have affected the responses. Another thing is, the sampling method used was convenient snowball method, which is not random.
Secondly, the intention to click on an ad does not necessarily mean that the consumer will purchase the product (Yan et al., 2009). Further research should take into consideration the purchase intentions because that is what consequently leads to business’ revenues.
Thirdly, the study investigated personally targeted ads on Facebook. Interesest based targeting should be more thoroughly investigated on other platforms, such as Instagram, where consumers can be more brand conscious and therefore attentive to interest matching. Further research should consider arguments other than relevance or congruence between user image and brand image in explaining attitudes towards ads, the click-through intention and acceptance of OBA. As explained by the theory on planned behavior (Ajzen, 1991), ulterior motives and beliefs seem to be guiding consumers to hold favorable attitudes towards ads.
References
Aguirre, E., Mahr, D., Grewal, D., de Ruyter, K., & Wetzels, M. (2015). Unraveling the personalization paradox: The effect of information collection and trust-building strategies on online advertisement effectiveness. Journal of Retailing, 91(1), 34-49.
Ajzen, I. (1991). The theory of planned behavior. Organizational behavior and human
decision processes, 50(2), 179-211.
Audience Targeting Options on Facebook (n.d.). Ads audiences. Retrieved January 12, 2019 from: https://www.facebook.com/business/help/717368264947302?
helpref=page_content
Baek, T. H., & Morimoto, M. (2012). Stay away from me. Journal of advertising, 41(1), 59-76.
Beales, H. (2010). The value of behavioral targeting. Network Advertising Initiative, 1.
Black, A., Asadorian, D., & Dunnett, H. (2017). 8 key truths about generation Z. Research
World, 2017(67), 12-14.
Bleier, A., & Eisenbeiss, M. (2015). The importance of trust for personalized online advertising. Journal of Retailing, 91(3), 390-409.
Boerman, S. C., Kruikemeier, S., & Zuiderveen Borgesius, F. J. (2017). Online behavioral advertising: A literature review and research agenda. Journal of Advertising, 46(3), 363-376.
Bosset, E. C., Frankel, S. J., Friedman, M. B., & Satterfield, S. P. (2011). Private actions challenging online data collection practices are increasing: Assessing the legal landscape. INTELL. PROP. & TECH. LJ, 23, 3-4.
Bryman, A. (2016). Social research methods. Oxford university press.
Celebi, S. I. (2015). How do motives affect attitudes and behaviors toward internet
advertising and Facebook advertising?. Computers in Human Behavior, 51, 312-324. Cho, C. H. (2004). Why do people avoid advertising on the internet?. Journal of advertising,
33(4), 89-97.
Czellar, S. (2003). Consumer attitude toward brand extensions: an integrative model and research propositions. International Journal of Research in Marketing, 20(1): 97–115. Ducoffe, R. H. (1996). Advertising value and advertising on the web. Journal of advertising
research, 21
Facebook Ads 101: Why Am I Seeing This? (2018). Kook. Retrieved January 12, from: https://www.kook.com.au/n/facebook-ads-101%3A-why-am-i-seeing-this/
Gangadharbatla, H. (2008). Facebook me: collective self-esteem, need to belong, and internet self-efficacy as predictors of the iGeneration’s attitudes toward social networking sites. Journal of Interacting Advertising, 8(2).
Ham, C. D., & Nelson, M. R. (2016). The role of persuasion knowledge, assessment of benefit and harm, and third-person perception in coping with online behavioral advertising. Computers in Human Behavior, 62, 689-702.
Kallweit, K., Spreer, P., & Toporowski, W. (2014). Why do customers use self-service information technologies in retail? The mediating effect of perceived service quality.
Journal of retailing and consumer services, 21(3), 268-276.
Lafferty J. (2014). HOW TO: See Why You’re Seeing A Facebook Ad. Retrieved 13 January, 2019, from: https://www.adweek.com/digital/see-youre-seeing-facebook-ad/
Lambrecht, A., & Tucker, C. (2013). When does retargeting work? Information specificity in online advertising. Journal of Marketing Research, 50(5), 561-576.
Li, W., & Huang, Z. (2016). The research of influence factors of online behavioral advertising avoidance. American Journal of Industrial and Business Management, 6(09), 947.
Low, G. S. & Lamb, C. W. (2000). The measurement and dimensionality of brand associations. Journal of Product and Brand Management, 9(6): 350– 368.
Lukka, V., & James, P. T. (2014). Attitudes toward Facebook advertising. Journal of
management and Marketing Research, 14(2), p1-26.
McDonald, A. M., & Cranor, L. F. (2009). An empirical study of how people perceive online behavioral advertising (CMU-CyLab-09-015).
Moore, R.S., Moore, M.L., Shanahan, K.J. & Mack, B. (2015), Creepy Marketing: Three Dimensions of Perceived Excessive Online Privacy Violation. Marketing
Management, 25 (1), 42–53.
Muehling, D. D. (1987). An investigation of factors underlying attitude-toward-advertising-in-general. Journal of Advertising, 16(1), 32-40.
Pandey, S., Aly, M., Bagherjeiran, A., Hatch, A., Ciccolo, P., Ratnaparkhi, A., & Zinkevich, M. (2011). Learning to target: what works for behavioral targeting. In Proceedings of
the 20th ACM international conference on Information and knowledge management
(pp. 1805-1814). ACM.
Pechmann, C., Stewart, D. W. (1990). The effects of comparative advertising on affection, memory, and purchase intentions. Journal of Consumer Research, 180–191.
Petty, R. E., Cacioppo, J. T., & Schumann, D. (1983). Central and peripheral routes to advertising effectiveness: The moderating role of involvement. Journal of Consumer
Research, 10(2), 135-146.
Rehman, F. U., Ilyas, M., Nawaz, T., & Hyder, S. (2014). How Facebook advertising affects buying behavior of young consumers: The moderating role of gender. Academic
Research International, 5(4), 395-404.
Roberts, K. K. (2010). Privacy and perceptions: How Facebook advertising affects its users.
The Elon Journal of Undergraduate Research in Communications, 1(1), 24-34.
Sanje, G., & Senol, I. (2012). The importance of online behavioral advertising for online retailers. International Journal of Business and Social Science, 3(18).
Schumann, J. H., von Wangenheim, F., & Groene, N. (2018). Targeted online advertising: Using reciprocity appeals to increase acceptance among users of free web services.
Journal of Marketing. 78(1), 59-75.
Sirgy, M. J., Grewal, D., Mangleburg, T. F., Park, J. O., Chon, K. S., Claiborne, C. B., Johar, J.S. & Berkman, H. (1997). Assessing the predictive validity of two methods of measuring self-image congruence. Journal of the Academy of Marketing Science,
25(3), 229.
Sjödin, H., & Törn, F. (2006). When communication challenges brand associations: a framework for understanding consumer responses to brand image incongruity.
Smit, E. G., Van Noort, G., & Voorveld, H. A. (2014). Understanding online behavioural advertising: User knowledge, privacy concerns and online coping behaviour in Europe. Computers in Human Behavior, 32, 15-22.
Southgate, D. (2017). The emergence of generation Z and its impact in advertising: Long-term implications for media planning and creative development. Journal of
Advertising Research, 57(2), 227-235.
Stipp, H. (2016). What 80 years of study means for the future of advertising research. Journal
of Advertising Research, 56(3), 231-234.
Summers, C. A., Smith, R. W., & Reczek, R. W. (2016). An audience of one: Behaviorally targeted ads as implied social labels. Journal of Consumer Research, 43(1), 156-178.
Tene, O., & Polenetsky, J. (2012). To track or do not track: advancing transparency and individual control in online behavioral advertising. Minn. JL Sci. & Tech., 13, 281.
The Beginner’s Guide to Facebook Advertising (n.d.). Retrieved January 14, from: https:// adespresso.com/guides/facebook-ads-beginner/demographic-targeting/
Tucker, C. E. (2014). Social networks, personalized advertising, and privacy controls.
Journal of Marketing Research, 51(5), 546-562.
Turow, J., King, J., Hoofnagle, C. J., Bleakley, A., & Hennessy, M. (2009). Americans reject tailored advertising and three activities that enable it.
Ur, B., Leon, P. G., Cranor, L. F., Shay, R., & Wang, Y. (2012). Smart, useful, scary, creepy: perceptions of online behavioral advertising. In proceedings of the eighth symposium
Van Doorn, J., & Hoekstra, J. C. (2013). Customization of online advertising: The role of intrusiveness. Marketing Letters, 24(4), 339-351.
Wang, Y., & Sun, S. (2010). Assessing beliefs, attitudes, and behavioral responses toward online advertising in three countries. International Business Review, 19(4), 333-344.
Yan, J., Liu, N., Wang, G., Zhang, W., Jiang, Y., & Chen, Z. (2009, April). How much can behavioral targeting help online advertising?. In Proceedings of the 18th international
conference on World wide web (pp. 261-270). ACM.
Yun Yoo, C. (2011). Interplay of message framing, keyword insertion and levels of product involvement in click-through of keyword search ads. International Journal of
