When you realize that big brother is watching: How informing consumers affects synced advertising effectiveness

ABSTRACT

New personalization technologies have made it possible to deliver personalized messages to consumers based on their offline media usage in real time, which is known as synced advertising. These developments go hand-in-hand with a rise in concerns related to consumers’ privacy. The aim of this study is to examine the effects of increasing consumer knowledge of synced advertising (SA) on resistance through critical attitudes and perceived surveillance. In two online experiments, we found that increased SA awareness and knowledge generates more critical attitudes and a greater level of perceived surveillance, which leads to more resistance to a synced ad. For consumers without prior SA experience (e.g., through education, work), providing technical information on SA is the most effective in increasing SA knowledge (both objective and subjective knowledge), but personally relevant information could help increase confidence in SA knowledge (i.e., subjective knowledge). These results advance theories of persuasion knowledge, as well as the underlying mechanism of synced advertising effects on consumer empowerment. The results contribute to literacy programs by showing what type of information could help consumers make informed decisions about this new personalization strategy.

KEYWORDS:

• Synced advertising
• persuasion knowledge
• personalization
• perceived surveillance
• resistance

Introduction

The majority of digital advertising messages are personalized based on consumers’ interests, behaviors, and needs (Varnali 2021). Consumers receive messages on websites that are related to previously visited websites or that are based on their current media usage. The latter is known as synced advertising (SA), which is a relatively new form of personalized advertising in which messages on mobile devices are synced with consumers’ concurrent media usage in real time (Segijn 2019). An example of SA is when a consumer receives a mobile ad for something that is simultaneously being discussed or shown on TV. Automatic content recognition technologies make it possible to synchronize ads across media in real time (Garrity 2018; Segijn and van Ooijen 2020a; Webwire 2017). SA is found to increase attention (Segijn, Voorveld, and Vakeel 2021), memory (Hoeck and Spann 2020), brand attitudes (Segijn and Voorveld 2021), and engagement with the ad (Webwire 2017) compared to a non-synced ad. However, different reports show that consumers lack knowledge about personalized advertising and data collection techniques (McDonald and Cranor 2010; Segijn and van Ooijen 2020b; Smit, van Noort, and Voorveld 2014). This is problematic because to be an informed consumer, it is important that consumers develop sufficient knowledge and skills (i.e., persuasion knowledge/advertising literacy) to resist such persuasive tactics (Hudders et al. 2017).

To date, scholars have a limited understanding of the extent to which SA practices affect consumers and how consumers could defend themselves against such practices. The proposed effects of SA are described (for an overview see Segijn 2019), but empirical studies are needed to understand consumers’ responses to SA, such as resistance. Filling this gap is needed because of SA prevalence in the industry (Kantrowitz 2014; Webwire 2017), consumers’ limited SA knowledge (Segijn and van Ooijen 2020b), and worrisome consequences for consumers’ privacy (Phelan, Lampe, and Resnick 2016). In line with developments in privacy regulations, such as the General Data Protection Regulation in Europe and the California Consumer Privacy Act, which aim to give consumers more control over their personal data (van Ooijen and Vrabec 2019; Strycharz et al. 2019, 2021), it is important to study how increased SA literacy could help consumers become informed decision makers. Theoretical knowledge is needed to understand whether and how SA literacy could influence resistance for consumers to defend themselves against SA practices. The aim of the proposed project is to examine to what extent increasing SA literacy affects consumer resistance to a synced ad through critical attitudes and perceived surveillance.

The proposed study is innovative and contributes to the literature because, to our knowledge, it will be the first to disentangle SA effects with a focus on SA literacy. This is important to study because when consumers are more literate (i.e., have knowledge and capacity to challenge SA), they are empowered to make their own informed decisions in response to SA (e.g., resistance). The study will advance theoretical knowledge into how SA literacy affects resistance toward SA by examining critical attitudes and perceived surveillance as underlying mechanisms of the effect. Although critical attitudes toward advertising have been examined in relation to persuasion knowledge and resistance toward ads (e.g., Boerman, Van Reijmersdal, and Neijens 2012; Eisend et al. 2020; Ham 2017), the examination of perceived surveillance is new in this context. However, such examination is much needed given the importance of this factor in personalized advertising research (Farman, Comello, and Edwards 2020; Segijn and van Ooijen 2020a). The results are relevant to the advertising literature by providing insights on persuasion and resistance in the context of this new personalization strategy. Finally, this line of inquiry contributes to the literature of personalized messaging, advertising literacy, and privacy. For example, insights can be used to better understand how consumers can defend themselves against innovative message strategies.

Furthermore, considering the growth and potential of SA and its influence on consumers, investigating SA will provide important implications for educational programs and policy makers. Findings from this research may generate debate on the ethical implications of these practices within the mainly self-regulated advertising industry. Moreover, knowledge of SA helps steer the societal debate regarding privacy concerns, ethics, and consumer empowerment. Insight into SA could help extend advertising literacy programs beyond traditional advertising strategies to consider emerging ethical dilemmas in a synced advertising environment. Finally, research into SA can provide more insights into the consequences of utilizing consumers’ data to personalize messages, which could inform state and national regulations for data privacy.

Theoretical background

According to the persuasion knowledge model (PKM), individuals develop knowledge and beliefs about how advertising persuades and the specific tactics that are used (Friestad and Wright 1994; Ham and Nelson 2019). People can access and use this knowledge to cope with persuasive attempts, which includes evaluating whether the goals of the persuader align with their own goals, and thus whether to accept or resist the persuasive message (Friestad and Wright 1994). Resistance is defined as ‘a motivational state, in which consumers have the goal to reduce attitudinal or behavioral change or to retain one’s current attitude’ (Fransen et al. 2015, 7). People can resist advertising using several strategies: avoidance, contesting, and empowerment (Fransen et al. 2015). We propose that consumers may resist SA by contesting it, which involves actively refuting the ad by challenging its message and the persuasive tactic that is used (Fransen et al. 2015). According to reactance theory, when consumers perceive that their freedom is limited by attempts to persuade them, they experience psychological reactance, which motivates them to resist attempts to restore their threatened freedom (Brehm and Brehm 1981). Although the PKM (Friestad and Wright 1994) emphasizes that merely the use of persuasion knowledge does not invariably means that consumers will resist a persuasive message, based on the theory of psychological reactance (Brehm and Brehm 1981), it is often assumed that consumers tend to resist messages that are perceived as intending to direct or control one’s choices (Sagarin et al. 2002).

Applying the PKM and reactance theory to the context of SA means that whether consumers are aware of SA as a persuasive tactic, and their level of factual knowledge of SA and its data collection practices, influences their responses to it. Recent research shows that consumers have limited knowledge about synced advertising compared to other personalized advertising strategies, such as online behavioral advertising (Segijn and van Ooijen 2020b). We expect that consumers that understand the practice of SA can access this knowledge and evaluate whether they believe that the tactic of using data about their media behavior is acceptable, aligns with their own goals, and whether it controls their choices, and thus decide whether to resist SA. Because practices such as SA require personal data, we expect that an evaluation of the use of these data is part of this coping behavior. Therefore, the current study looks into whether and how individuals’ knowledge of SA (i.e., awareness of SA as a persuasive tactic in Study 1, and objective and subjective knowledge about SA in Study 2) can increase resistance in consumers through critical attitudes and perceived surveillance.

Critical attitudes as an underlying mechanism of resistance

Although being aware of SA does not inherently mean that SA will also be resisted, we expect that higher awareness of the practice may induce critical attitudes toward SA, which will result in the resistance of a synced ad. In this study, we conceptualize critical attitudes toward SA as consumers’ critical feelings about advertising, such as skepticism toward or disbelief of advertising in general (Boerman, Van Reijmersdal, and Neijens 2012). Following the change-of-meaning principle (Friestad and Wright 1994), we expect that becoming aware of the actual persuasive tactic and goal of a message, can fundamentally alter the meaning of this message and thus the way a person responds to it. We therefore propose that when consumers are made aware of SA and the data collection practice of SA, they may develop critical attitudes toward synced advertising in general.

Scholars have proposed that critical attitudes toward advertising successfully alter consumer responses to advertising (Rozendaal et al. 2011). Given their insight into advertising, critical attitudes may promote consumers’ reflective thinking about advertising in general, which helps consumers critically process advertising messages rather than passively accept them. For example, research shows that adolescents who hold critical attitudes toward advertising experience more negative feelings when they are exposed to a specific ad, which leads to certain resistance strategies (Vanwesenbeeck, Walrave, and Ponnet 2016). Similar results have been found among adult consumers. Research shows that when adult consumers recognize the persuasive intent of embedded ads (e.g., brand placement and influencer marketing), they are motivated to critically process ads, which results in the development of negative feelings or attitudes toward the ads (e.g., Boerman, Willemsen, and Van Der Aa 2017; De Veirman and Hudders 2020; Eisend et al. 2020). Based on the discussion above, it is predicted that high levels of SA awareness will generate consumers’ critical attitudes toward SA in general, which makes consumers more resistant to a synced ad message. To this end, we formulated the following hypothesis (Figure 1):

H1: SA awareness and critical attitude toward SA serially mediate the relationship between providing consumers with SA information (vs. no information) and resistance, such that providing consumers with SA information (vs. no information) will increase SA awareness (H1a), which in turn will generate more critical attitudes toward synced advertising in general (H1b), which will lead to more resistance to a synced ad (H1c).

Figure 1. Conceptual model. Note: In Study 2, we examined the effect of technical, personally relevant, or no information on SA objective and subjective knowledge (RQ1). The rest of the model stayed the same.

Perceived surveillance as an underlying mechanism of resistance

SA is an advanced form of personalized advertising that targets the audience with personalized content based on real-time media behavior (Segijn 2019). It extends personalization to the offline sphere by personalizing ads based on, for example, what a consumer is currently watching on TV or listening to on the radio. However, as mentioned earlier, consumers’ knowledge about SA and its data collection techniques is limited (Segijn and van Ooijen 2020b) and according to the PKM, informing consumers about advertising techniques could increase their awareness of this technique (Friestad and Wright 1994). We argue that SA awareness can lead to perceived surveillance. Surveillance is defined as ‘the watching, listening to, or recording of an individual’s activities’ (Solove 2006, 490), and the perception of being watched is called perceived surveillance (Xu et al. 2012).

When consumers feel and recognize that the ad contains personally identifiable information, they might feel concerned about their personal data (White et al. 2008). The availability of consumers’ online and offline data increases the use of these data as input for messages, which increases the feelings of being watched (Duffy and Chan 2019). Research found that awareness of data collection practices for this type of personalization increase perceptions of surveillance (Segijn and van Ooijen 2020a). Additionally, when consumers feel that they are being watched and surveilled, it may limit their perceptions of freedom and autonomy and therefore lead to resistance. Indeed, Farman, Comello, and Edwards (2020) found that in the context of personalization based on past online behavior, perceived surveillance could lead to more resistance to personalized advertising. Based on previous research and in line with reactance theory (Brehm and Brehm 1981) discussed earlier, we suggest that SA awareness increases consumers’ sense that they are being watched, which increases their resistance to SA messages. To this end, we formulate the following hypothesis (Figure 1):

H2: SA awareness and perceived surveillance toward SA serially mediate the relationship between providing consumers with SA information (vs. no information) and resistance, such that providing consumers with SA information (vs. no information) will increase SA awareness, which in turn will increase levels of perceived surveillance (H2a), which in turn will lead to more resistance to a synced ad (H2c).

Method Study 1

Design and sample

We conducted an online experiment with a single factor (synced advertising (SA) information vs. no SA information) between-subjects design. We aimed to inform participants on SA by asking them to read a website text. Participants in the SA information condition were presented with a text about SA, while those in the no SA information condition were shown a text about a different advertising strategy to keep advertising priming consistent across conditions. Based on a pretest (see below), we selected online behavioral advertising (OBA) as the alternative strategy, which is a personalized advertising strategy based on past online behavior, such as searches, clicks, and likes (Boerman, Kruikemeier, and Zuiderveen Borgesius 2017).

The sample consisted of 93 U.S. adults recruited via Amazon MTurk. After excluding 12 respondents who failed the attention check, reported not viewing the stimuli, or noticed the true purpose of this study before debriefing, 81 usable responses remained (M_age = 35.2, SD_age = 10.9, 49.4% female). The participants received compensation of $1.50 for their participation.

Pretest website text

Two pretests were conducted to select an advertising strategy to be used for the second website text. In the first pretest, four advertising strategies (i.e., synced advertising, online behavioral advertising (OBA), native advertising, and email marketing) were tested to find advertising strategies in which consumers had similar perceptions as SA (N = 34; M_age = 35.56, SD_age = 11.69; 35.3% female). After the pretest, email marketing was excluded because it was significantly different from synced advertising in terms of familiarity (p < .001), resistance (p < .001), and creepiness (p < .001). In a second pretest (N = 77; M_age = 38.27, SD_age = 12.11; 39.0% female), we developed three texts explaining the three remaining advertising strategies. We selected OBA because it was not significantly different from SA across the measures (i.e., knowledge, perceived surveillance, resistance). Moreover, participants were able to distinguish between SA and OBA when they were exposed to a short clip of a SA case and asked to what extent the clip was an example of the strategy they had just read about (p < .001).

Procedure

We used a cover story stating that participants would take part in two separate studies with different purposes, where participants were asked to evaluate the clarity and design of a website page to be launched soon in ‘Study 1ʹ (We will call this Study 1, Part A hereafter), and to read a media multitasking scenario in ‘Study 2ʹ (We will call this Study 1, Part B hereafter). By doing so, we tried to ensure that participants would not notice that these two studies were part of the same study (Figure 2). Only one participant indicated having noticed the true purpose of the two studies in the cover story check, which led to the exclusion of this participant from the data analysis of this study.

Figure 2. Flowchart of Study 1.

In Study 1, Part A, in which the participants were either informed on SA or not, participants were exposed to an image of a website. Upon consent, participants were randomly assigned to one of the two conditions. The website text provided a brief description of the ad strategy (SA or OBA), explained the technology behind it, illustrated how it works with a concrete example, and explained some advantages of the strategy for advertisers. The two texts were kept as similar as possible across the conditions. After looking at the website image, the participants were asked to evaluate the texts and design of the website for clarity and design, which were filler questions.

In Study 1, Part B, which directly followed after the filler questions of Study 1, Part A, all the participants were presented with an identical media use scenario in which they were asked to imagine themselves watching a TV show on a television while browsing a mobile app on a tablet at the same time. The media scenario described a situation where a mobile banner ad popped up on the participant’s tablet, and the ad promoted a product mentioned in the TV show that they were watching simultaneously. An image of a mobile banner ad was displayed at the bottom of the media scenario. The brand was selected based on a pretest in which six brands were tested in terms of familiarity and brand attitudes (N = 39; M_age = 38.80, SD_age = 11.59; 30.8% female). In line with recommendations for advertising research (Geuens and De Pelsmacker 2017), we used a brand name that the consumers were not familiar with and had a neutral attitude toward. After looking at the media scenario and the banner ad, participants were asked to fill out an online questionnaire.

Measurements

To measure awareness of SA as persuasive tactic, we asked participants to what extent they agreed with three statements, starting with ‘In real life, I believe that companies …’, followed by 1) ‘collect information about my TV viewing habits’, 2) ‘use information about my TV viewing habits to show me advertisements’, and 3) ‘share information about my TV viewing habits with other companies’. (1 = totally disagree, 7 = totally agree). The three items focusing on consumers’ understanding of the collection usage and sharing of personal information to personalize advertising were based upon items from Ham (2017) and Boerman, Kruikemeier, and Zuiderveen Borgesius (2021). The three items were averaged into one measure of SA awareness (Cronbach’s alpha = 0.95; M = 5.70, SD = 1.37).

Critical attitude toward SA was measured by asking participants to indicate the extent to which they agreed that the presence of the brand on their mobile device while watching TV was 1) honest, 2) bad, 3) good, and 4) wrong (1 = totally disagree, 7 = totally agree) (van Reijmersdal et al. 2020). The items of ‘honest’ and ‘good’ were reversely coded, and all the items were averaged to create a single measure of critical attitude (Cronbach’s alpha = 0.80; M = 4.19, SD = 1.44).

We measured resistance with three items asking whether people contested the stimulus banned ad: 1) ‘I criticized the ad’, 2) ‘I thought up points that went against the ad’, and 3) ‘I was skeptical of the ad’ (1 = totally disagree, 7 = totally agree) (Silvia 2006). The three items were averaged into a single scale of resisting the banner ad (Cronbach’s alpha = 0.86; M = 4.56, SD = 1.76).

To measure perceived surveillance, participants were asked to indicate the extent to which they agreed with four statements starting with ‘When I imagined seeing the banner ad while watching TV, I had the feeling that advertisers were … ’ followed by 1) ‘looking over my shoulder’, 2) ‘entering my private space’, 3) ‘watching my every move’, and 4) ‘checking up on me’ (1 = totally disagree, 7 = totally agree) (Segijn, Opree, and van Ooijen, 2021). The items were averaged into a single scale of perceived surveillance (Cronbach’s alpha = 0.95; M = 5.35, SD = 1.97).

Result Study 1

To test the first hypothesis, we used PROCESS model 6 by Hayes (2013) with 5,000 bootstraps with the conditions (SA information vs. no SA information) as the independent variable, SA awareness and critical attitudes as the serial mediators, and resistance as the dependent variable. An overview of the results can be found in Table 1. The results showed no significant direct effect of the conditions on resistance to the synced ad (direct effect = −.17, SE = .32, 95% BCBCI [−0.81, 0.48]) and no significant indirect effect (indirect effect = −0.02., SE = .06, 95% BCBCI [−0.16, 0.07]). Despite the extensive pretesting, we did not find a significant difference between the two conditions (SA information vs. no SA information) on SA awareness (b = −0.13, p = .683). However, we found that a higher level of SA awareness led to more critical attitudes (b = .41, p < .001) and resistance to the synced ad (b = .48, p < .001). Additionally, as predicted, more critical attitudes led to more resistance to the synced ad (b = 0.40, p = .002).

Table 1. Serial mediation effects of SA information on resistance through SA knowledge and critical attitudes.

	Sample	SA knowledge measurement	Condition (reference group)	Information -> SA knowledge	SA knowledge -> Critical Attitude	Critical attitude -> Resistance	(Relative) Direct effect	(Relative) Indirect effect
Study 1
	All participants (N = 81)	SA awareness	SA information (vs. no SA information)	−0.13	0.41	0.40	−0.17 (0.32)	−0.02 (0.06) CI [−0.16, 0.07]
	No prior experience (n = 54)	SA awareness	SA information (vs. no SA information)	−0.48	0.55	0.84	0.24 (0.30)	−0.23 (0.18) CI [−0.59, 0.11]
Study 2
	All participants (N = 150)	Objective knowledge	Technical (control)	0.45	0.14	0.62	−0.06 (0.20)	0.04 (0.03) CI [−0.02, 0.11]
			Personal relevant (control)	0.19	0.14	0.62	0.33 (0.20)	0.02 (0.03) CI [−0.05, 0.08]
			Personal relevant (technical)	−0.26	0.14	0.62	0.39 (0.20)	−0.02 (0.03) CI [−0.11, 0.03]
		Subjective knowledge	Technical (control)	0.66	0.03	0.63	−0.04 (0.20)	0.01 (0.03) CI [−0.04, 0.07]
			Personal relevant (control)	0.64	0.03	0.63	0.32 (0.20)	0.01 (0.03) CI [−0.05, 0.06]
			Personal relevant (technical)	−0.02	0.03	0.63	0.37 (0.20)	−0.00 (0.02) CI [−0.05, 0.03]
	No prior experience (n = 82)	Objective knowledge	Technical (control)	1.06	0.18	0.65	−0.37 (0.29)	0.13 (0.08) CI [−0.02, 0.30]
			Personal relevant (control)	0.68	0.18	0.65	0.15 (0.29)	0.08 (0.07) CI [−0.05, 0.24]
			Personal relevant (technical)	−0.38	0.18	0.65	0.52 (0.27)	−0.05 (0.05) CI [−0.17, 0.05]
		Subjective knowledge	Technical (control)	1.32	0.02	0.68	−0.32 (0.30)	0.02 (0.07) CI [−0.15, 0.16]
			Personal relevant (control)	1.50	0.02	0.68	0.17 (0.30)	0.02 (0.08) CI [−0.18, 0.17]
			Personal relevant (technical)	0.18	0.02	0.68	0.49 (0.27)	0.00 (0.03) CI [−0.07, 0.06]

Note: Table presents unstandardized b-coefficients with bootstrap standard errors in parentheses. CI = 95% bootstrap confidence intervals. Significant relationships are bold.

To test the second hypothesis, we used PROCESS model 6 by Hayes (2013) with the conditions (SA information vs. no SA information) as the independent variable, SA awareness and perceived surveillance as the serial mediators, and resistance as the dependent variable. An overview of the results can be found in Table 2. Again, we did not find significant direct or indirect effects (direct effect = −.37, SE = .32, 95% BCBCI [−1.02, 0.27]; indirect effect = −0.03, SE = .09, 95% BCBCI [−0.20, 0.17]). However, we found that a higher level of SA awareness led to more perceived surveillance (b = .73, p < .001) and resistance to the synced ad (b = .39, p = .008). As predicted, more perceived surveillance led to more resistance to the synced ad (b = 0.35, p = .002).

Table 2. Serial mediation effects of SA information on resistance through SA knowledge and perceived surveillance.

	Sample	SA knowledge measurement	Condition (reference group)	Information -> SA knowledge	SA knowledge -> Perceived surveillance	Perceived surveillance -> Resistance	(Relative) Direct effect	(Relative) Indirect effect
Study 1
	All participants (N = 81)	SA awareness	SA information (vs. no SA information)	−0.13	0.73	0.35	−0.37 (0.32)	−0.03 (0.09) CI [−0.21, 0.17]
	No prior experience (N = 54)	SA awareness	SA information (vs. no SA information)	−0.48	0.75	0.33	−0.09 (0.40)	−0.12 (0.12) CI [−0.42, 0.08]
Study 2
	All participants (N = 150)	Objective knowledge	Technical (control)	0.45	0.25	0.31	−0.02 (0.24)	0.03 (0.03) CI [−0.01, 0.10]
			Personal relevant (control)	0.19	0.25	0.31	0.10 (0.23)	0.01 (0.03) CI [−0.04, 0.07]
			Personal relevant (technical)	−0.26	0.25	0.31	0.13 (0.23)	−0.02 (0.03) CI [−0.09, 0.02]
		Subjective knowledge	Technical (control)	0.66	0.15	0.33	0.02 (0.24)	0.03 (0.03) CI [−0.01, 0.09]
			Personal relevant (control)	0.64	0.15	0.33	0.11 (0.23)	0.03 (0.03) CI [−0.01, 0.09]
			Personal relevant (technical)	−0.02	0.15	0.33	0.09 (0.23)	−0.00 (0.02) CI [−0.05, 0.04]
	No prior experience (N = 82)	Objective knowledge	Technical (control)	1.06	0.33	0.28	−0.28 (0.35)	0.10 (0.07) CI [−0.02, 0.26]
			Personal relevant (control)	0.68	0.33	0.28	−0.05 (0.34)	0.06 (0.06) CI [−0.04, 0.19]
			Personal relevant (technical)	−0.38	0.33	0.28	0.23 (0.32)	−0.04 (0.05) CI [−0.15, 0.03]
		Subjective knowledge	Technical (control)	1.32	0.22	0.35	−0.13 (0.35)	0.10 (0.07) CI [−0.01, 0.27]
			Personal relevant (control)	1.50	0.22	0.35	0.05 (0.36)	0.12 (0.08) CI [−0.01, 0.28]
			Personal relevant (technical)	0.18	0.22	0.35	0.18 (0.32)	0.01 (0.04) CI [−0.09, 0.10]

Note: Table presents unstandardized b-coefficients with bootstrap standard errors in parentheses. CI = 95% bootstrap confidence intervals. Significant relationships are bold.

Additionally, we observed that some participants reported having previous experience with advertising or synced advertising through classes, lectures, work experience or previous research participation, which could have explained the nonsignificant findings of the websites on SA awareness. Therefore, although not hypothesized, we wanted to further explore the relationships within this subsample. Fifty-four participants (M_age = 38.3, SD_age = 11.9, 42.6% female) had no previous SA experience. However, running the model with this subsample did not change the results for the model with critical attitudes (Table 1) nor for the model with perceived surveillance (Table 2).

Discussion Study 1

Study 1 showed that higher levels of SA awareness could indeed lead to more perceived surveillance and resistance. However, informing participants with the SA website did not lead to a greater awareness of SA as a persuasive tactic compared to not informing the participants on SA (OBA website). Therefore, H1b, H1c, H2a, and H2b were supported but H1a was not supported. That the information did not increase SA awareness could be explained by the SA awareness measurement or the information used.

First, the SA awareness measurement may tap into consumers’ awareness of collection, usage, and sharing of media behavior data by organizations rather than about SA specifically. The items asked whether consumers believed that companies collect information on their TV viewing habits for personalized advertising (e.g., ‘In real life, I believe that companies use information about my TV viewing habits to show me advertisements’). This might explain the nonsignificant result and the high overall means (SA M = 5.63, SD = 1.46; OBA M = 5.76, SD = 1.29), which is significantly above the midpoint of the scale for both SA (t (41) = 7.27, p < .001) and OBA (t (38) = 8.52, p < .001), given that OBA also uses data as input for their message. Additionally, the measurement asks about beliefs rather than actual awareness. Finally, we did not include a group that did not receive any information; therefore, we cannot test whether SA awareness was increased in both groups or not at all.

Second, the information on the website might have been too technical or too general, or participants may not have perceived it as being personally relevant. Previous research has shown that providing technical information about personalization does not increase consumers’ perceived susceptibility or motivation to engage in privacy protection behaviors (Strycharz et al. 2019, 2021). Following the impersonal impact hypothesis (Tyler and Cook 1984), one could expect that vivid, more personally relevant information could be more effective than general, technical information. Additionally, the way the information was presented (i.e., on a website with a cover story) might have been too subtle for an online experiment. Participants may have just looked at the design of the website image rather than carefully reading the website text.

To account for these limitations, we conducted a follow-up study. First, we included a more detailed, explicit measure of participants’ knowledge of SA than merely awareness, by using factual knowledge statements as a measurement of objective SA knowledge, and participant’s confidence in their answers as a measure of their subjective SA knowledge, in line with previous research (Segijn and van Ooijen 2020b). Second, we split up the text of Study 1 into technical and personally relevant information to examine whether it matters what information is presented. Thus, next to retesting the hypotheses (H1 and H2) with SA knowledge instead of SA awareness, we ask whether it matters what type of information is most effective in increasing SA knowledge. Also, we presented the information without a cover story to increase the likelihood that participants would read the text. Finally, we added a control group as a baseline measure to test whether presenting information could increase SA knowledge. To this end, we ask:

RQ1: To what extent does the type of information (technical vs. personally relevant vs. no information) affect objective and subjective SA knowledge?

Method Study 2

Design, sample, and procedure

An online experiment was conducted with a single factor (type of information: technical, personally relevant, or no SA information) between-subjects design. A sample of 155 U.S. adults were recruited via Prolific. Participants (n = 5) who did not pass the attention check (i.e., selected other options than what they were asked to click on and speeders) were excluded, which led to a final sample of 150 U.S. adults (M_age = 26.78, SD_age = 8.26, 43.3% female).

Participants were randomly assigned to one of the three conditions where they were presented with 1) technical information about SA (n = 49), 2) a personally relevant SA text (n = 52), or 3) no information (n = 49, control group). After being exposed to the assigned SA text, the participants were asked to answer questions about their SA knowledge, meaning that participants in the control group were asked to answer SA knowledge questions right after their consent on this study without being exposed to any information. Then, all the participants were shown a short SA scenario and asked to fill out the questionnaire (see Figure 3 for an overview of the procedure). The SA information presented in Figure 4. Finally, the participants were debriefed and compensated with $1.90.

Figure 3. Flowchart of Study 2.

Figure 4. Different types of synced advertising information that was only shown to the participants in the respective conditions (Study 2).

a) Technical SA information b) Personal relevant SA information.

Stimuli

For Study 2, textual stimuli were used for the three conditions. First, the technical SA information focused on how technology of SA works, specifically about how consumer data are collected and utilized to send ad messages relevant to the TV programs that media users are watching. Second, the personal relevant SA information describes a specific situation where media users encounter synced ads, leaving participants to visualize it happening to themselves. Last, the participants in the control group were not exposed to any texts. Similar to Study 1, all the participants read a short SA media scenario. The SA media scenario described a situation where participants encountered a mobile ad that was synchronized with the TV content that they were watching concurrently.

Measures

Overall, the same measurements as in Study 1 were used in this study, except for the measurement of SA literacy. In Study 2, we measured both objective and subjective SA knowledge, which are both considered distinctive and play different roles in problem solving (Raju, Lonial, and Mangold 1995). Therefore, we measured objective knowledge and subjective knowledge by asking participants’ thoughts about the eight statements about SA. The statements were adapted from Segijn and van Ooijen (2020b). The statements about SA knowledge were presented in a random order.

Objective knowledge, which refers to factual knowledge (Raju, Lonial, and Mangold 1995), was assessed using an objective test about SA. Participants were asked to indicate whether each of the SA statements (e.g., ‘Companies can advertise on one device based on information collected through another device at the same time’) they believed to be true or false (Table 3). Correctly answered items were coded 1, and incorrect answers were coded 0 to calculate a sum score.

Table 3. SA knowledge statements (Study 2).

		Objective knowledge			Subjective knowledge
Statements	n	T/F	%		M	SD
The ads that people receive on their mobile device might differ depending on what other media content (TV, radio) is simultaneously consumed.	150	T	85.3		3.82	3.90
Companies know what people are watching/listening to because media content (TV/radio shows) sometimes contain a sound signal that can be picked up by a mobile device.	150	T	60.0		1.17	4.74
Companies can advertise on one device based on information collected through another device at the same time.	150^a	T	88.7		4.87	2.83
Technology already exists that makes it possible to receive ads on your smartphone based on your current (real-time) watching behavior on TV.	150	T	89.3		4.52	3.40
A company can show me an ad on my mobile device from a brand at the same time that I am watching a television commercial from that brand.	150	T	83.3		3.68	3.87
It is a coincidence when people receive an ad on their mobile device that is related to what they are concurrently listening to on the radio.	150	F	85.3		3.69	3.82
It is impossible for a mobile app to listen to a television show that people are watching, and use this information to provide those people with ads based on the show’s content.	149	F	82.6		3.07	3.74
It is impossible for companies to collect information about the shows that people watch on television, and simultaneously advertise relevant products/brands on those people’s mobile devices	150	F	86.0		3.33	3.59

^a n = 143 for subjective knowledge due to missing data.

Subjective knowledge was operationalized as confidence in knowledge (Park and Lessig 1981; Raju, Lonial, and Mangold 1995) and was measured by asking participants how confident they were in their previous answers to the true or false questions. Following the procedures by Segijn and van Ooijen (2020b), we recoded objective knowledge into correct (1) and incorrect (−1) and then multiplied it by the confidence score (1 = extremely unconfident, 7 = extremely confident). Thus, more negative values indicate more false confidence, while more positive values indicate more true confidence.

Result Study 2

To test H1 and H2 and answer RQ1, we used PROCESS model 6 by Hayes (2013) with 5,000 bootstraps, one for each type of knowledge. To test the first hypothesis, we used the type of information as the independent variable, knowledge and critical attitudes as subsequent mediators, and resistance as the dependent variable (Table 1). We used dummy coding for the independent variable with the control condition as the reference group.

The results showed that neither technical (b = 45, p = .147) nor personally relevant (b = 0.19, p = .534) information led to more objective SA knowledge than no information. However, we observed that objective knowledge led to more critical attitudes (b = 0.14, p = 0.30), which in turn affected resistance (b = 0.62, p < .001). Regarding the model for subjective knowledge, we also did not find that technical (b = 0.66, p = .127) or personally relevant information (b = .64, p = .131) led to more subjective knowledge compared to no information. In addition, subjective knowledge had no significant relationship with critical attitudes (b = 0.03, p = .473).

For the model with perceived surveillance (Table 2), the results also showed that neither technical (b = 0.45, p = .146) nor personal relevant (b = 0.19, p = .534) information led to more objective SA knowledge compared to no information. Similarly, we did not find that technical (b = 0.66, p = .127) or personally relevant information (b = 0.64, p = .131) led to more subjective knowledge than no information. However, we observed that both objective (b = 0.25, p = .001) and subjective knowledge (b = 0.15, p = .006) led to more perceived surveillance, which led to more resistance to a synced ad (b = 0.31–0.33, p < .001).

Similar to Study 1, we had several participants without any prior SA experience. Focusing on this subsample (n = 82, M_age = 27.80, SD_age = 8.11, 47.6% female), we observed a difference in the results. For participants without any prior SA experience, the results showed that technical information led to more objective SA knowledge compared to the no information condition (b = 1.06, p = .015); however, personally relevant information did not (b = 0.68, p = .113). No difference in objective SA knowledge was found depending on whether technical or personally relevant information was presented (b = −0.38, p = .354). Additionally, objective SA knowledge led to more critical attitudes (b = 0.18, p = .036), which in turn led to more resistance (b = 0.65, p < .001). Similarly, objective SA knowledge led to more perceived surveillance (b = 0.33, p = .002), which subsequently led to more resistance to the synced ad (b = 0.28, p = .007). Despite the significant individual paths, we did not observe a significant indirect effect for technical information (Tables 1, 2). Additionally, both personal relevant (b = 1.50, p = .011) and technical information (b = 1.32, p = .025) led to more subjective SA knowledge compared to no information, but again they did not differ from each other (b = 0.18, p = .749). Additionally, in this sample, subjective SA knowledge did not lead to more critical attitudes (b = 0.02, p = .796). However, subjective SA knowledge led to more perceived surveillance (b = 0.22, p = .004), which subsequently led to more resistance to the synced ad (b = 0.35, p = .001). Similar to the previous model, we did not observe a significant indirect effect (Tables 1, 2).

Discussion Study 2

The results of Study 2 showed that providing information can lead to more SA knowledge (RQ1), which increases critical attitudes (H1b) and perceived surveillance (H2a); this in turn subsequently increases resistance to the synced ad (H1c and H2b), but does so only for consumers without prior synced advertising experience (i.e., through classes, lectures, work experience or previous research participation). Given that the relations between all variables in the mediation are significant but the chain of mediation in total is not, we have some support for H1 and H2 for the subsample only. For this subgroup, technical information increases both objective and subjective knowledge, and personally relevant information increases subjective knowledge (i.e., confidence in knowledge) only (RQ1). Additionally, we found that objective knowledge increases both critical attitudes and perceived surveillance, which subsequently increases resistance to the synced ad. Subjective knowledge does not increase critical attitudes but increases perceived surveillance, which subsequently increases resistance to the synced ad. We should note that despite the significant individual paths, we do not observe a significant indirect effect. This means that we have some evidence for the individual paths but no strong evidence for the chain of mediation, which could be the result of a lack of power in the subsample (n = 82).

General discussion

Technology has made it possible to concurrently deliver personalized messages to consumers’ mobile devices based on their offline media usage, which is known as synced advertising. The developments of more sophisticated personalization techniques go hand-in-hand with a rise in concerns related to consumers’ privacy (Daems, De Pelsmacker, and Moons 2019; Yun et al. 2020). That is why there is a need to empower consumers through literacy programs that will help consumers become more aware of synced advertising and help them become informed decision makers. In two online experiments, we showed that increasing awareness and knowledge of SA results in more critical attitudes and perceived surveillance, which leads to more resistance to the synced advertisement.

Additionally, we found that it matters what type of information is presented to increase SA objective or subjective knowledge in consumers without prior SA experience. Presenting these consumers with technical information leads to an increase in both objective and subjective SA knowledge compared to no information. Personal relevant information only helped to increase SA subjective knowledge compared to no exposure to information. This difference goes against the impersonal impact hypothesis (Tyler and Cook 1984) and previous findings (Strycharz et al. 2019, 2021). A possible explanation could be that in the absence of any experience or prior knowledge, information on how the SA strategy works (i.e., technical information) might be helpful. However, when consumers already have (a basic) understanding, technical information might not contribute much. Future research is needed to further examine this relationship.

Although personal information does not increase objective knowledge (i.e., factual knowledge), providing inexperienced consumers with personally relevant information could make them more confident in their knowledge toward SA, which is enough to evoke perceived surveillance and resistance to a synced ad. We found evidence for the relationship between these variables, but we did not find a significant indirect mediation effect, which could be explained by the lack of power in the subsample. Therefore, future research with a larger sample is needed to validate the claims made. Moreover, it is recommended that future research use ‘experience in advertising’ as a selection criterion for the sample to ensure enough power. Another limitation of this study is the use of scenarios rather than an actual synced advertising experience. The nature of this manipulation could have increased the awareness of synced advertising in Study 2. On the other hand, the perceived surveillance might have been lower than when one would receive a synced ad in real life. Future research (e.g., lab experiment, observation study) is needed to examine these possibilities. Given that this is a first step in examining how to increase SA knowledge, these results serve as a good first indication and stepping-stone for future research.

The results of the study have important implications for theory. First, they advance theory on a relatively new personalized advertising strategy, namely, synced advertising. Research on this topic is still in its infancy. First studies have explored the effectiveness of this strategy (e.g., Hoeck and Spann 2020; Segijn and Voorveld 2021) and consumer knowledge of this strategy (Segijn and van Ooijen 2020b). However, research studying whether and how consumers will resist this strategy is lacking. To our knowledge, this is the first study to examine advertising literacy in the context of synced advertising, which advances theory by building on synced advertising from a consumer empowerment perspective. Specifically, we focused on the combination of TV with a mobile device in the current study. However, synced advertising could also involve synchronizing an ad with other (offline) media, such as radio or outdoor advertising. Future research is necessary to validate the claims for other media combinations as well.

Second, the study advances theory in the field of persuasion knowledge by applying existing theories (i.e., reactance theory (Brehm and Brehm 1981) and persuasion knowledge model (Friestad and Wright 1994)) to a new context. In line with previous research (Vanwesenbeeck, Walrave, and Ponnet 2016), this study confirms the role of critical attitudes in informed consumers. Moreover, we advance theory by adding a new explanatory factor of the relationship between SA knowledge and resistance to the synchronized ad, namely, perceived surveillance. Perceived surveillance might be specific to personalized advertising because of the data collection and privacy concerns involved. Future research is needed to examine whether this also applies to other forms of (personalized) advertising.

The current study sets the ground for future research on the topic. For example, future studies could further examine whether increasing SA knowledge also affects other advertising outcomes, such as brand attitudes and purchase intention. Farman, Comello, and Edwards (2020) showed that, in the context of OBA, perceived surveillance led to more resistance, which negatively affected attitudes towards the ad, and subsequently purchase intentions. Future research could study to what extent information consumers on SA practices (indirectly) affects attitudes towards the brand or ad and purchase intention. Moreover, as our study shows that critical attitudes play an important role in the effects of knowledge about SA on responses to SA, future research could investigate these evaluative dynamics further, for instance by examining the role of ad skepticism (Obermiller and Spangenberg 1998) and other attitudinal dimensions of persuasion knowledge (Boerman et al. 2018). Additionally, it was found that privacy concerns play an important role in synced advertising effects (Author, 2021). Future research could further examine the effects of SA awareness and knowledge on advertising outcomes by including privacy concerns as moderator, as well as other variables such as general attitude towards online advertising (For an overview see Segijn 2019).

Finally, this study provides important practical implications for literacy programs and government regulations. To increase consumer empowerment in the context of SA, educational institutes should pay attention to literacy programs to improve consumers’ knowledge of synced advertising strategies and their data collection practice. Our findings indicate that increased knowledge about personalization tactics such as SA can enable consumers to become informed decision makers. Knowledge of this new format of advertising increases the critical processing of SA messages, enabling consumers to decide whether to accept or reject them. We find that increased synced advertising knowledge and confidence in this knowledge help consumers resist synced ads. This means that increasing and activating consumers’ knowledge about this advertising strategy can increase consumer empowerment and help consumers respond to synced ads accordingly. Specifically, literacy programs should provide technical information on how synced advertising works when they want to increase objective knowledge. Moreover and in line with the findings by Segijn and van Ooijen (2020b), we found that objective knowledge of the statement regarding the personalization technique ‘watermarking’ is the lowest (Table 3), which emphasizes the need for explaining how this technique works. Additionally, it might be beneficial for governments to require advertisers to disclose how synced ad messages are presented to consumers. Disclosures have been shown to be an effective tool to help consumers recognize and understand new types of advertising, such as sponsored content, native advertising, and influencer marketing (e.g., Boerman and Van Reijmersdal 2020; Eisend et al. 2020; Evans et al. 2017).

With the development of digital technology and the growing needs for personalized communication, synced advertising is expected to become more prevalent than ever before. Because of the novelty of this new technique, many consumers still do not have a great deal of knowledge about this advertising strategy (Segijn and van Ooijen 2020b). Information on SA can make consumers more aware and more knowledgeable about this technique, which leads to the increased feeling of being watched (‘Big Brother’ is watching). This study is meaningful in that it examines what type of information increases SA knowledge and how this affects resistance toward synchronized ads through either critical attitudes or perceived surveillance.

Acknowledgement

The authors would like to thank Nathan Caspar for the design of the stimuli of Study 1 and Abbey Hammell for the help with the stimuli presentation in the survey.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported by the Office of the Vice President for Research, University of Minnesota [The Grant-in-Aid of Research, Artistry, and Scholarship].

Notes on contributors

Claire M. Segijn

Claire M. Segijn (PhD, University of Amsterdam) is an Assistant Professor of advertising at the Hubbard School of Journalism and Mass Communication, University of Minnesota – Twin Cities. Her research includes the simultaneous usage of multiple media (e.g. multiscreening, synced advertising) and how this affects information processing and message effectiveness.

Eunah Kim

Eunah Kim (MA, Michigan State University) is a PhD Candidate at the Hubbard School of Journalism and Mass Communication, University of Minnesota – Twin Cities. Her research interest lies in consumer engagement with digital advertising, particularly in the context of social media and personalized ads.

Asma Sifaoui

Asma Sifaoui (MA, University of Minnesota) is a PhD student at the Stan Richards School of Advertising & Public Relations at the Moody College of Communication, University of Texas-Austin. Her research interests are in computational advertising, personalized advertising, and social media advertising with a focus on privacy issues and surveillance. She collaborated on the work while finishing her MA at the University of Minnesota.

Sophie C. Boerman

Sophie C. Boerman (PhD, University of Amsterdam) is an Assistant professor of Persuasive Communication at the Amsterdam School of Communication Research (ASCoR) at the University van Amsterdam. Her research addresses empowerment and resilience in the context of persuasive and/or data-driven communication, with a focus on transparency, literacy and persuasion knowledge, and privacy.