Pre-roll advertising in YouTube videos
A study on the effects of a skip-button in pre-roll advertising in YouTube videos
Burcu Akcay
Student ID 10054219 University of Amsterdam
Master’s Thesis, Persuasive Communication Supervisor mw. A. Kranzbühler
Date of completion 2 February 2018
1 1. Introduction
2. Theoretical background
2.1 Un-skippable vs skippable ads 2.2 Advertising effectiveness 2.3 Perceived Control
2.4 Intrusiveness on ad effectiveness 3. Method
3.1Procedure & Design
3.2 Instruments & Measurements 3.3Participants 3.4Manipulation check: 4. Results 4.1Manipulation check 2.2 Correlations 4.3 T-test 4.4Regression 5. Conclusion & Discussion
References
Appendix A: Pre-roll ad
Appendix B: YouTube video Pandas Appendix C: Questionnaire
2
ABSTRACT
Pre-roll advertising on YouTube is the most dominant format of online video advertising. Advertisers can pick two types of pre-roll ads. A type where the user can skip the ad after five seconds or the other type where there is no option to skip the ad. In this case the user must watch the ad prior to the YouTube video. This study investigates the effectiveness of the option to skip an ad in pre-roll advertising. This study contains two conditions - a condition where users can skip the pre-roll ad and a conditions where the participants can not skip the ad. No effects of the option to skip the ad were found on the ad effectiveness. However, the results show that
skippable ads are being perceived as less intrusive than ads that provide no option to skip the ad. Another interesting finding is that intrusiveness influences the ad attitude of the users.
3 1.INTRODUCTION
The internet is the fastest growing advertising segment in the world (Gambaro & Puglisi, 2012). Within that category, especially online video advertising is a fast growing format of online advertisement. Research has shown that when it comes to recall and likeability, the impact of online advertising in video format can be greater than traditional TV ads (IAB, 2013). This provides interesting opportunities for marketeers. Online video ads can be placed before (pre-roll ad), during (mid-roll ad) or after (post-roll ad) an online video is played (Hegner, Kusse & Pruyn, 2016). YouTube is the most important name when discussing online video. YouTube is an online video sharing platform, which was founded as an independent website in 2005 and was purchased by Google in 2006 (Airoldi, Beraldo & Gandini, 2016). Founded in 2005, it is already the third most visited website worldwide, with one billion monthly users (Airoldi, Beraldo & Gandini, 2016). YouTube enables users to post, view, share and comment on videos on the site. Furthermore, the young generation between 18 and 34 watch more YouTube videos than any cable TV channel. This generation makes up two-third of the YouTube users, they are the most frequent users of YouTube (Alhabash, Baek, Cunningham & Hagerstrom, 2015; Perrin, 2015). All this makes YouTube an interesting platform for marketeers to advertise on. Pre-roll ads are with 60% the most dominant form of online video ad formats (Goodrich, Schiller & Galleta, 2015). Yet the impact of pre-rolled online video advertisements on YouTube is not yet fully understood. One of the challenges is that in many cases visitors do not watch the video
advertisement, if they get the chance to skip it (Dehghani, Niaki & Sali, 2016). But not providing an option to skip the ad (skip button), might lead to negative attitudes towards the ad (McCoy, Everard, Polak & Galletta, 2008). In this paper we will therefore focus on the potential effects that the presence of an option to skip the ad in pre-roll advertising on YouTube can have on the advertising effectiveness.
YouTube viewers are in most cases involuntarily exposed to pre-roll ads. One problem with pre-roll advertising in YouTube videos is that it can result in ad avoidance and irritation (Cho & Cheon, 2014; Hegner, Kusse & Pruyn, 2016). The study of Dehghani, (2016) states that only 5% of all the respondents in their study indicated that they sometimes watch the YouTube commercials. The majority of the respondents (88%) reported that they always skipped the YouTube commercials (when given the opportunity). Furthermore, the study of Chatterjee
4 (2008) states that intrusive ad formats that demand immediate response are more likely to be avoided by physically closing them. When searching for a video on YouTube, people are more goal oriented compared to when they are watching TV. Video ads on YouTube are therefore being perceived as more intrusive than ads on Tv. (Li, Edwards & Lee, 2002). This may provide an explanation for ad avoidance and negative attitudes towards the ad. Despite the drawbacks, pre-roll ads on YouTube are still most frequently used by marketeers. It is therefore important to have a better understanding of its effectiveness. In this paper we make an attempt to find out more about the effectiveness of pre-roll advertisement in YouTube videos. As mentioned before, the problem is that when visitors have the chance to skip it, they will in many cases not watch the pre-roll advertisement in its entirety. The pre-roll ads on YouTube can be provided with or without the ad skipping option. With the ad skipping option, the ad appears and the user is obliged to watch it for five seconds, before the visitor is provided with the option to skip the ad. Without the ad skipping option, the visitor is obliged to watch the whole pre-roll video
advertising (instead of just 5 seconds), before the intended video can be watched.
With this knowledge it might seem wise for marketeers to choose to advertise without an ad-skipping option, which would increase the number of users who would actually see the whole pre-roll ad.
However, some scholars state that not providing an option to skip an ad or control an ad, might be experienced as more intrusive and therefore will negatively influence the visitor’s perception of the brand (Hegner, Kusse & Pruyn, 2016; McCoy, Everard, Polak & Galletta, 2008). But on the other hand, if users just see five seconds of the advertisement, will the advertisement be effective? Therefore, we will more specifically look if the ad attitude, brand attitude and purchase intention of users differs for pre-roll advertisements that provide the option to skip the ad, and pre-roll ads that do not provide the option to skip the ad. To investigate this, the following research question is proposed:
RQ: “How does the option to skip an ad in pre-roll online video advertising influence its effectiveness?”
5
2. THEORETICAL BACKGROUND
2.1 Un-skippable vs skippable ads
As mentioned in the introduction above, YouTube provides different options for advertising. One specific option of advertising is pre-roll advertising. These are online video ads placed before a YouTube video. These pre-roll ads automatically start when a user clicks on the video they want to watch (Goodrich, Schiller & Galletta, 2015). Pre-roll ads come in two different formats: the un-skippable and the skippable format. If a un-skippable ad appears before a video, the user is obliged to watch the entire ad before he/she can watch the video they intend to watch. Skippable pre-roll ads provide the option to skip the pre-roll ad after viewing it for five seconds (“Non-skippable in-stream ads”, n.d). Although both formats are quite similar to each other, there seems to be a difference in effects. YouTube statistics show that un-skippable pre-roll ad formats on YouTube have a higher abandonment rate than other ad formats on YouTube (“ Non-skippable in-stream ads”, n.d). Which is in line with the psychological reactance theory by Brehm (1966). This theory explains how people react when they feel like their freedom to engage in a behavior is being threatened or restricted. The reactance theory states that when a freedom is being restricted or threatened, it becomes more attractive to the person losing that freedom and it will then seek for ways to restore freedom (Edwards, Lee & Li, 2002; McCoy et al., 2008; Morimoto & Chang, 2006). Thus, in the case of pre-roll advertising, the freedom is already being threatened by showing an ad (that was unasked for) prior to a YouTube video. According to the psychological reactance theory one will seek for ways to restore his/her freedom, for example by abandoning the video. This is even more the case for the un-skippable pre-roll advertising, where one is forced to watch an entire ad, than for the skippable pre-roll advertising.
Furthermore, scholars suggest that pre-roll ads might negatively influence the ad attitude, brand attitude and the purchase intention (Li, Edwards & Lee , 2002). According to Campbell,
Mattison, Thompson, Grimm & Robson (2017), pre-roll advertising can cause avoidance and irritation with users. However, Campbell et al. (2017) do not differentiate between skippable and un-skippable pre-roll ads. It is therefore unclear if and how the effects of brand and ad attitude and purchase intention differ between skippable and un-skippable pre-roll advertising. But
6 drawing upon the psychological reactance theory (Brehm,1966) it can be expected that
un-skippable ads will have a more negative influence than un-skippable ads.
2.2 Advertising effectiveness
In this research, the advertising effectiveness is measured by looking at three different elements: attitude towards the ad, brand attitude and purchase intention.
One of the important things for advertisers is to make effective advertising. One factor that indicates the effectiveness of an ad is the attitude towards the ad (Belanceh, Flavián & Pérez-Rueda, 2017; Goldsmith, Lafferty & Newell, 2000). MacKenzie and Lutz (1989) define attitude towards an ad as “a predisposition to respond in a favorable or unfavorable manner to a
particular advertising stimulus during a particular exposure occasion”. The attitudes towards the ad can affect the attitude towards the brand, which is another indicator of ad effectiveness (Belanceh, Flavián & Pérez-Rueda, 2017; Gardner, 1985). The most basic definition of brand attitude is the overall evaluation of a brand (Low & Lamb, 2000). Previous research states that advertising that interrupt this flow can negatively influence ad attitude brand attitude and purchase intention. Purchase intention is the likelihood that a customer will buy a particular product (Fishbein and Ajzen, 1975). Spears and Singh (2004) define purchase intention as “an individual’s conscious plan to make an effort to purchase a brand”. Furthermore, scholars state that there is a positive relationship between advertising attitude and purchase intentions (Kim & Han, 2014). From the above the following hypotheses are arrived at:
H1a: The option to skip an ad will lead to a more positive brand attitude H1b: The option to skip an ad will lead to a more positive ad attitude
H1c: Purchase intention will be more likely when there is an option to skip an ad
2.3 Perceived control
Pre-roll advertising in YouTube videos that provide an option for users to skip the ad might give back a sense of control to the users (Hegner, Kusse & Pruyn, 2016). This can be seen as an option for the visitor to control their freedom, which was just taken away from them. McCoy et al.’s (2008) study indicates that having control over removing or skipping an ad reduces
7 perceived intrusiveness. In the case of skippable pre-roll advertising in YouTube videos, this might give back a sense of control to the users, which will reduce the perceived intrusiveness of the ad (Hegner, Kusse & Pruyn, 2016). At the same time, the interruption of an ad or having to click to skip the ad in order to view the video can elicit negative feelings in the user. These will be even more when there is no option to skip the ad and the user is being forced to watch it entirely (McCoy et al., 2008).However the intrusive feeling that is being caused by an ad can be reduced by a sense of control (Mormito & Chang, 2006).
The main purpose of advertising in general is to increase potential buyers (Adegoke & Popoola, 2018). In order to reach the desired effect or any advertising effects, it is important that attention is being paid to the advertisement itself. To achieve this, advertising aims to interrupt the flow of editorial content (Ha, 1996). These interruptions are often perceived as intrusive, it can therefore be stated that intrusiveness is in the very nature of advertising itself (McCoy et al., 2008). Li, Edward and Lee (2001) define intrusiveness as “a perception or psychological
consequence that occurs when an audience’s cognitive processes are interrupted”. This definition of Li, Edward and Lee (2001) states that ads are not intrusive themselves, but that ads must be perceived as interrupting the goals of the viewers to be considered intrusive.
As described above, the psychological reactance theory (Brehm, 1966) states that once freedom is being threatened or restricted, an individual is motivated to restore this freedom. According to this theory, perceived control serves as a mediator. By restoring freedom, the perceived intrusiveness will be lessened. This explains why the perceived intrusiveness of an ad can be reduced by providing a button that enables users to skip the ad. It is therefore expected that in this experiment, if people are given the option to skip the ad (after 5 seconds), they will feel more in control than when they are given no option to skip the ad. Thus, their perception of control will be greater and the perceived intrusiveness of the ad will be lessened
accordingly.Based on the literature above the following hypotheses are proposed:
H2: The option to skip an ad will lead to more perceived control H3: More perceived control will lessen the perceived intrusiveness
H4: The effect of being able to skip an ad on the perceived intrusiveness is mediated by the perceived control
8
2.4 Intrusiveness on ad effectiveness
Ironically, the interrupting and intrusive aspects of advertising that are important to catch the attention of the users, can have negative consequences that may elicit the opposite of the desired effects (Campbell, Mattison Thompson, Grimm, Robson, 2017; Li, Edwards & Lee, 2002). Scholars state that intrusiveness can cause irritation and negative attitudes towards the
advertisement, which can negatively impact the brand attitude and purchase intention (Cho & Cheon, 2014; Goodrich, Schiller & Galletta, 2015; MacKenzie and Lutz, 1989). Many scholars explain the link between perceived ad intrusiveness and negative effects on the brand image by referring to the Psychological reactance theory by Brehm (1966). This theory explains how people react when they feel like their freedom to engage in a behavior is being threatened or restricted. The reactance theory states that when a freedom is being restricted or threatened a person wants to restore there freedom. It states that once a person loses his/her freedom, this freedom becomes more attractive and the individual will then seek for ways to restore said freedom and control (Edwards, Lee & Li, 2002; McCoy et al., 2008; Morimoto & Chang, 2006). For example, an individual wants to watch a YouTube video about how to make lasagne. She clicks on the video to watch it, but then all of a sudden a pre-roll ad appears before the video that the person intended to watch is played. In this case, the ad might be perceived as a restriction to access the video and thus, according to the psychological reactance theory, watching the video becomes more attractive and at the same time the restriction may elicit negative feelings towards the ad. The following hypotheses are based on the theory that is discussed above:
H5a: Low perceived intrusiveness will generate a more positive attitude towards the ad than when the perceived intrusiveness is high
H5b: Low perceived intrusiveness will generate a more positive attitude towards the brand than when the perceived intrusiveness is high
9
H6a: The effect of being able to skip an ad on the ad attitude is mediated by the
perceived intrusiveness
H6b: The effect of being able to skip an ad on the brand attitude is mediated by the
perceived intrusiveness
H6c: The effect of being able to skip an ad on the purchase intention is mediated by the
perceived intrusiveness
10
3. METHOD
In order to test the hypotheses that were set up in section 2, we have conducted an
online-experiment. Participants were exposed to both an ad and a YouTube video, the ad appeared prior to the video. The online experiment had a one factor design, with two conditions: a condition that shows an ad which is skippable after five seconds and a condition with no option to skip the ad. Participants were randomly assigned to one of the two conditions and were asked to complete a questionnaire afterwards.
3.1 Procedure & Design
The online experiment was available online between 16-29 January 2018. Participants were first explained the purpose of the study and were asked to agree with the informed consent. The participants were randomly assigned to one of the two conditions (ad could be skipped after five seconds/no option to skip the ad). After that the participants were instructed to watch a video about pandas, but when the participants proceeded to the next page they would see a video-ad first, instead of the video about pandas. The participants in the skippable condition were provided with the option to skip the ad after five seconds. The participants who were randomly assigned to the non-skippable condition had to watch the entire video ad (28 seconds) before the video about pandas would automatically play. In an attempt to eliminate effects of the use of different devices and audio output, the participants were asked to watch the videos on a PC/Laptop, with sound. After the pre-roll ad and video, participants were presented a questionnaire with questions concerning their perceptions, attitudes and demographics.
The video advert that is used for this online experiment is a 30 second ad for Nikon that is promoting the 350D video camera (see Appendix B). This ad was chosen because the product promoted in it is a gender-neutral product and because the logo of Nikon appears directly from the start of the ad. Therefore the participants of both the skippable and un-skippable condition should be able to see what brand the ad is about. The video that the participants are instructed to watch is a short informational BBC video fragment about pandas (see appendix C). This choice
11 for this fragment about pandas derives from the fact that the topic is relatively neutral (gender neutral, no political views are being shared) and informative.
3.2 Instruments/Measurements
All items of the scales below were measured on a 7-point Likert or a 7-point semantic differential scale.
Perceived intrusiveness
Perceived intrusiveness was measured with 7 items. The scale measuring the intrusiveness of advertising is developed by Li, Edwards and Lee (2002). “When the ad was shown, I thought it was … Distracting, Disturbing, Forced, Interfering, Intrusive, Invasive, and Obtrusive”
Perceived control
To measure perceived control, items from Azjen (2002) were selected and adjusted to the
purpose of this study. “When I wanted to watch the video but the ad appeared, I felt like:” “What was happening was mostly up to me”, “I was not capable of deciding what was happening at all”, “I was in total control of the situation” were measured on a Likert scale ranging from 1 (strongly disagree) to 7 (strongly agree).
Ad attitude
To measure the attitude towards the ad was based on a semantic differential scale by Choi Miracle & Biocca (2001): Bad/Good, Dislike/Like, Unfavorable/Favorable,
Unenjoyable/Enjoyable, Unpleasant/Pleasant and Awful/Nice.
Brand attitude
A semantic differential scale by Choi Miracle & Biocca (2001) was used to measure the brand attitude: Positive/Negative, Satisfactory/Unsatisfactory, Favorable/Unfavorable, Good/Bad, and Likeable/Dislikeable
Purchase intentions
A three-item scale by Rosengren, Modig & Dahlen (2015) was used to measure purchase
12 this product” and “I would consider buying this product” were measured on a Likert scale
ranging from 1 (strongly disagree) to 7 (strongly agree)
3.3 Participants
The videos and the questionnaire used for the online experiment were spoken and written in English. The link to the online experiment was posted on several Facebook groups aimed at people with sufficient English knowledge and people with the minimum age of 18. The link was also shared through WhatsApp with relatives and friends who speak sufficient English in order to participate. Furthermore, participants were asked to share the link of the online-experiment with friends they know that also fit the age criteria and had sufficient knowledge of English. The participants (N = 102) were selected by means of convenience sampling. After inspection 8 participants were taken out of the sample, because of incorrectly answering the question “Which brand was the advertising about”. The responses from the remaining participants (N=93) were used for further analysis. The average age of the participants was 25.34 (SD=24.11). 41.9% of the participants were male (N=39) and 58.1% of the participants were female (N=54). More than half of the participants, 65.6% (N=61) use YouTube every day or several times a day.
3.4 Manipulation Checks
For the manipulation check, a multiple-choice question was used. Asking: “What brand was the
advertising about’’? Participants could pick an answer out of the following options: “Sony”, “Nikon”, “Harman Kardon” or “I don’t know”.
4. RESULTS
First a Factor Analysis was done and all the reliability of all scales were tested as is shown in Table 1.The Result of the exploratory Factor Analysis to check discriminant validity, with principal axis factoring (PAF) show that the total R² = 0.74, this means that the five dimensions explain 74% of the total variance of the 24 items. The Kaiser-Meyer-Olkin (KMO) = .84 this means that the factor analysis yields distinct and reliable factors, KMO values between 0.8 and 0.9 are considered as, great (Field, 2015).
13 Table 1
Results of Reliability and Factor Analysis (PAF) with oblimin rotation (n = 93)
Factor loadings (??) construct AA INT PC BA PI ad attitude Q10_1 Bad:Good .79 -.03 .03 -.16 -.02 Q10_2 Dislikable:Likeable .85 -.07 -.02 .03 .00 Q10_3 Unfavorable:Favorable .79 .03 -.02 -.05 -.10 Q10_4 Unenjoyable:Enjoyable .89 .02 .04 .10 -.15 Q10_5 Unpleasant:Pleasant .84 -.02 -.04 -.03 .02 Q10_6 Awful:Nice .78 -.06 .02 -.19 .06 brand attitude Q12_1 Negative:Positive -.10 -.12 -.01 -.89 -.09 Q12_2 Unsatisfactory:Satisfactory .09 .10 -.06 -.87 -.03 Q12_3 Unfavorable:Favorable .13 -.02 .02 -.82 .09 Q12_4 Bad:Good -.06 .01 -.04 -.84 -.05 Q12_5 Dislikable:Likable .16 .02 .01 -.77 -.01 purchase intention
Q14_1 I want to try this product .03 -.05 -.10 -.22 -.67 Q14_2 I want to buy this product .17 -.10 -.12 .01 -.86 Q14_3 I would consider buying this product .32 .04 .11 -.07 -.59 perceived
control
Q15_1 What was happening was mostly up to me -.11 .03 -.73 -.10 -.16 Q15_2 I was not capable of deciding what was happening -.03 -.02 .60 .00 -.06 Q15_3 I was in total control of the situation .05 .00 -.86 .03 .02
perceived intrusiveness Q16_1 Distracting -.06 .27 .14 -.03 -.24 Q16_2 Disturbing -.08 .47 -.05 .01 -.05 Q16_3 Forced -.05 .72 .13 -.15 .07 Q16_4 Interfering -.09 .73 .04 .01 .06 Q16_5 Intrusive .13 .98 -.08 .01 .12 Q16_6 Invasive .06 .92 -.07 .11 -.06 Q16_7 Obtrusive -.03 .80 -.01 .02 .00 Eigenvalues 8.51 4.15 2.19 1.69 1.28 % of variance (R2) 35.46 17.31 9.13 7.02 5.35 cumulative % (R2) 35.46 52.76 61.89 68.91 74.26 Cronbach's α .95 .94 .87 .76 .87
bold factor loadings are above |.30|
χ2
(276) = 1803.2, p < .001. Bartlett test of sphericity Based on Eigenvalue criterion (eigenvalues > 1) *
AA= Ad Aditude, BA= Brand Attitude, PI= Purchase Intention, INT= Perceived Intrusiveness, PC= Perceived Control
Furthermore, scales for perceived control, perceived intrusiveness and advertising effectiveness and were tested for their reliability. The six items that together create the scale for ad attitude has a Cronbach’s Alpha score of α=.95, which meets the standard of an excellent reliable scale. The items that together created the scale for brand attitude was tested and has a Cronbach's Alpha score of α= 0.94. The scale for purchase intention consists out of 3 items has a Cronbach's Alpha
14 score of α= 0.87, therefore both scales to measure brand attitude and purchase intentions meet the standards of a reliable scale. The scale for perceived control was tested for its reliability. The scale consisted three items and has a Cronbach's Alpha score of α= 0.76, which indicates that the scale is a reliable scale to measure perceived control. The 7 items that together create the scale of perceived intrusiveness was tested for its reliability, this scale has a Cronbach's Alpha score of α= 0.87 and therefore meets the standards of a reliable scale
4.1 Manipulation & Randomization Check
To see whether the participants were aware of which brand the advertising was about the question “Which brand was the advertising about?” 91.2% (N=93) answered: ‘Nikon’ which was the correct answer 7.8% answered: ‘I don’t know’ and only 1% answered ‘Sony’. The question:“Could you skip the ad?” answer yes/no was originally a second manipulation check. However many participants reached out after the experiment to, explain that they did not pay close attention to the ad, as they thought it was real. (e.g. “I was looking at my phone because I thought it was a real ad, but I did really pay attention to everything else”) Excluding the
participants that incorrectly answered this question leads to the same results pattern. However the sample size would have been very small. Therefore we decided not to exclude the these
participants from the sample size
In order to see whether the conditions are randomized, the characteristics of age, gender, education and the frequency of YouTube consumption of both groups are compared. Age was measured as continuous variable and is therefore tested with an independent T-test Chi- Square test will be used for the binary variable, gender. And a Mann-Whitney U test, will be used for the ordinal variables, education and the frequency of YouTube consumption Results showed that the randomization was successful for all variables: age Age, t(91) = -0.24, p = .81; Cohen’s D = .05 and for gender Gender χ 2 (1) = 0.42, p = .52; ϕ = .09 . The results of the Mann- Whitney U test showed there was no difference between the two groups for Education, U = 957, p = .34, as well as the frequency of YouTube consumption, U = 944, p =.29.
15
4.2 Correlations
Table 2
Correlations between the observed variables in the sample (n = 93)
The correlations between the dependent and independent variables are presented in Table 2. In this table a positive relation between the following dependent variables: brand attitude (BA), purchase intention (PI) and ad attitude (AA) can be found. The results show that a strong positive relation exists between these variables (R2 > .25). These results also display a negative
relationship between ad attitude and perceived intrusiveness r = -.34, p < .001; R2 = .12, thus a medium to large effect. This means that when the perceived intrusiveness is higher the ad attitude is lower Furthermore a relation between Gender and the frequency of YouTube consumption is noticeable, women appear to watch less YouTube than men, rho = .35, p < .001.The PI is higher for women than for men r = .21, p = .04; R2 = .04.
AA BA PI PC INT Age gender edu
You Tube AD Attitude 1 Brand attitude .55** 1 Purchase intention .57** .54** 1 Perceived control .08 .15 .12 1 Perceived intrusiveness -.34** -.06 -.04 -.12 1 Age .03 -.18 .04 .04 .12 1 Gender .16 .20 .21* -.05 -.13 -.17 1 Educationa .17 -.01 .04 .13 .12 .35** -.11 1 frequency YouTubea -.05 -.10 -.12 .06 .12 -.08 -.29** .16 1 ** p < .01; * p < .05 (2-tailed). a
16
4.3 T-test & Descriptive statistics
Table 3
Descriptive statistics and independent sample T-tests comparing group means of conditions to answer H1 and H2
Unskippable (n = 50) Skippable (n = 43) Differences between conditions
M SD M SD MD t-value p-value AD Attitude 4.42 1.40 4.56 1.05 -0.14 -0.53 .60 Brand attitude 5.33 1.04 5.33 0.86 0.00 -0.01 .99 Purchase intention 4.14 1.40 3.78 1.16 0.36 1.32 .19 Perceived control 3.51 1.54 3.84 1.22 -0.34 -1.16 .25 Perceived intrusiveness 5.49 0.95 4.88 1.03 0.61 2.99 <.01 Age 25.26 3.68 25.47 4.60 -0.21 -0.24 .81
In order to test H1a,b,c and H2 a T-test is performed. No significant difference in means are found for ad attitude between the un-skippable (M=4.42, SD= 1.40) and skippable condition (M= 4.56, SD =1.05), t(91) = -0.53, p = .60; Cohen’s D = .11. The results for brand attitude show no difference between the un-skippable condition (M= 5.33, SD=1.04) and skippable condition (M=5.33, SD= 0.86, t(91) = -0.01, p = .99; Cohen’s D = .01. There is no difference found between the conditions for purchase intention un-skippable (M= 4.14, SD=1.40) condition and the skippable condition (M=3.78, SD= 1.16), t(91) = 1.32, p = .19; Cohen’s D = .49. Thus (H1a, b, c) can be rejected. Furthermore, the results show no significant difference in the means for perceived control, between the un-skippable (M= 3.51, SD=1.54) condition and the skippable condition (M=3.84, SD=1.22), t(91) = -1.16, p = .25; Cohen’s D =.24. And therefore H2 can be rejected. Although no predictions were made for the following finding, a significant difference between the conditions for perceived intrusiveness was found. The condition with the un-skippable ad ( M=5.49, SD=.95) was perceived significantly more intrusive that the condition with the skippalbe ad (M=4.88, SD=)1.03, t(91)=2.99, p=004; Cohens’s D=0.62T
17
4.4 Regression models
In order to check for the assumptions of regression, the models are cheked for no severe multicollinearity, independence of observations , normality of residuals , homoscedastic distribution of residuals and multivariate outliers. Multicollinearity is checked with VIFs (<5 ) (Rogerson, 2001), independence of observations with the Durbin- Watson, these are all between 1 and 3 Normality of residuals is checked by inspecting a histogram of the residuals, this looks normally distributed. The homoscedastic distribution of residuals is checked by the use of a scatter plot, the scatterplot shows no severe heteroscedasticity. Finally the multivariate outliers are being checked by the standardized residuals (all z values are between -3 and 3) (Field, 2013).
Table 4
Results of regression analyses in order to test the hypotheses (H3-H6), n = 93
Dependent variable
INT INT AA AA BA BA PI PI
Model 1 Model 2 Model 3 Model 4 Model 5 Model 6 Model 7 Model 8 Constant 5.536*** 5.717*** 6.610*** 6.760*** 5.612*** 5.653*** 4.221*** 4.739*** (18.56) (19.44) (10.40) (9.49) (10.90) (9.79) (6.00) (6.07) Perc. control -0.090 -0.064 (-1.18) (-0.87) Perc. intrusiveness -0.408*** -0.425*** -0.054 -0.059 -0.047 -0.109 (-3.40) (-3.38) (-0.56) (-0.58) (-0.36) (-0.79) CONDITION -0.593*** -0.123 -0.034 -0.424 (-2.86) (-0.47) (-0.16) (-1.50) R2 .015 .097 .113 .115 .003 .004 .001 .026 R2adjusted .004 .077 .103 .096 -.007 -.018 -.010 .004 F 1.391 4.841*** 11.59*** 5.856*** 0.313 0.168 0.127 1.183 p-value .241 .010 <.001 0.004 .577 .846 .722 .311 t-statistics in parentheses, *** p<0.01, ** p<0.05, * p<0.1
18 To test H3- H6 regression analyses were performed. H3 expected that a higher level of perceived control would lessen the perceived intrusiveness. Yet the results of the regression analyses (Model 2) show that there was no significant effect of perceived control on intrusiveness b* = -0.06, t(90) = -0.87, p =.37; bèta = -.09. Therefore H3 can be rejected. As the results of H3 already show there is no significant effect of perceived control on perceived intrusiveness (path B). Not surprising that H4 is then also rejected as it shows that the effect of condition on perceived intrussiveness is not mediated by control. However a significant difference between the conditions for perceived intrusiveness was found (Model 2) , The regression analyses shows that when controlled for perceived control, participants in the un-skippable condition perceived the ad as significantly more intrusiveness b* = -0.59, t(90) = -2.86, p = .005; bèta = -.29. This is in line with the results of the t-test (Table. 3) the condition with the un-skippable ad (M=5.49, SD=.95) was perceived significantly more intrusive than the condition with the skippable ad (M=4.88, SD=1.03), t(91) = 2.99, p = <.01; Cohen’s D = .62.
The perceived intrusiveness had a significant negative effect on the ad attitude of the participants b* = -0.41, t(91) = -3.40, p <.001; r = -.34. (Model 3) Which means that the higher the ad intrusiveness is the lower the ad attitude will be, this is in line with the predictions made for H5a, and therefore H5a can be accepted. But there was no significant effect of perceived intrusiveness on brand attitude found b* = -0.05, t(91) = -0.56, p =.58; r = -.06. (Model 5) or on the purchase intention b* = -0.05, t(91) = -0.36, p =.72; bèta = -.09. (Model 7) Therefore H5b, c can be rejected.
Finally we proposed that the effect of being able to skip an ad on ad effectiveness (ad attitude, brand attitude and purchase intention) would be mediated by the perceived intrusiveness The results show that a significant negative effect is found of perceived intrusiveness on ad attitude after controlling for condition, b* = -0.43, t(90) = -3.38, p = .001; bèta = -.35. However there is no direct effect of the condition on the ad effectiveness. An additional test shows that there is no difference for the effect of condition on ad attitude, with or without perceived
intrussive. Thus perceived intrusiveness does not mediate the relationship between condition and ad attitude. This is in line with the mediation model of Baron and Kenny (1986) (Figure 2.) that states that mediation is not possible if path C, which is in this case the effect of condition on ad attitude is not significant. Thus, a mediation by perceived intrusiveness is not possible.
19 Furthermore, no mediation by perceived intrusiveness is found for the effects of condition on brand attitude, b* = -0.06, t(90) = -0.58, p = .56; bèta = -.06 and purchase intention, b* = -0.11, t(90) = -0.79, p = .43; bèta = -.09., even after controlling for condition. Therefore, H6 a, b, c can be rejected
Figure 2. Details from Baron and Kenny’s mediation model
5. CONCLUSION & DISCUSSION
The purpose of this study was to add to the literature of online advertising and to gain more knowledge of the effectiveness of pre-roll advertising on the user’s attitude and behavioral intentions. The results of this study will be discussed in the light of the hypotheses that were set up in this study. First we proposed that the option to skip an ad would increases the ad
effectiveness. The results of an independent t-test found that the participants who had an option to skip did not had a more positive ad attitude (H1a) or brand attitude(H1b) and there was there a higher likelihood of purchase intention (H1c). Second it was expected the perceived control was higher for the participants who had the option to skip the ad than the participants who did not have the option skip the ad (H2). However the findings do not support this. Furthermore it was proposed that more perceived control would lessen the perceived intrusiveness (H3) and the effect of being able to skip an ad on the perceived intrusiveness would be mediated by the perceived control (H4). The regression analysis proved these assumptions to be false. Thus there was no effect of the option to skip an ad on control (see path A in the mediation of model Baron and Kenny (1986) Figure 2) and there was also no mediation effect found (Path B). However an unexpected, yet significant effect was found for the option to skip an ad on perceived
20 intrusiveness (Path C). This finding indicates that the presence or the absence of an option to skip the ad is not relevant for the advertising effectiveness. It also shows that perceived control is affected by the option to skip an ad. Perceived control also does not influences the perceived intrusiveness. An explanation for this might be that with or without the option to skip an ad the user still feels in control because one can passively avoid the ad (by not watching the ad, looking somewhere else clicking on another screen).However the results do show that the option of a skip button does impact the perceived intrusiveness of the users, the un-skippable ad seems to be perceived as more intrusive than the skippable ad ad.
Furthermore perceived intrusiveness negatively influenced the ad attitude (H5a). It is remarkable that this was not the case for brand attitude or purchase intention of the participants (H5b,c).The literature states that ad attitude affects the brand attitude, and a positive relationship exists between ad attitude brand attitude and purchase intention (Belanceh, Flavián & Pérez-Rueda, 2017). Perhaps the results of this study can be explained by the brand that is used in the experiment. Almost all participants 96.8% (N=90) were already familiar with the brand Nikon, before participating in this study. It could be that the brand attitudes of the participants were already well established and therefore the presence or an absences of a skip button could not interfere with the brand attitude or purchase intention. Perhaps the result for unknown brands will be different. Although the option to skip an affected the perceived intrusiveness, and the perceived intrusiveness influenced the ad attitude. No evidence was found that the effect of an option to skip an ad was mediated by the perceived intrusiveness.
It is important to note that this study is limited. The findings of these study are limited to just one format of pre-roll ads, the length is not taken into consideration. The ad used in this experiment can be considered as a relatively short ad (28 seconds), perhaps the results could be different when a long ad (e.g. 120 seconds) was used. A second limitation of this study is the placement of the button that allows users to skip the ad. On YouTube this button is implemented in the video itself, the button appears after five seconds in the same screen as where the pre-roll ad is shown. Due to technical reason this was not possible in this experiment and therefore the button that allowed the participants to skip the appeared underneath the pre-roll ad. Again due to technical reasons a timer was added under the video, to make visible how long the ad would last, and how long the YouTube video about panda was. This is normally embedded in the video itself.
21 The knowledge gained from this study, is useful for advertisers. Currently the cost of pre-roll ads on YouTube are based on the following. For the pre-roll ads that are not skippable the advertiser obviously needs to pay. But for the pre-roll ads that are skippable. The advertiser is only charged when a viewer watches for at least 30 seconds or to the end of the video, whichever is shorter. If a user decides to skip the ad after five seconds or before the video has ended the advertiser does not have to pay for the ad (Trueview, 2016). Regarding the results of this study it is advisable for advertisers to always go for the skippable option.
22
LITERATURE
Adegoke, A. S., & Popoola, O. N. (2018). Promotional Mix as a Determinant of Consumer Buying Decision in the Food and Beverages Industry: A Case Study of Nigeria Bottling Company Plc., Asejire Ibadan. World Academy of Science, Engineering and Technology, International Journal of Economics and Management Engineering, 5(4).
Alhabash, S., Baek, J. H., Cunningham, C., & Hagerstrom, A. (2015). To comment or not to comment?: How virality, arousal level, and commenting behavior on YouTube videos affect civic behavioral intentions. Computers in human behavior, 51, 520-531.
Airoldi, M., Beraldo, D., & Gandini, A. (2016). Follow the algorithm: An exploratory investigation of music on YouTube. Poetics, 57, 1-13.
Baron, R. M., & Kenny, D. A. (1986). The moderator–mediator variable distinction in social psychological research: Conceptual, strategic, and statistical considerations. Journal of personality and social psychology, 51(6), 1173
Brackett, L. K., & Carr, B. N. (2001). Cyberspace advertising vs. other media: Consumer vs. mature student attitudes. Journal of advertising research, 41(5), 23-32.
Brehm, J. W. (1966). A theory of psychological reactance.
Campbell, C., Mattison Thompson, F., Grimm, P. E., & Robson, K. (2017). Understanding Why Consumers Don't Skip Pre-Roll Video Ads. Journal of Advertising, 46(3), 411-423.
Chatterjee, P. (2008). Are unclicked ads wasted? Enduring effects of banner and pop-up ad exposures on brand memory and attitudes. Journal of electronic commerce Research, 9(1), 51.
Chen, Q., & Wells, W. D. (1999). Attitude toward the site. Journal of advertising research, 39(5), 27-38.
23 Cho, C. H., & as-, U. O. T. A. A. I. A. (2004). Why do people avoid advertising on the internet?. Journal of advertising, 33(4), 89-97.
Choi, Y. K., Miracle, G. E., & Biocca, F. (2001). The effects of anthropomorphic agents on advertising effectiveness and the mediating role of presence. Journal of Interactive Advertising, 2(1), 19-32.
Dehghani, M., Niaki, M. K., Ramezani, I., & Sali, R. (2016). Evaluating the influence of
YouTube advertising for attraction of young customers. Computers in Human Behavior, 59, 165-172.
Edwards, S. M., Li, H., & Lee, J. H. (2002). Forced exposure and psychological reactance: Antecedents and consequences of the perceived intrusiveness of pop-up ads. Journal of Advertising, 31(3), 83-95.
Field, A. (2013). Discovering statistics using IBM SPSS statistics. sage.
Fishbein, M., & Ajzen, I. (1975). Belief, attitude, intention and behavior: An introduction to theory and research
Gambaro, M. and Puglisi, R. (2012), “Complement or substitute? The internet as an advertising channel, evidence on advertisers on the Italian market, 2004-2009,” in: Departmental Working Papers from Department of Economics, Management and Quantitative Methods at Università degli Studi di Milano
Gardner, M. P. (1985). Does attitude toward the ad affect brand attitude under a brand evaluation set?. Journal of Marketing Research, 192-198.
Goldsmith, R. E., Lafferty, B. A., & Newell, S. J. (2000). The impact of corporate credibility and celebrity credibility on consumer reaction to advertisements and brands. Journal of advertising, 29(3), 43-54.
24 Goodrich, K., Schiller, S. Z., & Galletta, D. (2015). Consumer reactions to intrusiveness of online-video advertisements. Journal of Advertising Research, 55(1), 37-50.
Ha, L. (1996). Advertising clutter in consumer magazines: Dimensions and effects. Journal of advertising research, 36(4), 76-85.
Hegner, S. M., Kusse, M. D. C., & Pruyn, A. T. (2016). Watch it! the influence of forced pre-roll video ads on consumer perceptions. In Advances in Advertising Research (Vol. VI) (pp. 63-73). Springer Fachmedien Wiesbaden.
IAB (2013), “A Comprehensive Picture of Digital Video and TV Advertising: Viewing, Budget Share Shift and Effectiveness,” retrieved 10th November 2013 from
http://www.iab.net/media/file/Digital-Video-and-TV-Advertising-Viewing-Budget-Share-Shiftand-Effectiveness-FINAL.pdf
Kim, Y. J., & Han, J. (2014). Why smartphone advertising attracts customers: A model of Web advertising, flow, and personalization. Computers in Human Behavior, 33, 256-269.
Lee, J., Ham, C. D., & Kim, M. (2013). Why people pass along online video advertising: From the perspectives of the interpersonal communication motives scale and the theory of reasoned action. Journal of interactive Advertising, 13(1), 1-13.
Li, H.; Edwards, S. and Lee, J. (2002) “Measuring the Intrusiveness of Advertisements: Scale Development and Validation,” in: Journal of Advertising, Vol. 31, No. 2, 37-47.
Li, H., & Lo, H. Y. (2015). Do you recognize its brand? The effectiveness of online in-stream video advertisements. Journal of Advertising, 44(3), 208-218.
Low, G. S., & Lamb Jr, C. W. (2000). The measurement and dimensionality of brand associations. Journal of Product & Brand Management, 9(6), 350-370.
25 MacKenzie, S. B., Lutz, R. J., & Belch, G. E. (1986). The role of attitude toward the ad as a mediator of advertising effectiveness: A test of competing explanations. Journal of marketing research, 130-143.
MacKenzie, S. B., & Lutz, R. J. (1989). An empirical examination of the structural antecedents of attitude toward the ad in an advertising pretesting context. The Journal of Marketing, 48-65.
McCoy, S., Everard, A., Polak, P., & Galletta, D. F. (2008). An experimental study of antecedents and consequences of online ad intrusiveness. Intl. Journal of Human–Computer Interaction, 24(7), 672-699.
Morimoto, M., & Chang, S. (2006). Consumers’ attitudes toward unsolicited commercial e-mail and postal direct mail marketing methods: intrusiveness, perceived loss of control, and irritation. Journal of Interactive Advertising, 7(1), 1-11.
Pasadeos, Y. (1990), “Perceived informativeness of and irritation with local advertising,” in: Journalism & Mass Communication Quarterly, Vol. 67, No. 1, 35-39
Pikas, B., & Sorrentino, G. (2014). The effectiveness of online advertising: consumer's perceptions of ads on Facebook, Twitter and YouTube. The Journal of Applied Business and Economics, 16(4), 70.
Rogerson, P. A. (2001). Statistical methods for geography. London: Sage.
Rosengren, S., Modig, E., & Dahlén, M. (2015). The value of ambient communication from a consumer perspective. Journal of Marketing Communications, 21(1), 20-32.
Spears, N., & Singh, S. N. (2004). Measuring attitude toward the brand and purchase intentions. Journal of Current Issues & Research in Advertising, 26(2), 53-66.
26 Trueview. (2016, August). Retrieved from https://www.thinkwithgoogle.com/products/youtube-trueview/
27 Appendix A
28 Appendix B
29 Appendix C :Questionnaire
30
